Luring (was: DCSS has a power creep problem)

[zxc23]

One of the defining characteristics of this game for me is how fair it is. Most other roguelikes will simply destroy you in some games even if you played optimally. Many of the suggestions in this thread, if employed in isolation, will increase the randomness of the early game. I've been streaking for the last year now and I have become acutely aware of how great a factor RNG is.

The D1 entrance vault is a dice roll, as some provide no room / no pillars. The first fight can be overwhelming with no room to retreat or pillar dance, leading to unavoidable deaths. Adders can spawn on D1, which are entirely a gamble for most characters on D1, skill playing a miniscule part. A gnoll on D1 is usually untouchable, and will tend to force a premature search for the downstairs, which can lead to a situation impossible to deal with due to a lack of consumables and HP from XL. On entering D2, Sigmund, Grinder, orc wizards, orc priests, gnolls and adders all can spell instant death. I am reminded of Zooty's streak character in the 0.16 tournament which entered D2 right next to Sigmund; he died after missing the first curare and while the second hit, he couldn't survive long enough. Then there are shafts, tele traps, first turn confuse/paralysis/banishment, chain confusion, enemies with wands that can 1-2 shot you in the early dungeon... There is a lot of RNG in crawl if you play to win every single game.

Luring is a tactic that I employ constantly to reduce the element of RNG and to control situations better. I doubt anyone utilises luring as much as I do. I don't find it tedious; it's a necessity to stop fights from being a complete gamble. By playing for score or dur instead, luring is reduced.

Thing one is pack breaking, crawl acutally has a specific definition of a pack, a pack spawns together, and has special logic to try to stay together,

Similarly there is also "breaking apart arbitrary groups of critters who happened to be closer to each other" this is actually fairly similar to pack breaking, but it is slightly easier since the critters aren't actively trying to stay together. They are just all chasing you.

a third thing that happens is "drawing critters into cleared areas so as to minimize noise and risk of attracting things that aren't in the immediate area"

A fourth thing is "drawing creatures to a tactically advantageous choke point to limit the number of attackers you face at s time"

A fifth thing, related to the fourth is "drawing creatures back to a single location on the level that you have set up to be especially tactically advantageous" kill holes, oklob farms.

A sixth thing very simular to the fifth and fourth is "leading creatures up stairs to kill them without having to deal with any others"

A seventh thing is leading a creature around in cleared areas, in order to reset a fight by regenning hps or mana for which you either were unlucky or simply overestimated your chances.

An eighth thing is simply trying to disengage from a fight that you estimate yourself insufficiently likely to win.

A ninth thing is attempting to use energy randomization to gain a space to escape from a combat by staircase.

#1 does not appear to be an issue. Bees, yaks, elephants, gnolls are all quite hard to separate. The open terrain of lair potentially plays a part in this. If the monster AI were more stringent about staying together in a pack, that would seem OK to me.

#3 can be reduced by reducing the noise of combat as this reduces the incentive to lure as far back as possible. HP/MP rewarded on kill / increased regen also reduces the incentive to lure in order to limit the chain of foes one must fight consecutively. However, much of the reason why I lure far back is so that my escape route is not blocked off randomly by a new enemy. This would appear to be a very reasonable concern.

#5 can be tweaked by changing enemies (elf trunk changes), terrain (stone walls in some locations), and changing abilities (Fedhas oklob ability adjustment?).

#7 is absolutely necessary in order to win fights which are dice rolls. Every fight in the early game is a dice roll.

#9 is very rare to come across but is important as a way of escaping a fight when there are simply no consumables (early D, usually used in fights vs ogres).

The rest seem OK to me.

Is luring a problem? I don't think it is. I would instead advocate a series of changes to reduce RNG deaths in crawl. Every death should be your fault. Indirectly this would reduce luring as well since a good deal of luring is done to mitigate some of the craziness of the early game.

[Siegurt]

Just to address item 6 on the list: If stairdancing is considered bad, it is easy to fix just by making it impossible to use stairs while a hostile creature is next to you. Some roguelikes have this feature.

Not to single you out particularly here, but this is an example of what I am talking about, I think if we want to have a conversation about what to do about stair dancing, it should go in another topic, however, when some people see 'luring' they include stair dancing in the topic, when it is imho a totally separate issue.

I suspect to have a productive conversation we need to start with a clear definition of which behavior we are trying to talk about. I am going to resist the temptation to talk about stair dancing now as it feels off topic to me

[carbonbasedlifeform]

What about

1. Introducing 'sprint' walk mode for some monsters which reflects swiftness mechanic( 1.2moving speed for x turns, 0.8 moving speed for y turns), where x is significantly higher than y.
2. Enhancing guard mode-mechanic - monsters protect their leader ( highest HD enemy out of a monster LOS) by wandering around him and not leaving his LOS.

Would that ( combining those two points) work in a favor of resolving luring problem?

[Jarlyk]

I think if we want to have a conversation about what to do about stair dancing, it should go in another topic, however, when some people see 'luring' they include stair dancing in the topic, when it is imho a totally separate issue.

They are different tactics, but I think they are closely intertwined. A very common luring destination is on or near stairs, precisely so you can go upstairs if things escalate unfavorably. Likewise, 'anti-luring' proposals need to take stair-dancing into account, since anything making luring more dangerous (whether it's noise or preventing splitting packs or just making monsters faster) can mean stair-dancing becomes an even more dominant tactic. Both tactics are ultimately symptoms emerging from the mechanics of crawl that favor shortish encounters, with retreating to explored territory being the safest way to mitigate risk of escalation. It's difficult to address the issues separately, since they develop from the same underlying causes.

[dynast]

Is luring a problem? I don't think it is. I would instead advocate a series of changes to reduce RNG deaths in crawl. Every death should be your fault. Indirectly this would reduce luring as well since a good deal of luring is done to mitigate some of the craziness of the early game.

Luring as in pillar/stair dancing is the equivalent of refusing to roll the dice, thats why its problematic. The whole RNG unfair deaths can be solved by giving players consumables (as wanderers normally start with) and by removing traps. The kiting, avoiding fights, luring in general should not be something completely in the hands of the player, since this is a dice rolling game, there has to be a random factor when trying to avoid a fight, or just running when you dont have any other options.

[all before]

Siegurt: I don't think #3 (luring to a safe area) can be separated from #6 (stairdancing) because luring to or up stairs is almost always the optimal form of luring to a safe area.

[zxc23]

Luring as in pillar/stair dancing is the equivalent of refusing to roll the dice, thats why its problematic.

These are two separate things. Pillar dancing is about rolling the dice many times in order to reduce variance. Stair-dancing is to do with separating a pack, where fighting the entire pack at once would be suicidal. By the way, one change could be to make enemies that are part of a pack less likely to follow you through stairs so that you can't separate packs that way.

The whole RNG unfair deaths can be solved by giving players consumables (as wanderers normally start with) and by removing traps. The kiting, avoiding fights, luring in general should not be something completely in the hands of the player, since this is a dice rolling game, there has to be a random factor when trying to avoid a fight, or just running when you dont have any other options.

Not having consumables is one of the main reasons why the early game is so random, I agree there. One of my main ideas for reducing RNG deaths is to start the game at about XL 6 and a few floors down, with an assortment of items you often find by that point in time. This would get rid of most pillar dancing, most RNG deaths, and skips most of the more conservative luring that occurs in a game.

I don't see what you mean about crawl being a dice rolling game. Lair onwards, chance plays almost no part at all. The very early dungeon is the odd part out.

[dynast]

Pillar dancing is about rolling the dice many times in order to reduce variance.

I assume you are talking about energy randomization. Running to recover hp/mp is not reducing variance, its reseting it.

Stair-dancing is to do with separating a pack, where fighting the entire pack at once would be suicidal.

Stair-dancing is when you dont wanna deal with whatever is happening on that floor at that current moment, be it a yak pack, a ood spawn, your lack of mp/hp. You dont need stairs to separate packs in the first place. To be fair, stairs are its own can of worms so its hard to talk about it.

One of my main ideas for reducing RNG deaths is to start the game at about XL 6 and a few floors down, with an assortment of items you often find by that point in time. This would get rid of most pillar dancing, most RNG deaths, and skips most of the more conservative luring that occurs in a game.

Then what? you head upstairs and explore a empty floor or a floor filled with trivial enemies and get more loot/exp? If i started the game already "set" i probably wouldnt be playing it anymore.

I don't see what you mean about crawl being a dice rolling game. Lair onwards, chance plays almost no part at all. The very early dungeon is the odd part out.

Thats like you dont aknowledge the very RNG you are talking about, just because they game gets too much in the hands of the player doesnt mean the rolls where never been there to begin with.

[dpeg]

Alright, I've just re-read the whole freaking thread because I have too much time on my hands, and because I am really interested in doing something against luring.

But before that: a number of postings start with "The obvious solution is ...". Hey guys, it's certainly not that simple! There are many different mechanics blending into each other here, which has been explained a couple of times. It's not clear to me whether there is a decent solution at all; if so, there's no reason that is unique and/or simple.

Okay, so what I hate very much is dragging monsters one by one into clear territory. This is my definition of luring. It has nothing to do with packs, and it is exclusively a safety measure to avoid risks. I don't think rewarding low turn counts can work to address this; I am a firm believer in the stick, screw the carrot. I don't mind if the game gets harder (by the way, whatever change we come up with, it will probably affect "naive" newbies the least). I like to play stealthy characters a lot and am highly interested in keeping them fun to play.

Hence the idea should be that luring is punished in some ways. I already made a posting where I listed four options for such punishment (these were collected from this very thread).
Siegurt suggested gradual punishment: whenever you lure (defined by me as: moving while a monster tries to reach you, whether in sight or not; but not if decreasing the distance), then random asleep monsters may be awoken (but not told your position). After reading all of this, I think this is a very gentle concept.
I think it can be expanded: What if the game rolls a monster, or three, and finds nobody asleep (anymore)? Well, then the rolled monster could move to a staircase. Or it could now be told about player position. This would create a local timer, on each level, defined by how awake, and aware of your position, monsters are. If you lure for a long while, then everyone will be awake, and try to get to your place, and await you at a staircase should you leave the level. (By the way, I have no problems with cheating: in such a sitation, I'd be fine with you going down *any* staircase, and everyone awaiting you. Yes, it's not realistic, but the point will be clear. They are awake, they know where you are, and they'll show up wherever you come up/down to get you.)

I imagine that such a system could lead to a state where going *towards* the crowd and fighting them, even though the neighbours will get awake is actually safer than luring. Note how I am totally fine with using a tunnel/chokepoint etc. These are not my concerns.

I intentionally mostly ignored stairdancing. Perhaps there's something to get out of this thread.

[HardboiledGargoyle]

To me "change whatever is making luring monsters optimal" means changing the fact that monsters will stand in a line for you, waiting for their turn to die.

No. Just because we want to give monsters one anti-suicidal tendency (which makes the game more fun) does not mean we want to give monsters other anti-suicidal tendencies (which may make the game less fun, or be hard to code).

My main point is still that luring is more important for a naga than for something speed ten like a human because a human can fight an orc in a hallway and if an orc warrior or two followed by a bunch more orcs shows up just run to the stairs and go down deifferent ones. The naga can't outrun the orc warrior so the naga has to lure the orc back a bunch of spaces. TLDR; speed 11 monsters doesn't solve the problem.

That's a good way of demonstrating that swift monsters don't solve the particular problem of Degenerate Luring.

#3 can be reduced by reducing the noise of combat as this reduces the incentive to lure as far back as possible. HP/MP rewarded on kill / increased regen also reduces the incentive to lure in order to limit the chain of foes one must fight consecutively. However, much of the reason why I lure far back is so that my escape route is not blocked off randomly by a new enemy. This would appear to be a very reasonable concern.

If keeping your escape route not blocked off is your main reason for "luring" far back, then you should hardly mind incidentally dragging entire packs with you to your staircase or whatnot.

I think if we want to have a conversation about what to do about stair dancing, it should go in another topic, however, when some people see 'luring' they include stair dancing in the topic, when it is imho a totally separate issue.

They are different tactics, but I think they are closely intertwined. A very common luring destination is on or near stairs, precisely so you can go upstairs if things escalate unfavorably.

Whether or not you lead them up stairs as well, you draw your hand-picked cluster of monsters into an otherwise safe/cleared area, putting all the other monsters on pause so they don’t get to interrupt. Don’t we have proposals for lvl-9 transloc spells that do mostly that? Effectively it’s like if the game spawned only 1-8 monsters per floor. You have a point in how the 2 are competing tactics. Is there a third?

I imagine that such a system could lead to a state where going *towards* the crowd and fighting them, even though the neighbours will get awake is actually safer than luring.

new negative status: "drowsy", Aizul permanently has it

[HardboiledGargoyle]

I think dpeg’s splitting is tighter and more useful than Siegurt’s and I’d like to address it:

As I see it, there are the following problematic features (problematic as in, useful to do but tedious to carry out):

1. Luring: Spot a monster and make it follow you. Attempts to only ever fight one on one, in safe terrain.
2. Kill-holing: Makes sure to fight and see only a single monster, no matter how large the group.
3. Pilllar dancing: When in one on one combat with a monster and low on MP or HP, run in circles to regain HP/MP. Potentially also get an extra action due to energy randomisation for safe stairs.
4. Stair dancing: Various uses, including (1) leading monsters in small groups (between 1 and 8 depending on layout of destination level) to a safer place, (2) skipping a fight to regain MP (or HP).

It is hard to solve all of these at once. Each one is a "tactic", i.e. you may feel smug when it's the margin to win a fight. But Crawl throws hundreds of monsters at you, so it becomes very stale if you apply them all the time.

Perhaps it helps to think about this in the following way: suppose Crawl was a board game and you're the game master, and your friend is the player. If your friends keeps using the above methods, all the time, well, you can throw a spanner in his works... How would you do it? Can you algorithmise it?

A - luring
dpeg is tempted to do what, in order to have more fun? avoid luring. what does avoiding luring do? what is the consequence of not-luring? answer: more monsters noticing you all at once (assuming you engage at all and don't 1HKO everything). that is the fun consequence dpeg is tempted to experience. and that is the very consequence of continuously shouting monsters, a non-radical (and gentle, too; what dpeg calls gentle 2 posts above sounds brutal and spoilery to me) solution which I recommend. (forgive me, o dpeg, for divining on what you think.) I’d even argue that continuous shouting would be less of a new feature and more of a removal. Look: monsters shout, but quickly stop after seeing you. In other words, monsters have a % chance to shout on the first turn that they notice you, but not on subsequent turns. A new player that notices this behavior may suspect this to be an exploitable bug, like good old "shoutless". That’s right, boys, I just called Degenerate Luring “Shoutless 2: The Revenge”.

B - killholing
Reliable non-formicid non-shatter non-IOoD ways of making killholes should become not so, yeah. There are only a few. Naturally occurring killholes can remain on levels where they’re deemed acceptable.

C - pillar dancing
Where a player would pillar dance, this is worse than an in-combat ability to “press 5 and rest” only because it requires more keystrokes. Energy randomization actually does create the chance of the monster getting a chance to hit you, incentivizing you to use a consumable. The one turn it takes to use a staircase gives the monster an even greater chance of hitting you, so stairs work better than energy randomization.

D(1) - stair dancing, dragging monsters onto safe levels
This can be removed by making all monsters (or almost all? maybe not uniques) incapable of using stairs. There are already several mechanisms that punish/prevent using stairs; monsters following you does not have to be one of them. Importantly: it is bad if stair-using monsters work as intended, i.e. as a threat. Let me explain:

The specific way Crawl lets you skip so much fighting is really nice and makes it stand out from other games. I'd rather Crawl do that well than turn around and try to be like another game. If you put in features that make running away less convenient while keeping it viable - all monsters having old-swiftness, stair-using monsters, Broguelike stairchasing (how did that work out in Crawl Lite anyway?) - you signal to new players that every fight is to the death. That one of those is in the game already is one of the reasons every guide and even the official manual must keep saying, "remember, you don't have to kill every monster you encounter!"

Crawl is "a game where you can run from your problems" unlike many other games. That is not a problem to solve. That is awesome. What Crawl should not be is "a game where your problems split themselves into easily defeated bite-sized pieces, if you do a bunch of easy tedious stuff", but Degenerate Luring (and this here loophole) make it that game. Which leads us to:

D(2) – stair dancing, bailing from a fight to recover
is not ideal, perhaps, but I think we should let it stay, as it’s the lesser of all evils, and not worry about it that much. If you can kill dudes while on upstairs, letting yourself get swarmed, without moving away, then letting monsters follow you up the stairs makes your job even easier. If you can't do that (bit of a false dichotomy, I admit), then you can't camp at those upstairs, so you need to: move off the upstairs, or use different downstairs. Both of those alternatives involve i) more dynamism and player decisions than stairdancing and luring[d.(1) and a.] (even if we were to grant that they're all technically rinse-and-repeat) and ii) less tedium than monsterstairdancing and pillar dancing[d.(1) and c.] and, while we're at it, DD^Mak in CPA. Which kinda answers, BTW, Ceann/svendre's query about why we're singling out luring. Re: svendre’s game modes: Crawl has a ton of these and almost all enable and encourage Degenerate Luring (I guess Qaz is different).

[Tiktacy]

Personally, I don't think there is anything wrong with luring as it is. Generally speaking, the devs want to try and not encourage tedious behavior, but luring popcorn monsters in an effort to marginally increase winrates is so superfluous that I think it deserves a pass(which in my mind is the only thing truly problematic with luring). Especially considering the amount of work you would have to put into fixing this without detracting from other parts of the game. Aside from that, I really don't see anything truly problematic with it. Perhaps it could use some improvement, but I think people are making this out to be far more problematic then it actually is.

Perhaps there is something I am missing though.

[dpeg]

Tiktacy: If you argue that luring is not necessary or even less than optimal, I'll be all ears: it'd mean that its status changes from "tedious no-brainer" to "negative action", and we don't have to punish already bad behaviour. However, I am afraid that luring is basically everywhere a good idea. In my game right now, I should be doing it in Elf, in Snake, in Depths, in Vaults... seriously everywhere. (Or at least, that's what I think I should do.)

HbG: "gentle" as a design, not balance wise. I think you got my drift on luring very well. I believe that even with continued shouting, stealth would be useful (I guess this goes to Ceann).
I believe stair rules could be changed, too, but in my experience it is better to separate issues, and I think it is possible to do this here.

[prozacelf]

I don't lure very often, but my abysmal winrate also reflects that. Once I get beyond a certain part of the game I decide that I'm invested enough in the character that I basically come to the conclusion that I should lure whenever possible.

[svendre]

Tiktacy: If you argue that luring is not necessary or even less than optimal, I'll be all ears: it'd mean that its status changes from "tedious no-brainer" to "negative action", and we don't have to punish already bad behaviour. However, I am afraid that luring is basically everywhere a good idea. In my game right now, I should be doing it in Elf, in Snake, in Depths, in Vaults... seriously everywhere. (Or at least, that's what I think I should do.)

HbG: "gentle" as a design, not balance wise. I think you got my drift on luring very well. I believe that even with continued shouting, stealth would be useful (I guess this goes to Ceann).
I believe stair rules could be changed, too, but in my experience it is better to separate issues, and I think it is possible to do this here.

When luring is less than optimal:

* I've got inner flame cast on a guy that I'm killing and I need the pack to all be right on his heels rather than seperated.
* I've confused one guy, and I want to confuse a second one so they hit each other
* I'm using discord, and I want as many as possible to follow me back - and stay tightly grouped
* I'm using discord, and if I can have all the monsters on the floor in one place coming down the stairs, I can get them all dead at once as I blink/tp away to a location sure to not be near monsters (since they're all tightly packed together in one location)
* I'm worshipping Vehemut and using AOE spells, I really need that fireball to be hitting 4 guys instead of 1 so I get more mana back, so hopefully more guys are coming besides just one, and staying together... tightly.
* I'm worshipping Mahkleb and using AOE spells and I want a bigger heal for my mana spent, so that I can sublimate

etc. etc.

You mentioned you play a lot of stealthy games. I play a lot of the opposite. I found stealth to be somewhat dangerous. On the one hand, if you approach the monsters in the first place, you might get the drop, at least a turn or two before the rest wake up - but the battle is on the terms of RNG. On the other hand, if you're like me and you like to shout and be as loud as possible - when you're where you want to fight, it can be a larger advantage.

You may have monsters all wake themselves up and start hunting and think you're making a disadvantage for the player because of the way you play, but you may also be giving an advantage to other styles of play.

I also let those vault guys pull their summons a lot of the time, intentionally. I also walk onto mark traps intentionally sometimes, if the location is close enough to somewhere where I feel I would have the advantage.

Please don't misunderstand me, I understand your point and what you're saying. It's from your viewpoint however, and I'm playing devil's advocate here.. but as you said already - I think you know it's not "that simple".

~~~
If you really want to piss off people who lure, degenerately... the best way to do it would be to make the monsters not take the bait and hold a defensive position which would be determined by if they were melee or ranged. Melee would hide around some corner to ambush. Ranged would line up in a row next to their other ranged guys, so when you round the corner, you're facing a horrible volley. Smarter AI that sometimes tries to sneak around behind you, where they think you came from... these are things that would screw someone who massively lured as much if not more than everything on the floor going ballisic psycho get-the-chase-on-now-everyone.

[Ceann]

Tiktacy: If you argue that luring is not necessary or even less than optimal, I'll be all ears: it'd mean that its status changes from "tedious no-brainer" to "negative action", and we don't have to punish already bad behaviour. However, I am afraid that luring is basically everywhere a good idea. In my game right now, I should be doing it in Elf, in Snake, in Depths, in Vaults... seriously everywhere. (Or at least, that's what I think I should do.)

HbG: "gentle" as a design, not balance wise. I think you got my drift on luring very well. I believe that even with continued shouting, stealth would be useful (I guess this goes to Ceann).
I believe stair rules could be changed, too, but in my experience it is better to separate issues, and I think it is possible to do this here.

As I mentioned "and have had erroneously called out as bikeshedding" is that there are plenty of actions are "tedious no-brainers" I would be extremely hard pressed to go out and find a game short of gambling that winning does not revolve around the exploitation of advantages. If you eliminate the ability to do so, if me finding an Ogre on D:4 and getting knocked to 5 hp in one attack and wishing to run, my choices should not change to die, spam un-id'd consumables, granted you may not have even found a heal woulds yet, or starting over don't seem very appealing. If I wanted to gamble I would just gamble, not play a roguelike. An IE can get Throw Icicle and clear nearly an entire 3 rune run with that spell alone, or an FE Fireball, or an Mi just got ^Mak.

Dynast wants to refer to variance, well AC,EV,SH contribute way more to loss of variance than luring does or pillar dancing, once you reduce variance enough you win game essentially because your goal is to craft a character that doesn't die to random variance. Does luring save you from getting poisoned to death in spider without rPois? Nope. Does luring save you from getting killed by Sirens in shoals? Nope. Does luring stop a Deep Elf annihilator from hitting you for 90 with LCS? Nope.
I can go on but I hope that you get my point, resistances, gods "makleb healing...really?" consumables, evo's, do way more to remove monster variance in regards to killing the player than luring does. Now I do understand the argument that pretty much anyone can lure, but I think the game has too many things ingrained into what would be required to "fix" this that any solution I can think of would create more problems than it solves and if the actual task at hand here is remove the ability to reduce variance, there are many many other candidates that contribute to loss of variance more substantially than luring does. Teleporting on speed runs is way more degenerative than luring, speed run high scores are essentially a matter of gambling anyway, you just keep doing them until events slide into your favor, you find some broken item early game and it enables you to offset variance to win faster.

I don't really lure myself, rarely if I see an excessively dangerous room I might lure something away, typically moving back to a choke point is the minimum that is required. No one really has any respect for someones 160k turn, 4 rune win, so if people get them who cares? If the assessment of skill comes down to speed runs, multiple zig completions for final score, what does it matter if people tank their score by doing these things?

If you want to have a good score it is probably entirely less than optimal. If luring matters that much then remove scoring entirely, because if luring matters, then scores don't matter.
I know you said that you prefer the stick over the carrot, but how many dogs have you met that you want to be around that are vicious and hostile? Probably none.

If you give players an incentive, something that they can lose by playing slower, for example maybe have monsters destroy loot if they see you and you run away, or have them attack shops if you run away, this forces players to be greedy because they don't want to lose resources, losing resources gets you killed. Then they have to balance their greed against their caution, hence they have an incentive to not runaway and problem of luring resolves itself.

If you have two dogs and only one treat and you always give it to the dog who shows up first, they will compete to show up first. That is incentive.
Having the system as, if one dog shows up and the other doesn't then neither gets it, then they will probably both just stop showing up at all. That is the stick.

Luring is small potatoes, if you want to fix it though, I think the answer lies in using an loss of an incentive, not inflicting a punishment.

[ydeve]

As a reference, I play mostly noisy casters, and I've found myself luring quite a bit. Spells make a huge amount of noise, causing monsters to cone in from all over, and I only have a limited amount of mp. I find it safer to lure the enemy back a little bit (not multiple screens though, too tedious) before blasting them. I always constantly walk back to the stairs to keep my exit path open because autoexplore does weird things.

Does luring save you from getting poisoned to death in spider without rPois? Nope. Does luring save you from getting killed by Sirens in shoals? Nope. Does luring stop a Deep Elf annihilator from hitting you for 90 with LCS? Nope.

Actually, luring saves you from these things very effectively. You don't usually die to a single spider, even without rPois. A siren by itself is harmless. A Deep Elf annihilator is squishy enough that if you can't kill it before it kills you you have no business being in Elf. These monsters, indeed almost all monsters in Crawl, are dangerous only when there are other monsters around. 3 spiders could kill you without rPois. Sirens prevent you from escaping or killing more dangerous monsters. DE annihilators hitting you with LCS is devastating when you're taking damage from other monsters.

I can go on but I hope that you get my point, resistances, gods "makleb healing...really?" consumables, evo's, do way more to remove monster variance in regards to killing the player than luring does.

Resistances, extra AC, gods, consumables, etc are no where near as good as fighting each monster one at a time and completely healing before each fight. If you can't kill a single monster you are way ood, but if you can't handle a group that's a different story.

If you want to have a good score it is probably entirely less than optimal. If luring matters that much then remove scoring entirely, because if luring matters, then scores don't matter.

You've hit the nail on the head here. For a large number of players, I'd guess even the majority of players, scores don't matter at all. I'd argue that most players don't speedrun. They enjoy playing different combos and count "score" mostly by how many runes they got and if they managed to win. So the score disincentive to luring is completely meaningless to the majority of players.

[Ceann]

As a reference, I play mostly noisy casters, and I've found myself luring quite a bit. Spells make a huge amount of noise, causing monsters to cone in from all over, and I only have a limited amount of mp. I find it safer to lure the enemy back a little bit (not multiple screens though, too tedious) before blasting them. I always constantly walk back to the stairs to keep my exit path open because autoexplore does weird things.

Does luring save you from getting poisoned to death in spider without rPois? Nope. Does luring save you from getting killed by Sirens in shoals? Nope. Does luring stop a Deep Elf annihilator from hitting you for 90 with LCS? Nope.

Actually, luring saves you from these things very effectively. You don't usually die to a single spider, even without rPois. A siren by itself is harmless. A Deep Elf annihilator is squishy enough that if you can't kill it before it kills you you have no business being in Elf. These monsters, indeed almost all monsters in Crawl, are dangerous only when there are other monsters around. 3 spiders could kill you without rPois. Sirens prevent you from escaping or killing more dangerous monsters. DE annihilators hitting you with LCS is devastating when you're taking damage from other monsters.

I can go on but I hope that you get my point, resistances, gods "makleb healing...really?" consumables, evo's, do way more to remove monster variance in regards to killing the player than luring does.

Resistances, extra AC, gods, consumables, etc are no where near as good as fighting each monster one at a time and completely healing before each fight. If you can't kill a single monster you are way ood, but if you can't handle a group that's a different story.

If you want to have a good score it is probably entirely less than optimal. If luring matters that much then remove scoring entirely, because if luring matters, then scores don't matter.

You've hit the nail on the head here. For a large number of players, I'd guess even the majority of players, scores don't matter at all. I'd argue that most players don't speedrun. They enjoy playing different combos and count "score" mostly by how many runes they got and if they managed to win. So the score disincentive to luring is completely meaningless to the majority of players.

A siren is not harmless if you don't have enough MR. As a caster it is less of an issue if you can just blast them, but I am sure people have gotten into situations from time to time that Siren's have led to their death. You are also correct correct that fighting one monster at a time is optimal, but really you are just saying kill holes or single lane hallways are optimal, which does not fall under luring specifically. In spider you will never find a single spider and spiders move faster than essentially every one, so attempting to insinuate that people are luring in spider is extremely unlikely. You could try to lure away a DE annihilator, he could then end up blinking behind you and then killing you with a single LCS on his next action, I am not saying its impossible or that happens with an astounding frequency, but it can happen and when it does the act your luring was irrelevant.

As I mentioned, the best way to eliminate the behavior, is the provide an incentive not to do it rather than a punishment for doing it. In crawl your #1 objective is to not die, if you eliminate the most defensive behavior then people will just default to the next most defensive behavior, because that is the mindset of the players who are doing it. If you add incentives to encourage them to play more risky you won't really modify the behavior, you will be forcing a behavior. I really feel like it is a smaller subset of players that lures nigh every single monster.

So an idea...sub branches "volcano,bailey,ice cave etc" I feel are a reasonably good incentive to get players to play in a more risky manner, an option might be to maybe increase misc branches and their spawning and reduce actual loot on dungeon floors. This would put more loot into misc branches and would penalize players for excessive luring because if they take too long to get to the misc branches they would lose out on valuable loot which could then result in them getting killed, while still leaving luring available as a valid tactic for situations that would warrant it.

You then boil it down too, well you could lure a lot but if you do... you are going to lose out on this loot. That still leaves the player with the choice to do it, but adding a risk for them by doing so.
I would also like to add that it would be a dev. opportunity because then you could design new sub-branches and it would be a way to add content.

[HardboiledGargoyle]

Especially considering the amount of work you would have to put into fixing this without detracting from other parts of the game.

Except that we have access to the source code, and code for fixing it, so it should not be much work. And how do you even detract from other parts of the game?

You may have monsters all wake themselves up and start hunting and think you're making a disadvantage for the player because of the way you play, but you may also be giving an advantage to other styles of play.

Please don't predictably derail the thread with "smart AI". First off, how do you know that this is the best way? What is the consequence of the defensive AI you propose, and what does it have to do with luring? And what makes you think we "want to piss off people who lure, degenerately"? Pissing people off is easy. People who lure degenerately are pissed off already, because they can lure degenerately with no drawback.

~~~
If you really want to piss off people who lure, degenerately... the best way to do it would be to

Where not-luring is optimal, we don't mind so much, because not-luring is fun. Where luring is optimal, we do mind, because luring is not fun. The goal is to align the optimal and the fun. Get it? Giving an advantage to fun styles of play won't make good designers cry.

Now I do understand the argument that pretty much anyone can lure, but I think the game has too many things ingrained into what would be required to "fix" this that any solution I can think of would create more problems than it solves

There was a game getting developed once. It was a hard game. In Crawl terms, it had no semblance of a sound system. Monsters only started seeking you if they got hurt or saw you. It was a hard game, as I said, but you could move through the dungeon slowly, a few steps at a time, so that very monsters at a time noticed you. If you were really careful, you could reduce most of a floor to a series of 1v1 fights. A major complaint about the game was that it was primarily about luring away monsters, making for pretty mindless combat. Then devs figured out how to make monsters who can see the player activate all monsters within a small area. And the monster state was pretty binary: either knowing where you are and seeking you, or "asleep". It seems like a huge change, but it just made people play the way the game was intended. It forced difficulty rather than leaving it as an option, if you were so kind and inclined to refrain from cheesing the game. Luring was a way to opt out of the game's difficulty, but was never meant to be a balancing factor. If Degenerate Luring balances Crawl, that's a pretty big issue, because it shouldn't. Good AC/EV/SH reducing variance is a different matter, and not an issue in the same way. It's not like the game is balanced around having 0/0/0 defenses all game long and having defenses is a cop-out.

If you have two dogs and only one treat and you always give it to the dog who shows up first, they will compete to show up first. That is incentive.
Having the system as, if one dog shows up and the other doesn't then neither gets it, then they will probably both just stop showing up at all. That is the stick.

Luring is small potatoes, if you want to fix it though, I think the answer lies in using an loss of an incentive, not inflicting a punishment.

What you call incentive is more specifically competition (Crawl is single-player) and I don't understand the rest - it sounds weird.

[Ceann]

Especially considering the amount of work you would have to put into fixing this without detracting from other parts of the game.

Except that we have access to the source code, and code for fixing it, so it should not be much work. And how do you even detract from other parts of the game?

You may have monsters all wake themselves up and start hunting and think you're making a disadvantage for the player because of the way you play, but you may also be giving an advantage to other styles of play.

Please don't predictably derail the thread with "smart AI". First off, how do you know that this is the best way? What is the consequence of the defensive AI you propose, and what does it have to do with luring? And what makes you think we "want to piss off people who lure, degenerately"? Pissing people off is easy. People who lure degenerately are pissed off already, because they can lure degenerately with no drawback.

~~~
If you really want to piss off people who lure, degenerately... the best way to do it would be to

Where not-luring is optimal, we don't mind so much, because not-luring is fun. Where luring is optimal, we do mind, because luring is not fun. The goal is to align the optimal and the fun. Get it? Giving an advantage to fun styles of play won't make good designers cry.

Now I do understand the argument that pretty much anyone can lure, but I think the game has too many things ingrained into what would be required to "fix" this that any solution I can think of would create more problems than it solves

There was a game getting developed once. It was a hard game. In Crawl terms, it had no semblance of a sound system. Monsters only started seeking you if they got hurt or saw you. It was a hard game, as I said, but you could move through the dungeon slowly, a few steps at a time, so that very monsters at a time noticed you. If you were really careful, you could reduce most of a floor to a series of 1v1 fights. A major complaint about the game was that it was primarily about luring away monsters, making for pretty mindless combat. Then devs figured out how to make monsters who can see the player activate all monsters within a small area. And the monster state was pretty binary: either knowing where you are and seeking you, or "asleep". It seems like a huge change, but it just made people play the way the game was intended. It forced difficulty rather than leaving it as an option, if you were so kind and inclined to refrain from cheesing the game. Luring was a way to opt out of the game's difficulty, but was never meant to be a balancing factor. If Degenerate Luring balances Crawl, that's a pretty big issue, because it shouldn't. Good AC/EV/SH reducing variance is a different matter, and not an issue in the same way. It's not like the game is balanced around having 0/0/0 defenses all game long and having defenses is a cop-out.

If you have two dogs and only one treat and you always give it to the dog who shows up first, they will compete to show up first. That is incentive.
Having the system as, if one dog shows up and the other doesn't then neither gets it, then they will probably both just stop showing up at all. That is the stick.

Luring is small potatoes, if you want to fix it though, I think the answer lies in using an loss of an incentive, not inflicting a punishment.

What you call incentive is more specifically competition (Crawl is single-player) and I don't understand the rest - it sounds weird.

Sorry if its complicated...its a matter of psychology.
If the player is doing something you don't want them to do, give them a reason not to do it. You can make the reason either negative or positive. Positive reasons are usually a better way to adjust behavior than negative ones. Making bad things happen to the player for moving away from monsters ""because you really have no way to differentiate between luring, funneling to a choke, ranged kiting behavior etc"" is disruptive to valid tactics. All they will do "if they are already playing defensively in the first place" is gravitate towards whatever the next most defensive behavior becomes as the end result.

Loot is what ultimately decreases variance and allows the player to live, so as I suggested if you remove some loot from the dungeon and place more misc branches which have timed portals, you will force players to lure much less or only when absolutely necessary, because they now have an incentive to go faster, if they go slow they will lose out on loot and die from lack of it. So if they want to play by luring excessively it becomes to their own detriment. This allows the player to retain the option of luring but balancing it by placing a risk onto them for going slowly. Which I think is more in tune with where crawl seems to be going these days... you can do this "thing" but there is a cost to doing it.

[savageorange]

Sorry if its complicated...its a matter of psychology.
If the player is doing something you don't want them to do, give them a reason not to do it. You can make the reason either negative or positive. Positive reasons are usually a better way to adjust behavior than negative ones. Making bad things happen to the player for moving away from monsters ""because you really have no way to differentiate between luring, funneling to a choke, ranged kiting behavior etc"" is disruptive to valid tactics. All they will do "if they are already playing defensively in the first place" is gravitate towards whatever the next most defensive behavior becomes as the end result.

Loot is what ultimately decreases variance and allows the player to live, so as I suggested if you remove some loot from the dungeon and place more misc branches which have timed portals, you will force players to lure much less or only when absolutely necessary, because they now have an incentive to go faster, if they go slow they will lose out on loot and die from lack of it. So if they want to play by luring excessively it becomes to their own detriment. This allows the player to retain the option of luring but balancing it by placing a risk onto them for going slowly. Which I think is more in tune with where crawl seems to be going these days... you can do this "thing" but there is a cost to doing it.

In other words, you want to impose an opportunity cost on the player so they think "because *I* am not progressing, I don't have X", instead of reactively responding "you are doing this bad behaviour [luring], and so you will be punished with [increased monster spawning,etc]". Makes sense to me, and promotes sense of agency rather than retarding it.
I think it's definitely worth a try, myself. If we can reduce equipment item generation without reducing consumable item generation, that seems like the right place to start.

[HardboiledGargoyle]

In spider you will never find a single spider and spiders move faster than essentially every one, so attempting to insinuate that people are luring in spider is extremely unlikely. You could try to lure away a DE annihilator, he could then end up blinking behind you and then killing you with a single LCS on his next action,

Mmm? Spiders being faster doesn't invalidate luring, it just decreases the distance by which you can lure them, and it still splits packs. And, OK, suppose a lone Annihilator can kill you, despite perfect luring. Then, two Annihilators can kill you much more easily. So you'd still rather lure.

In crawl your #1 objective is to not die, if you eliminate the most defensive behavior then people will just default to the next most defensive behavior, because that is the mindset of the players who are doing it.

Umm, sure. Is that a problem? We don't hate luring because it's defensive. We hate it because it's cheesing and boring. At least much more so than the next most defensive behavior, whatever that is (will probably still involve retreating; you mentioned chokepoints on page 1 of this thread).

If the player is doing something you don't want them to do, give them a reason not to do it.

How about instead, just get rid of the reason they want to do that something? They lure because luring splits packs. Just make luring not split packs. Seems much simpler IMHO. You really take the "valid tactics" of luring for granted.

reward speedrunning
punish slowrunning

In other words, you want to impose an opportunity cost on the player so they think "because *I* am not progressing, I don't have X", instead of reactively responding "you are doing this bad behaviour [luring], and so you will be punished with [increased monster spawning,etc]". Makes sense to me, and promotes sense of agency rather than retarding it.
I think it's definitely worth a try, myself. If we can reduce equipment item generation without reducing consumable item generation, that seems like the right place to start.

Doom clock, timed portals, incentives to go deeper/faster are good in the kind of unwinnable game, where you just dive for score or for greatest depth reached. It's okay for them to exist in Crawl, but not as the default. Crawl is a finite game, and quite random in what it can give you. It must be fun with characters of varying strength. I feel that letting people take a long time to clear levels, as well as letting them disengage and rest more frequently, is a good buffer, and without it, weaker characters will feel painfully difficult and stronger characters will feel grossly overpowered.

[Hurkyl]

No one really has any respect for someones 160k turn, 4 rune win, so if people get them who cares?

I have to disagree.

Okay, sure, all other things being equal it's unimpressive, but when it's the by-product of a very high win rate or a long winning streak, it is impressive.

Even just in terms of self-motivation, I would find "I can win nearly every game of crawl" more satisfying than "I can get a high score at crawl".

[svendre]

Please don't predictably derail the thread with "smart AI". First off, how do you know that this is the best way? What is the consequence of the defensive AI you propose, and what does it have to do with luring? And what makes you think we "want to piss off people who lure, degenerately"? Pissing people off is easy. People who lure degenerately are pissed off already, because they can lure degenerately with no drawback.

How am I derailing the thread talking about monster behavior with regards to monster movement and tracking of the player when the topic is about luring monsters? What does monster AI have to do with luring.. I don't even know where to begin answering that. How do I know what is the best way to discourage it? How do you know what isn't the best way? It's an opinion, I assumed that was understood - I'll be more clear next time. What are the consequences of sometimes more defensive AIs? It's hard to predict everything but I have at least seen two changes done in the past.

1) Melee monsters sometimes not chasing, and hiding around corners to ambush.. as described *I felt* it was effective in some games I've played.
2) I've seen ranged get smarter about maintaining range the way they come around a corner with a little distance instead of melee range, and *I feel* it has benefited them and made the game slightly more challenging and interesting.

Both changes were done in a non-cheat, cheesy way.. (they didn't just get a random speed boost) which is *my opinion* based on playing the game. I don't think monsters behaving the same all the time is ideal (i.e. always ballistic chase, or always hiding in defensive positions.) I believe that mixing it up a bit would yield the most interesting results, a randomly rolled disposition for a monster or group.
What makes me think you want to piss people off? Nothing... I said: "If you really want to", which is a conditional statement.

[HardboiledGargoyle]

How am I derailing the thread talking about monster behavior with regards to monster movement and tracking of the player when the topic is about luring monsters? What does monster AI have to do with luring.. I don't even know where to begin answering that. How do I know what is the best way to discourage it? How do you know what isn't the best way? It's an opinion, I assumed that was understood

This is a thread about luring. I'm asking about the consequences of what you propose, because you don't state what it will do, other than piss people off, which is not a respectable design goal. It's okay for it to be an opinion. You're the one saying "it's the best way". What reason do you have to think that smart AI is the best way to solve luring unpissed players something?

And I asked why you associate luring with defensive AI. More aggressive AI can counter luring too.

What makes me think you want to piss people off? Nothing... I said: "If you really want to", which is a conditional statement.

Why would you say "Ah, you want to achieve X, so you naturally want to use method Y, but if you *earnestly* want to achieve X, you should really do Z" if there is nothing to make you think that people want X?

The smart AI argument derailing tends to take one of two forms:

"You say luring is a problem but its root cause is that the AI is dumb and suboptimal. Patching up one of the symptoms barely does anything. To really make a difference, you need to implement smart AI. Otherwise, might as well let the current AI stay as dumb and exploitable as it is."

or

"What these anti-luring measures do is enhance monster AI, but smart/optimal monsters are bad for the game and unfun. We had monsters that flee and centaurs that kite, and we got rid of those for a good reason. [insert hypothetical nightmare scenarios of annihilators spamming LCS and orb guardians surrounding the player on turn 1.] Luring and stupid monsters are there by design. Do not mess with it."

Both positions are fallacious, but tend to be phrased in an indirect way that makes addressing them challenging.

[Siegurt]

"What these anti-luring measures do is enhance monster AI, but smart/optimal monsters are bad for the game and unfun. We had monsters that flee and centaurs that kite, and we got rid of those for a good reason. [insert hypothetical nightmare scenarios of annihilators spamming LCS and orb guardians surrounding the player on turn 1.] Luring and stupid monsters are there by design. Do not mess with it."

Actually this isn't a single argument, you're collapsing two separate things, 1. Those two AI mechanics were implemented and they were awful and 2. The hypthetical is the conclusion one reaches if you try to "make the monsters as smart as possible" which is a completely different thing then "Let's try to make the monsters have this specific behavior" it's not a response to any specific proposal

Just because the two specific *listed* behaviors were awful and tedious, that doesn't mean all proposed behaviors would be, and people who say "Make the AI smarter" aren't actually saying anything at all, it's just useless white noise like someone posting "Just make the game better" the example responses are directed at people who just say that we should "Make the monsters as smart as possible" which is a useless thing, because yes, if you were a collection of smart monsters who wanted to keep the player from reaching the orb, you'd just blow up the stairs so they could never reach it, or surround the entrance with a group of critters who could kill anyone walking down the stairs with no possibility of reprisal. Making the monsters "as smart as possible" doesn't make for a thing you can call a game at all, maybe a really boring electronic novel or something.

If there's a specific proposal for smarter (or dumber for that matter) AI rules that would make the game more fun, then by all means propose them, they'll have to be picked apart by people looking for how to best pervert them to their advantage (as all game mechanics eventually are) but if they pass baptism by forum-fire, and are reasonable to implement, then it's certainly worth considering them, if they happen to address luring in some fashion, then this might even be a reasonable topic in which to discuss them. However the monster AI in crawl has a *very* complex impact, to the extent where any AI rule change significantly effects a very large portion of the game, so don't be surprised when someone finds a hole in such a proposal such that it will actually make the game worse overall, that doesn't mean you should quit trying (Well, maybe if you're really bad at it) but rather that it's actually *hard* to come up with good AI rules that work and make the game better than it is.

[svendre]

How am I derailing the thread talking about monster behavior with regards to monster movement and tracking of the player when the topic is about luring monsters? What does monster AI have to do with luring.. I don't even know where to begin answering that. How do I know what is the best way to discourage it? How do you know what isn't the best way? It's an opinion, I assumed that was understood

This is a thread about luring. I'm asking about the consequences of what you propose, because you don't state what it will do, other than piss people off, which is not a respectable design goal. It's okay for it to be an opinion. You're the one saying "it's the best way". What reason do you have to think that smart AI is the best way to solve luring unpissed players something?

And I asked why you associate luring with defensive AI. More aggressive AI can counter luring too.

What makes me think you want to piss people off? Nothing... I said: "If you really want to", which is a conditional statement.

Why would you say "Ah, you want to achieve X, so you naturally want to use method Y, but if you *earnestly* want to achieve X, you should really do Z" if there is nothing to make you think that people want X?

The smart AI argument derailing tends to take one of two forms:

"You say luring is a problem but its root cause is that the AI is dumb and suboptimal. Patching up one of the symptoms barely does anything. To really make a difference, you need to implement smart AI. Otherwise, might as well let the current AI stay as dumb and exploitable as it is."

or

"What these anti-luring measures do is enhance monster AI, but smart/optimal monsters are bad for the game and unfun. We had monsters that flee and centaurs that kite, and we got rid of those for a good reason. [insert hypothetical nightmare scenarios of annihilators spamming LCS and orb guardians surrounding the player on turn 1.] Luring and stupid monsters are there by design. Do not mess with it."

Both positions are fallacious, but tend to be phrased in an indirect way that makes addressing them challenging.

If you look back on the things I originally said, I was not a proponent of adding tons of anti-luring mechanisms. If certain things were added to counter luring which were extreme, how do you know they wouldn't piss me off? So there you have one reason why I could use that phrase accurately, if only for one person. I realize more aggressive AI can also counter luring.. that's mostly what was being talked about as a counter. I brought up the point of defensive AI *also* having a potential impact, and went so far as to say that I don't think one or the other should be the only solution. Exactly how, the details of mechanics what would be changed, I didn't go into that because as I said, I wasn't actually advocating big changes to stop luring in the first place.

I agree with:

* AI is certainly connected with the topic
* Changing AI would be very difficult to implement, an actual proposal (far from my earlier comments) would be very challenging

If I were to make a serious effort to think about how monster AI could be improved as a net gain and have a positive impact against luring (if that were even my stance to begin with) it would require a lot more thought. I like to be challenged by "smarter" monsters as long as it is done in a non tedious or annoying way (I didn't like the fleeing monsters.)

My quick take on the situation, going back to one of my earlier examples of the monsters which try to encircle the player so they become trapped in their repeated retreat attempts leads me to think about how many games solve the situation of too careful players: using stealthy stabber type monsters. Please don't tear this up as some polished idea (at least I'm talking about ideas instead of only throwing stones at those who do), but perhaps you could have a monster something similar to those ghosts that stab from invisible status, that have an ability to stealth and stab a player. Their effectiveness would need to be tied to the amount of time spent travelling back and forth across the same route and excessive resting in the same spot. A player constantly changing their location and making noise might throw off their otherwise slow approach, perhaps they dislike too much noise and move away from it. A player going from the stairs to a pile, quietly luring one back to the stairs rinse, repeat 5 times would be an ideal target for this monster, giving it time to understand the route the player was taking and time for it's slow moving (in stealth form) to a location it thinks the player will stop and rest at (again) and waits for the return and rest of the player to that specific location, then can deal a nasty surprise blow.

[dpeg]

Maybe this thread is beyond its sell-by date, but I've got to mention this: Any non-trivial change (and many trivial ones too) is going to piss off some players. That's never an argument against a change. (Some changes pissed off a pretty vocal set of potentially many players, and it didn't really matter.)

Several developers have explained, in this thread, that they hate luring, and why. Many players have explained this, but the responsibility of coming up with a decision (including: no change) lies with developers. If we (developers and interested parties) can come up with a solution that improves the matter for us, then that's alright: while it will never be alright for everyone, we expect (and know) that it will be for sufficiently many players as well.

Also note that nobody has even suggested to curb luring wholesale. The proposals (including the "AI changes") are about making it less universally good, not impossible

[HardboiledGargoyle]

I agree with:

* AI is certainly connected with the topic
* Changing AI would be very difficult to implement, an actual proposal (far from my earlier comments) would be very challenging

If I were to make a serious effort to think about how monster AI could be improved as a net gain and have a positive impact against luring (if that were even my stance to begin with) it would require a lot more thought. I like to be challenged by "smarter" monsters as long as it is done in a non tedious or annoying way (I didn't like the fleeing monsters.)

Current monsters get lured away because they lack even a very primitive way of communicating to or controlling their friends. Your train of thought can be divided two ways:

1) Monsters continue to play lone wolves, but have special AI that, in effect, prevents you from luring individual monsters from a group. Well, my first impression of this is: dang, that sounds extremely hard to make work on a technical level, let alone work well.

2) Monsters can coordinate with each other to retreat/encircle/etc as a group. But… you’re giving them the ability to communicate with each other. Which would be enough to fix degenerate luring with the current simple/stupid/aggressive AI. You may like the challenge of playing against smarter monsters, but that goes beyond fixing luring.

Also note that nobody has even suggested to curb luring wholesale.

Actually, I kind-of did, by taking away from luring the thing that makes it degenerate, rather than giving degenerate luring a drawback to make it less universally good. I see it as a loophole, an unintended emergent property - simply an unfortunate consequence of what happens when you set up noise and AI the way Crawl does.

(others suggested curbing luring wholesale too, in a different way…)

Maybe this thread is beyond its sell-by date

Why? Did something within the thread happen to render all following discussion futile and hopeless? It would be helpful if you explained what breaks or forfeits a thread or argument, or, if you’ve done it many times before, link to one of your favorite explanations.

[edgefigaro]

Creatures that are tied to a pack are faster based on how far away they are from the leader. Something like speed=(speed + 0.33/square.) So if there is a lead yak and you pull him through a choke and all the yaks file in and elongate, the ones in the back get a speed bonus, larger if they are further behind, but the pack will never advance past the lead animal.

[partial]

"luring" seems like a really difficult mechanic to identify/define and control (I'm no designer so maybe this isn't true at all)

edit: missed a page in the discussion and want to reform my ideas

[HardboiledGargoyle]

I had a parallel conversation with Siegurt by PM – fortunately, since it allowed a focus that would have been lost had other people contributed. If anything, this delineates two perspectives on the matter. In retrospect, Siegurt has been saying it all along: the gap in power between luring and not-luring, and the number of times the game has you make a “do I lure or not” decision, do not matter, because luring is tedious by definition. I argued against it, and I don't have anything to add at this point. Siegurt's messages are in blue.

Similarly there is also "breaking apart arbitrary groups of critters who happened to be closer to each other" this is actually fairly similar to pack breaking, but it is slightly easier since the critters aren't actively trying to stay together. They are just all chasing you.

a third thing that happens is "drawing critters into cleared areas so as to minimize noise and risk of attracting things that aren't in the immediate area"

I can't tell the difference between thing 2 and thing 3
it's like you're paraphrasing the same thing
or rather, thing 3 is what follows thing 2
since, if you've done thing 2, what else are you gonna do, other than thing 3?
this may be the source of our misunderstanding

(you can quote this in thread if you feel necessary)

Thing 2 is "breaking apart things that are close together so you only have v to fight them one c at a time"

Thing 3 is "drawing any creature away from undiscovered areas into already cleared areas *regardless of the presence of other creatures* because the combat might potentially attract unwanted attention from things you don't know about and have never seen"

The actual actions are similar, but the situation and reasons for the actions are different. In thoery, it is optimal to lure every creature in the game into cleared areas to fight it whether or not there is anything else nearby, just on the off chance the noise attracts something else, or something wanders by while you are injured.

Again, I fail to see what differentiates the two scenarios. What is the variable that changes across the two scenarios?

Is it the distance – far enough, or not far enough, for anything else to hear the combat? Suppose I ensure I get noticed by an orc warrior and lure it out of a room full of enemies into an adjacent hallway. I’ve taken it out of LOS of unaware monsters. I’ve already done 80% of the work needed to take it out their earshot. I might as well lure it far enough for nothing to hear.

Is it the player’s awareness of monsters? Suppose I open a door straight-on and see the whole room, and catch the attention of 1 orc warrior among other monsters, and lure it – that’s non-degenerate. But suppose I open the door from the side, diagonally, and see only half the room, and only 1 orc warrior, and catch its attention, and lure it, just in case there are other monsters – that’s degenerate. Correct?

These distinctions seem really arbitrary to me, I mean coin-flip level of arbitrary.

Actually, this got me thinking. Maybe many people are making this distinction? They want to keep one scenario, but not the other? This could be why we see very odd mechanics getting brought up, like “the chasing monster starts doing X voodoo stuff after getting lured for Y number of turns”. It’s a tension – a desire to keep some lurings and eliminate others – that I don’t understand.

If you dedicate to fighting (AKA making noise) within earshot or until more threats appear, you enforce that which would happen without your consent if monsters continuously alerted other monsters, and you might as well give up the pretention that you’re luring, because all you’re doing is fleeing or repositioning.

I can guess “c” is “creature”. What is “v”?

(I suggest we continue this conversation privately, and then I’ll put it all on Tavern, if you don’t mind.)

It has nothing to do with distance, it's all about whether it's optimal for *every creature in the game* or not.

If I stumble across a lone quokka, with nothing else in LOS, it's presently *still optimal* to not merely draw it to your current position (or a nearby one) but back into an area far away from any possible source of danger.

It's all about motivation, with the orc warrior situation you outlined, it's about mitigating a known threat through positioning (in particular with orc warriors, they're almost never alone), with the quokka situation, it's about mitigating possible unknown threats, which are unlikely to even exist.

It's hard for me to extract the main idea from what you're saying. I try to lay everything clean and bare. You're saying "motivation" and "every" (as opposed to "all but one" AKA "not every"??) like everything's supposed to click into place for me. Eh, maybe it's just me. So:

good: mitigating a known threat
bad: mitigating an unknown (and unlikely?) threat
pfff, I think I'm too tired to seriously ponder that at the moment.

So the holy grail solution would be to make it so:
optimal: not-luring quokka and luring orc warriors
sub-optimal: luring quokkas and not-luring orc warriors
Huh, I guess I'm starting to see where those odd proposals, doom clock etc are trying to wedge themselves.

I'm merely pointing out that they are two different things, I'm not trying to posit a holy grail of solution, only point out that two different reasons for luring require two different discussions about whether it's: tedious, optimal, and degenerate, and possibly two different solutions (If, indeed, both types of luring are considered bad for the game by that metric)

Even if the actions are the same, the reasons for taking those actions are different.

I am not *certain* it's bad for the game that it's good tactics to use luring to break apart groups of critters (it might be, but I'm really not convinced that it's so.) I *am* sure that it's bad for the game that you can marginally reduce your risk at no cost by luring every single critter in the game without regards to whether or not that creature specifically poses more risk in one position vs another.

I actually think that luring being a powerful positive tactic is a good thing, but presently it has no drawback or counter balance (I suspect originally the intention was for the food clock and OOD timers to provide such a drawback, but they fail at that task) and without such it becomes the optimal action for all encounters everywhere in the game, with a drawback that discouraged too-frequent use, you'd only want to use it where it had a definitive and measurable impact, as opposed to when it the benefit was marginal.

Making "lurable" the default, and attaching a cost to the luring of popcorn only, sounds like a difficult goal. And it's difficult to communicate. But that's how many people seem to be thinking.

You just striked upon the reverse of that. That is, so that the default is "not lurable", and the drawback to luring is related to frequency of use. In other words, luring is possible, but has an in-game cost attached every time you elect to use it. What comes immediately to mind: 1)XP-gated timers like on evokers, 2)long-term cost via consumable depletion or something like draining, 3)reward for completing a "quest", like killing a certain monster or getting to a certain place on time. The thing that enables luring can be increased sound attenuation, or increased ambient noise, or something that restricts monsters' ability to shout.

Why aren't proposals based around this, then? "Rampant luring is bad but some forms of luring should be granted" is a position that people can legitimately take. Imagine if Crawl's evocables, for example, didn't have XP-gated or long-term penalties, and were boring due to lack of drawback. And to prevent them from being spammable, people proposed that the solution is to make overusing the evoker "wake up monsters all over the level" or "give swiftness to monsters" or "make you incapable of moving away from a monster". That sounds ridiculous if you're viewing luring as just another tool along with potions and evocables, but that's the spirit of much the discussion in the thread.

But the player should be the one deciding whether an encounter is worth luring, and tanking that cost. You're gonna have problems if you think the game is going to do it well by itself.

In any case, this is much more easily done if Degenerate Luring is impossible by default, and even then, it might turn out to be unnecessary.

Well, i and dpeg both made suggestions along those lines, mine wasn't a very *interesting* suggestion, nevertheless my suggestion is generally speaking "attach an accumulative penalty for using luring, which is small enough that you could still use it in important situations, thereby attaching a real decision about overusing luring" I never suggested that the game should decide if the encounter was worth using luring or not.

As a *bad* suggestion, lets say we attached a significant food cost to moving with creatures in LOS, this would effecitvely not be a penalty if you only lured occasionally, or only used non-degenerate luring tactics, and it would become a progressively larger penalty the more you used it.

The reason that happens to be bad is that it's both hard do to communicate, and food is a terrible mechanic to attach more things to, it does however address the problem by attaching a cost to overusing luring while still making 'normal' use acceptable.

Some other suggestions in the thread along those lines are removing loot if you take too long (Just generally attaching a cost to taking too long to go through the level) and mine of increasing the OOD timer when moving when creatures are in LOS (thereby increasing risk) I don't consider that a very good solution, because I don't think the OOD timer mechanic works overly well, Also Lasty's suggestion for a 'doom clock' which increases risk generally on a level the longer you've been there.

Suppose I find simulacra and decide to go get some stashed rC+ items, and retreat with them in tow. Or, I'm fighting normal-speed monsters in a place where I'm not afraid of anything hearing our combat, but not near stairs, and my HP gets uncomfortably low, so I track back to stairs and the monsters follow. Or, autotravel stops upon a quokka and I just decide to manually move to where I want instead of tabbing the quokka. Or, I'm trying to round up a bunch of monsters to get the most use out of a long-lasting buff or a mass bomb (immo/torment/slouch/etc). There is no luring there, but if the game qualifies that as luring, the player accumulates a penalty. That's not fair. If you want the game to discriminate between genuine luring and not-really-luring (and where does pillar dancing belong?), that's a super-hard task, hard to program, would likely be broken and breakable anyway, like Nemelex's Genuine Card Use conduct back in the day.

I don't think it's just a food cost that would be difficult to communicate. So would waking distant monsters, destroying loot, increased out-of-depth spawns, etc. It would be a weird, strange, obnoxious mechanic, which does not need to be there, and exists to nerf some obscure player behavior, but interferes with normal gameplay.

That's why I'm not saying the solution is to provide incentives for non-luring, or disincentives for luring. If you want to curtail the behavior of luring, just... make it not a behavior. Walking away from a monster has a side-effect of luring it away from other monsters, due to how noise and AI work. Just get rid of the side-effect. Then you could still retreat, but retreat won't be hopelessly intertwined with luring.

Actually I still don't even get why unlimited luring of dangerous enemies like single orc warriors, which is usually equivalent to insta-gibbing them, is a great feature to have in the game.

I can't see why people are attached to their precious luring. Could be Mountain Dwarf Syndrome.

It sounds like your proposed solution is 1. If a monster has seen you ever, it will track you indefinitely, and 2. Combat noise does wake up or attract the attention of other critters.

1. Is problematic, it moves the pack breaking aspect of letting back to staircases, making it more tedious, not less. If your proposal includes making creatures follow up stairs that is problematic from both a coding and logistical problem, coding wise because once you move of the level all critters aren't in memory any longer to track, and logistic wise because one end of the stairs may have more room than the other, and there isn't any where to put all the following critters, if the solution is don't bring critters upstairs if they don't fit, the you re introduce pack breaking, but move it back to an appropriate stair set, which makes it significantly more tedious than it is now, but no less powerful.

2. Wouldn't be significantly problematic from a coding or logistical standpoint, but again, it means if you can see one b pay off a group of critters, but not the rest, you can pick them of one at a time by luring, and it makes the process safer and more predictable, making luring more powerful, while less tedious, not what you were going for.

You might mean something else, I am interpreting vs going of what you said directly. But eliminating the behavior is actually worse in any implementation I have seen proposed than just leaving it alone.

Your response is full of non-sequiturs. Why do you bother interpreting? I do in fact say it directly. And I proposed neither 1 nor 2. But I even disagree with you about why 1 and 2 are problematic. Actually I can't understand half the things you say, so I suspect you're drunk or something. "b pay off a group of critters, but not the rest" - ??

It sounds like your proposed solution is 1. If a monster has seen you ever, it will track you indefinitely, and 2. Combat noise does wake up or attract the attention of other critters.

1 is already part of Ash wrath, and is not problematic logistically or coding-wise. Letting monsters forget you is not a crucial element of luring, so 1 does not impact luring in any obvious way.

2 would actually be like current Crawl, but less spoilery. If combat noise were zero, then you could lure an orc warrior around a corner and safely kill it. But a newbie sees it summon friends from just out of LOS. Presumably, the orc warrior is shouting for help. But actually, no - that's just the noise of you hitting it, attracting attention. The newbie is confusing combat noise for group AI. That is what I meant by calling "group-clustering AI" an illusion in my first post on the thread.

If monsters shouted continuously, combat noise would be almost redundant most of the time. And you can't reasonably claim it's problematic from a coding or logistical standpoint. Like, even berserk monsters already make noise continuously when they see you. You just copy-paste those lines of code for all hostile monsters that see and seek you.

Aww hell, I just realized you're letting auto-suggest give me crap. *sigh*

Your response is full of non-sequiturs. Why do you bother interpreting? I do in fact say it directly.

I have re-read your initial post and don't see a specific proposal at all, you say

"Walking away from a monster has a side-effect of luring it away from other monsters, due to how noise and AI work. Just get rid of the side-effect. Then you could still retreat, but retreat won't be hopelessly intertwined with luring."

But you don't specify what *should* happen instead.

There's a few problems with this:
1. What does "separated" mean, If I encounter an orc warrior, and haven't seen some nearby orcs who are part of his pack, and start walking away, does he follow me, or his pack that I haven't seen yet, how about the hobgoblin who is also near the pack but not part of it, what about the ogre who happens to be near-ish to both myself and the orc warrior, but not in LOS of either one of us? You need to define what you mean by "other monsters" so that we can define what you mean by "separated from"
2. Even if we presume that we have a definition of "other creatures" that shouldn't be separated from the one (or group of) creatures you can see, and we have no "other creatures" per that definition, it's presently *still optimal to lure that singular creature* because while you're fighting one creature another creature could wander nearby and *become* part of that definition, so even if we eliminate the 'breaking creatures apart from each-other' aspect, *it doesn't fix the problem at all, because luring is still tedious and optimal* You've mis-identified the problem, which is why your solution doesn't work.

And I proposed neither 1 nor 2. But I even disagree with you about why 1 and 2 are problematic. Actually I can't understand half the things you say, so I suspect you're drunk or something. "b pay off a group of critters, but not the rest" - ??

As it happened I wasn't drunk, I just typed that response on my cell phone, and my auto correct sometimes does things without me noticing when I'm going quickly.

Here's the paragraph corrected for readability:

2. Wouldn't be significantly problematic from a coding or logistical standpoint, but again, it means if you can see only one out of a group of critters, but not the rest, you can pick them of one at a time by luring, and it makes the process safer and more predictable, making luring more powerful, while less tedious, not what you were going for.

It sounds like your proposed solution is 1. If a monster has seen you ever, it will track you indefinitely, and 2. Combat noise does wake up or attract the attention of other critters.

1 is already part of Ash wrath, and is not problematic logistically or coding-wise. Letting monsters forget you is not a crucial element of luring, so 1 does not impact luring in any obvious way.

2 would actually be like current Crawl, but less spoilery. If combat noise were zero, then you could lure an orc warrior around a corner and safely kill it. But a newbie sees it summon friends from just out of LOS. Presumably, the orc warrior is shouting for help. But actually, no - that's just the noise of you hitting it, attracting attention. The newbie is confusing combat noise for group AI. That is what I meant by calling "group-clustering AI" an illusion in my first post on the thread.

If monsters shouted continuously, combat noise would be almost redundant most of the time. And you can't reasonably claim it's problematic from a coding or logistical standpoint. Like, even berserk monsters already make noise continuously when they see you. You just copy-paste those lines of code for all hostile monsters that see and seek you.

Aww hell, I just realized you're letting auto-suggest give me crap. *sigh*

Hehe, yes, sorry about that I shouldn't type long replies on my cell, it gets me into trouble, and I miss things it doesn't do properly.

If monsters shouted all the time, it'd be even more optimal than it is now to draw every fight back into cleared areas, because shouting is generally louder than combat, also when you initially encounter a critter, it shouts (typically) and monsters are attracted *to that spot* which is yet another reason to fight the thing you want to fight elsewhere.

Monsters not losing track of you isn't problematic on it's own, it's just not a solution for luring (Which I think now reading your current message that we both agree on) If anything it makes luring more predictable. When I said "track you indefinitely" when I thought you were proposing that as a solution for luring (which it appears you are not) I also thought you meant "across levels" which *would* be a problem coding wise (not impossible, but really really annoying)

Again, I feel like the point has been missed, the problem with luring isn't *specifically* that it breaks apart packs or groups of monsters, it's that it's both tedious and optimal, and having it not break apart groups of monsters doesn't make it not-optimal, only slightly less powerful than it is now.

Anything that preserves monsters being attracted to whatever you're fighting now, and doesn't attach any cost to taking the turns needed to lure things away from possible trouble, doesn't fix the 'optimal' part of the problem, and there's nothing to be done to make luring any less tedious (Well, maybe allow auto-travel while monsters are in LOS, but I don't think a UI tweak is really the way to fix the problems here.)

So that reduces us to either: 1. We make it so there's no cost to fighting monsters you discover in place (it's equivalent to moving the fight elsewhere) which means pretty much the whole noise system needs to be thrown out in it's entiretly. Or 2. We attach some kind of real cost to relocating combats, whether that's a general low-level cost for spending extra turns, or a specific cost for certain circumstances certainly does bear discussion. Also worth discussing whether to make it an opportunity cost (give a reward for fighting in place that you don't get if you relocate the combat) or a simple direct cost.

You define "other creatures" well enough in your own words:

when you initially encounter a critter, it shouts (typically) and [other] monsters are attracted *to that spot*

so pretty much whatever you'd end up fighting if you stood and fought instead of drawing some monsters back; or somewhat less than that.

If monsters shouted all the time, it'd be even more optimal than it is now to draw every fight back into cleared areas

OK, but you wouldn't be drawing back just whatever monsters *noticed you*. If you *just* draw back monsters that *notice you*, that's degenerate luring. But if you draw *them*, *and all their friends*, then that's tactical repositioning.

because shouting is generally louder than combat

I'm not qualified to tell what the perfect loudness for continuous shouting should be, but anything between zero and normal shouts would be an improvement over the current situation.

the problem with luring isn't *specifically* that it breaks apart packs or groups of monsters, it's that it's both tedious and optimal

ugh, how do you keep these apart in your mind? Luring breaks apart monsters. Breaking monsters apart is the property that makes luring tedious+optimal. Because breaking apart monsters is tedious+optimal. Luring inherits "tedious+optimal" from the breaking apart of monsters. Basic logic, no?

having it not break apart groups of monsters doesn't make it not-optimal, only slightly less powerful than it is now.

only slightly less powerful? ummm... I'd be happy enough with "slightly" nerfed luring, which just doesn't break apart groups anymore

So that reduces us to either: 1. We make it so there's no cost to fighting monsters you discover in place (it's equivalent to moving the fight elsewhere)

Yes, precisely...

which means pretty much the whole noise system needs to be thrown out in it's entiretly

WUT
I gave an example, which you understood, of how to do this with existing sound mechanics (continuous shouting). Granted, it's just an example. I'm sure you can contrive another system, for example: monsters get an 'alert' flag when they notice you, and awake monsters are visually attracted to 'alert'-flagged monsters or their last-seen location, which get priority over noise. But, I just think continuous shouting is simpler and easier to implement and debug, as well as discuss, making it better as a proposal.

WUT
I gave an example, which you understood, of how to do this with existing sound mechanics (continuous shouting). Granted, it's just an example. I'm sure you can contrive another system, for example: monsters get an 'alert' flag when they notice you, and awake monsters track these 'alert'-flagged monsters (like they'd track a player), which get priority over noise if they're more than ~2 tiles away from the awake monster. But, I just think continuous shouting is simpler and easier to implement and debug, as well as discuss, making it better as a proposal.

FIXED, what was I even thinking...

WUT
I gave an example, which you understood, of how to do this with existing sound mechanics (continuous shouting). Granted, it's just an example. I'm sure you can contrive another system, for example: monsters get an 'alert' flag when they notice you, and awake monsters track these 'alert'-flagged monsters (like they'd track a player), which get priority over noise if they're more than ~2 tiles away from the awake monster. But, I just think continuous shouting is simpler and easier to implement and debug, as well as discuss, making it better as a proposal.

FIXED, what was I even thinking...

It still doesn't make luring go away (which was your stated goal)

Continuous shouting doesn't make luring go away either, in both cases, it is optimal to get the creature(s) who are awake and have noticed you as far from any who haven't (and might not be in your los yet) as possible as quickly as you can.

It still doesn't make luring go away (which was your stated goal)

How does it not? I'll assume you're referring to the part in bold font: An orc warrior sees you, gets an 'alert' flag, shouts, and now everything in ITS line of sight is following IT as IT follows YOU, so you got a LOS-ful of monsters on your tail. You can't lure away the single orc warrior anymore, which is my main intention. It is the great Luring Nerf.

You define "other creatures" well enough in your own words:

when you initially encounter a critter, it shouts (typically) and [other] monsters are attracted *to that spot*

Ok, so i see at least part of the ccommunication problem here, the creatures which are attracted to any given noise (say a shout) aren't a pre defined set of creatures, each critter has a chance of being attracted to a noise, depending on how far away it is, what kinds of walls are in the way, whether it is asleep or wandering etc. And it may or may not actually find its way onto your los, it may give up and start wandering. The "set of things which would come find you if you just had the combat on the spot" is unknown and unknowable in the current system, and changes drastically if move even slightly, or change how much noise you make in combat.

We could go with the greatest extent (say, all things that could in the current system be attracted to your current position) but what noise do we use to determine it? The loudest one you could make? A specific one (say, a shout?) What if I stab something silently, does that also cause the tracking behavior? What about when I encounter one of the things I attracted, does it re trigger for the things that are in range of that creature? What about the things that are in range of the things that are attracted, does it chain? If not, how do we decide who the center point is?

So a couple points here: if the number of creatures attracted are based off the noise level of your current action, rather than an arbitrary one, it is optimal to lure, because you can always take a quiet action move away, and take a louder action (that is the current situation, you just remove randomness, which just makes luring more defined and predictable), if it is based on an arbitrary noise level then we need to move on to the secondary creatures, if secondary creatures might attract more creatures, then it is still optimal to lure so that they won't.

Then there is the interaction with stealth, if I don't wake up a critter and haven't made any noise yet, I assume we don't have anything awake and tracking us, but if I do, what if I move to a place where just that monster is in range of shouting and start shooting myself, or flinging fire storms, who comes running? Anyone? no one?

Defining "nearby creatures" in a consistent and specific way is not as simple as you have made it out to be, and noise is a pretty complex and subtle thing presently, you might have a specific set of rules in mind for how all that should work (it might even be a very good one), but I haven't heard it yet.

Whoa, this isn't about defining the outcome of an encounter, only the laws that dictate how it unfolds. You can't suppose that a solution to luring must involve some preordained selection of creatures that shall follow the player. You wouldn't need to set "the number of creatures attracted" nor consider the player's future or past actions. It is dynamically generated. On the fly.

That's how the current system does it too. It runs off a number of rules that monsters must follow. And with the current set of rules, you get such behaviour as "lured monsters leave their pack behind and wandering aimlessly". Does the game specifically order monsters to run away from their pack? No, but that's what happens.

Very simple example: you have 0 stealth, a monster wanders into your LOS, and you can't move or blink out of its LOS. Only one outcome is possible next turn: if it's still alive by then, the monster will notice you. Now, at this point, the game hasn't figured out that the monster will notice you. But it doesn't need to. It doesn't matter that the game doesn't know. So as a developer, you don't need to make the game know that. Thus the questions you ask do not need to be answered, yet the game will work properly.

Indeed, with constantly shouting monsters, you don't know how many will end up following you. Hell, maybe just one monster notices you, but before it gets any friends to notice you, you blow it up with a wand of disintegration, and walk away scot-free.

And you say a thing (what thing?) needs to be based on a noise level you must set, as a dev? Are you sure you didn't misunderstand what I mean by continuously shouting monsters? They just shout continuously - that's pretty much it - and it's this shouting that attracts other monsters, because shouting is noise, and noise attracts monsters.

Whoa, this isn't about defining the outcome of an encounter, only the laws that dictate how it unfolds. You can't suppose that a solution to luring must involve some preordained selection of creatures that shall follow the player. You wouldn't need to set "the number of creatures attracted" nor consider the player's future or past actions. It is dynamically generated. On the fly.

But that is my question, if it is dynamically generated, *what system do we use to dynamically generate it*

By definition it is either determined by player actions, or by some other pre set algorithm, those being the only two categories of things that exist.

Whoa, this isn't about defining the outcome of an encounter, only the laws that dictate how it unfolds. You can't suppose that a solution to luring must involve some preordained selection of creatures that shall follow the player. You wouldn't need to set "the number of creatures attracted" nor consider the player's future or past actions. It is dynamically generated. On the fly.

But that is my question, if it is dynamically generated, *what system do we use to dynamically generate it*

By definition it is either determined by player actions, or by some other pre set algorithm, those being the only two categories of things that exist.

We use the existing system, the system of noise. You understand noise. So what's the problem? What part of this is ambiguous? I don't see your question.

When you cast LRD on a wall, and monsters come to you, what system is used to dynamically generate the movement of monsters from their original position to you? The system of LRD making noise, and monsters being attracted to noise. Or, when you're following Qazlal, how do monsters keep finding you? Qazlal makes noise at your location, which is constantly changing (or not), and that noise draws monsters. It's perfectly analogous.

Player actions and pre set algorithms interact, you know? So I don't agree that dynamic generation is, by definition, determined by only one of those things.

BTW yet more evidence that luring is not an obvious tactic (this is from an old advice thread; newbies have to be told this, i.e. spoiled): "Many enemies make noise upon seeing you, but that only draws other enemies to the spot where they made the noise. (They don't shout continuously or anything.)"

What is ambiguous is what change you are proposing to noise (or that uses the existing noise system?) to prevent luring from occurring

I don't recall ever saying it was obvious, only that it was tedious and optimal.

Having monsters shout all the time doesn't make it not optimal, it just changes the optimal location for luring from "the nearest cleared area" to "upstairs" If you propose a secondary change to remove the ability to go upstairs (although I don't know how you do that while preserving retreating as an option, the mechanism is irrelevant to luring ) then you luring to the nearest cleared area is still optimal, just slightly less glaringly so than it was when critters didn't shout.

Also, yes some things are player actions, some things are programming, and sometimes it is a combination of the two, I didn't mention the "combination" alternative because it seemed obvious to me and that it didn't need explicitly stated, however the pedantry doesn't help your case at all.

It seems like you are trying to solve the "luring is too powerful when used in this one way" problem, when the actual problem isnt that it is particularly powerful it is that it is always the best choice, if you reduce its power, *but it is still the best choice* then you have accomplished nothing in terms of making the game more fun (you might have made it "more balanced" but that is unimportant)

What is ambiguous is what change you are proposing to noise

I'm not proposing any change to noise. Noise will work like noise works now. That's why I literally call it noise and not X-that-works-somewhat-like-noise.

Having monsters shout all the time doesn't make it not optimal, it just changes the optimal location for luring from "the nearest cleared area" to "upstairs"

How on earth are the two related?! They're independent! At least have the decency to explain your thought process or rescind. (I'm for the “monsters can't use stairs” option.)

still optimal, just slightly less glaringly so

True. (Probably true.) But. I don't care much, and neither should you, if “luring THAT HAULS ENTIRE PACKS TOGETHER“ remains optimal. The important thing is that luring a single creature at a time (which you pinpointed as the problem earlier in our conversation IIRC) stops being an option. 90% of the cheese in luring is gone. Problem solved, mission accomplished! Or would you rather bathe in cheese for a few more years until some perfectionist brings that 90% up to 100% or whatever your standard is? I addressed this a few messages back; apparently you ignored it:

OK, but you wouldn't be drawing back just whatever monsters *noticed you*. If you *just* draw back monsters that *notice you*, that's degenerate luring. But if you draw *them*, *and all their friends*, then that's tactical repositioning.

as well as how you keep calling new, non-degenerate luring “tedious and optimal”:

Luring breaks apart monsters. Breaking monsters apart is the property that makes luring tedious+optimal. Because breaking apart monsters is tedious+optimal. Luring inherits "tedious+optimal" from the breaking apart of monsters.

What pedantry do you accuse me of? My proposal is simplistic to the extreme. I don't see how your sentence here has anything to do with what we've said: “By definition it is either determined by player actions, or by some other pre set algorithm, those being the only two categories of things that exist.”

the actual problem isnt that it is particularly powerful it is that it is always the best choice

No, it's not. If the best choice is very very safe and boring and takes no decisions, then it's a big problem. If taking the best choice *leads* to such interesting situations as having to deal with entire packs at the same time, then it's not a big problem. People are tempted to lure precisely because of the huge gap in power between luring and staying to fight. If the tedium of luring stays the same, but its power is reduced dramatically, then the temptation to boringly lure is that much smaller.

you might have made it "more balanced" but that is unimportant

Ah, balance is unimportant apparently. So, why does all importance lie in making luring suboptimal sometimes? (which this actually does; for example, imagine a monster-filled bottle-like vault on an open level - it could be optimal to rush into bottleneck with monsters beyond it, instead of luring them out of the bottle and into the open.) Lots of things are optimal, always, like having better defenses, or killing monsters before they kill you, or winning. You conflate all the ways you can "lure", as if they would all be the same under new "luring". In many games you're crouching behind obstacles for cover all the time, and their gameplay isn't in the binary decision of whether to use cover or not, but in using/changing cover well, and the other things you do while under cover, since it's presumed that you use cover.

You shouldn't base a game around a degenerate element; basing a game around a legitimate element is... inevitable to some degree - there's no way around it. But it looks like you have never distinguished degenerate and legitimate luring.

True. (Probably true.) But. I don't care much, and neither should you, if “luring THAT HAULS ENTIRE PACKS TOGETHER“ remains optimal. The important thing is that luring a single creature at a time (which you pinpointed as the problem earlier in our conversation IIRC)

That's quite literally the exact opposite of my opinion, the ONLY thing I think is important is if a behavior is the best choice *AND* is tedious, if a thing is always the best choice, and makes the game boring, then it's bad for the game, whether it's "cheesy" or not.

If it's less powerful *but it's still the best option* then it doesn't matter if it's less powerful, people are still encouraged to do it, and if people are encouraged to do something boring, then it's bad for the game.

stops being an option. 90% of the cheese in luring is gone. Problem solved, mission accomplished! Or would you rather bathe in cheese for a few more years until some perfectionist brings that 90% up to 100% or whatever your standard is? I addressed this a few messages back; apparently you ignored it:

OK, but you wouldn't be drawing back just whatever monsters *noticed you*. If you *just* draw back monsters that *notice you*, that's degenerate luring. But if you draw *them*, *and all their friends*, then that's tactical repositioning.

as well as how you keep calling new, non-degenerate luring “tedious and optimal”:

Luring breaks apart monsters. Breaking monsters apart is the property that makes luring tedious+optimal. Because breaking apart monsters is tedious+optimal. Luring inherits "tedious+optimal" from the breaking apart of monsters.

That's incorrect, it's tedious because walking halfway across the level to fight something *is inherently tedious*, It's optimal because it's more powerful than any other option with no cost. Making it not break apart packs makes it less powerful, *but still more powerful than any other option and without any cost* and hence it's still tedious and optimal.

[dpeg]

Many thanks for the interesting read, HbG & Siegurt!

I really think that luring is a substantial problem, so I'll keep posting. So it has been established that the major problem is "drawing single monsters from wherever they were generated to a place you like better". This leads to a new possible solution, which I added below to my previous list (from page 1); I hope I got your proposal right.

1. The monster shouts (without losing an action).
2. The monster gets a speed boost.
3. Wake up asleep monsters on the level.
4. Relocate a monster from elsewhere on the level "on the other side of the player" (there is a direction monster-player). Keep its status (awake, asleep).
5. The monster does not follow you.

That new idea has its charm, because it tries to solve the problem at the root. However, something more would be needed: otherwise we'd change monster AI to not pursue the player, making every monster easy pickings for any sort of ranged damage.

You also discussed positive incentives: I don't think this can work to solve luring. It can be used at times, but there are way too many encounters to reward non-luring by trophies (whether items or xp).

Regarding the disinction between "good luring" and "bad luring": I don't make it. When I wrote (in a reply to HbG) that I'm not about scrapping luring wholesale, I just meant that the punishment I have in mind would not disable luring (like 5 above). Instead, I want every form of luring (i.e. having a monster follow you, regardless of why) to have a drawback. Due to the nature of the problem (monsters, threat assessment, risk mitigation), I hope that drawbacks along the following lines are good enough: (a) fewer monsters asleep, (b) more monsters aware of your position, (c) monsters relocated to worse position, e.g. clustering, eventually near stairs. Note these are all local, i.e. only use the monsters on a given level. In principle, it'd also be possible to have luring on this level inflict drawbacks on every other level -- that'd be the direction of the doom clock.

[lethediver]

Has anyone proposed just getting rid of the whole noise system yet?

It seems like removing the whole incentive to lure is a LOT simpler and more straightforward than creating punishments to counterbalance the incentive (which will inadvertently penalize a lot of other vaguely related things like pack breaking, straight up retreating, tactical movement, etc).

IF we operate on the assumptions:

A) Degenerate Luring defeats or removes all risks associated with combat noise/shouting

B) Degenerate Luring is impossible to fix without adding unintuitive, scattershot penalties that will worsen the game more than improve it (I happen to believe this after reading all the proposals here)

Then it seems to me the only conclusion is either live with it, or just get rid of noise for ALL players under the assumption that they would have done so anyway if playing hyper-optimally, and that therefore, combat noise is a mechanic which only adds tedium to the game.

To use an analogy here, luring is to noise what dropping your items before each fight was to item destruction. And the proposed solutions in this thread thusfar seem to be along the lines of, "Make the monsters get angrier and smarter when the player drops all his items." Perhaps removing noise itself would be the more elegant solution to this whole mess.

[HardboiledGargoyle]

dpeg, I think svendre half-heartedly suggested that monsters don't follow you, and nobody else.

So it has been established that the major problem is "drawing single monsters from wherever they were generated to a place you like better".

That is exactly opposite to Siegurt's opinion, as he established most definitely in his last message. If luring has a chance of making you face 20 creatures instead of 21, it's still optimal, and therefore as problematic as it ever was.

I want every form of luring (i.e. having a monster follow you, regardless of why) to have a drawback.

But really? Even accidentally or carelessly letting a monster follow you, or pillar dancing? Does the monster have to be visible? And you would communicate this to the player?

Noise is just a mechanic designed to do something. I don't think a mechanic can be tagged as inherently problematic purely because a half-assed implementation makes it not do what it's supposed to do. Stabbing aside, if you get rid of noise without compensation, then luring monsters into 1v1 fights wouldn't even take time or effort, and would resemble pokemon-style "let's have a duel" combat. IMO Crawl doesn't have what it takes to make that kind of game interesting, and should force group fights and ganging. If no noise, then you need at least something like monsters waking each other up, and communicating the player's position, by looking.

[Siegurt]

So it has been established that the major problem is "drawing single monsters from wherever they were generated to a place you like better".

That is exactly opposite to Siegurt's opinion, as he established most definitely in his last message. If luring has a chance of making you face 20 creatures instead of 21, it's still optimal, and therefore as problematic as it ever was.

That is acutally a mis characterization of my opinion.

I believe that it is bad if luring is *always optimal with no cost* the largest chunk of "always" is "you encounter a creature by itself" therefore dpegs statement isn't directly opposed to my opinion at all.

I would like it the most if every situation had a plausible use for, and commensurate drawback for, using luring, presenting the player with an interesting choice. Aka "is it worth it for this instance"

Second choice for me would be if luring was occasionally optimal, and the rest of the time is was clearly not optimal, this degrades it from interesting choice, to "follow this script for luring or not"

Third choice for me would be "luring is never optimal" this effectively removes a large chunk of the tactical positioning that I personally find interesting, but it would probably be barely worth it to get rid of something that is presently always optimal and tedious.

Fourth choice is the status quo, and to be clear, it isn't a terrible option, as is luring is only marginally optimal in most situations, and ignoring it in all but some of the more critical ones, and not implementing it in the most tedious way possible is only very slightly detrimental to your win rate.

Fifth choice would be a lot of the terrible suggestions made on this topic, mostly designed around addressing the wrong problem.

[HardboiledGargoyle]

I believe that it is bad if luring is *always optimal with no cost* the largest chunk of "always" is "you encounter a creature by itself" therefore dpegs statement isn't directly opposed to my opinion at all.

Sorry, you cannot use such pithy repetitions with tiny add-ons each time and expect to be understood.

bad part of luring = *[you encounter a creature by itself] optimal with no cost*

^ you tell me what that means and how it's compatible with dpegs statement

I almost quoted you. You said it does not matter
whether luring is much more optimal or a tiny bit more optimal
whether a lured creature breaks away from its pack or not
and that nothing is solved if the optimal tactic is nerfed but remains optimal.

But maybe that's not what you meant at all; it's hard to tell.

[Siegurt]

I believe that it is bad if luring is *always optimal with no cost* the largest chunk of "always" is "you encounter a creature by itself" therefore dpegs statement isn't directly opposed to my opinion at all.

Sorry, you cannot use such pithy repetitions with tiny add-ons each time and expect to be understood.

bad part of luring = *[you encounter a creature by itself] optimal with no cost*

^ you tell me what that means and how it's compatible with dpegs statement

I almost quoted you. You said it does not matter
whether luring is much more optimal or a tiny bit more optimal
whether a lured creature breaks away from its pack or not
and that nothing is solved if the optimal tactic is nerfed but remains optimal.

But maybe that's not what you meant at all; it's hard to tell.

Dpeg said the major problem with luring is luring a single creature by himself, I said luring is bad if it is always optimal with no cost, if I consider the largest number of cases where you could lure are when you encounter a creature by itself, then the largest part of the problem (percentage wise) is with luring a single creature by itself.

That isn't it say that I believe the problem would be eliminated if you made it only optimal to lure things that weren't alone, I don't. But I do believe that if you did that, you would make the problem smaller. I also believe that there are ways to make the problem smaller in a more interesting way than arbitrarily drawing a line between encountering a creature alone or not, which I am not even sure is even relevant at all to what dpeg was saying.

[HardboiledGargoyle]

Dpeg said the major problem with luring is luring a single creature by himself, I said luring is bad if it is always optimal with no cost, if I consider the largest number of cases where you could lure are when you encounter a creature by itself, then the largest part of the problem (percentage wise) is with luring a single creature by itself.

Your conclusions may have some validity, but the way you think about it logically is nonsense, and contradicts what you were slowly hammering away at in that long conversation of ours. What you say here is like if I said that guns are bad for some reason, and most guns are black, then the largest part of the problem with guns is that they're black...

That isn't it say that I believe the problem would be eliminated if you made it only optimal to lure things that weren't alone, I don't.

...and if I further said that the problem would not be eliminated if you made only non-black guns bad, but it would be made smalller:

But I do believe that if you [made it only optimal to lure things that weren't alone], you would make the problem smaller.

On purely theoretical, axiomatic grounds, I guess?

I also believe that there are ways to make the problem smaller in a more interesting way than arbitrarily drawing a line between encountering a creature alone or not, which I am not even sure is even relevant at all to what dpeg was saying.

Dude, like, you are the only one trying to arbitrarily draw lines. Here you are imagining lines drawn between encountering a creature alone or not. In our PM conversation you were trying to draw other lines, regarding which monsters would be attracted by continuous shouting and which wouldn't. I'm pretty confident dpeg singled out single monsters because that's just the apex of degeneracy.

"encounter a creature by itself" is not even meaningful, it's so vague I'm offended you assume I'd understand what you're thinking about. An "encounter" can be defined as some episode that happened, with the player in a neutral state (alone and not pursued) before and after, like "I met Sigmund and he confused me but it wore off and I ran away" or "I found an orc pack and lured a warrior away and killed it" or "I found 7 yaks and drew them into a corridor and killed them". You may have meant "get noticed by a single creature", but you'd have to expand on that...

I apologize to everyone reading this thread for the exhaustive deconstruction. Don't mind us.

Also, stop using run-on sentences. They're hard to read. Periods are our friends.

[Siegurt]

Dpeg said the major problem with luring is luring a single creature by himself, I said luring is bad if it is always optimal with no cost, if I consider the largest number of cases where you could lure are when you encounter a creature by itself, then the largest part of the problem (percentage wise) is with luring a single creature by itself.

Your conclusions may have some validity, but the way you think about it logically is nonsense, and contradicts what you were slowly hammering away at in that long conversation of ours. What you say here is like if I said that guns are bad for some reason, and most guns are black, then the largest part of the problem with guns is that they're black...

That isn't it say that I believe the problem would be eliminated if you made it only optimal to lure things that weren't alone, I don't.

...and if I further said that the problem would not be eliminated if you made only non-black guns bad, but it would be made smalller:

But I do believe that if you [made it only optimal to lure things that weren't alone], you would make the problem smaller.

On purely theoretical, axiomatic grounds, I guess?

More like In the same way that if all black guns were destroyed I think any problem that might exist with guns generally would be lessened, just by virtue of sheer numbers involved.

I also believe that there are ways to make the problem smaller in a more interesting way than arbitrarily drawing a line between encountering a creature alone or not, which I am not even sure is even relevant at all to what dpeg was saying.

Dude, like, you are the only one trying to arbitrarily draw lines. Here you are imagining lines drawn between encountering a creature alone or not. In our PM conversation you were trying to draw other lines, regarding which monsters would be attracted by continuous shouting and which wouldn't. I'm pretty confident dpeg singled out single monsters because that's just the apex of degeneracy.

"encounter a creature by itself" is not even meaningful, it's so vague I'm offended you assume I'd understand what you're thinking about. An "encounter" can be defined as some episode that happened, with the player in a neutral state (alone and not pursued) before and after, like "I met Sigmund and he confused me but it wore off and I ran away" or "I found an orc pack and lured a warrior away and killed it" or "I found 7 yaks and drew them into a corridor and killed them". You may have meant "get noticed by a single creature", but you'd have to expand on that...

When I say "encounter" I mean "discover" or "come across", and by "a creature by itself" I was trying to mean the same thing as dpeg meant when he said "single monsters".

For future information when I say "creature by itself" I intend it to mean the same thing as "single monster".

[HardboiledGargoyle]

That remains undefined since when someone like dpeg talks about a single monster, they're talking about luring it, to the exclusion of other monsters, regardless of whether or not it is found single. But if you've only encountered it, you have not started luring. So you open up the question of what it means for a monster to be single when encountered. But the point is moot since any mechanic that separates single- vs multiple- monster encounters, rewards luring in one case, and punishes in the other, would be weird.

More like In the same way that if all black guns were destroyed I think any problem that might exist with guns generally would be lessened, just by virtue of sheer numbers involved.

Then wouldn't you like continuous shouting, since it would lessen the number of times you'd "lure"? Because instead of being able to lure 50 times on a level, once per monster, you'd only be able to "lure" 5 times on a level, with 10 monsters lured each time. Roughly. It would be like destroying all black guns and filling most of the rest with blanks. Noise is not broken - we know that Qazlal's conduct works. What's broken is monsters not really using noise. Imagine if Qazlal made a noise only once per level.

[dpeg]

Sorry to interrupt the domestic quarrel, but luring is always problematic, no matter how far you lure, or how many monsters. When I emphasised *single* above, I meant that you can usually just do that: no matter how many monsters are together, autoexplore will stop when you see just one, and you can diligently pull that one back to where you came from.

It happens that you lure several monsters, or even a whole pack, and that is in no way better than luring them one by one (it is more rare).

However, I am not sure if these semantic digressions move us forward to eliminating luring.

[HardboiledGargoyle]

but dpeg, if luring pulls back the same number of monsters that you would have faced had you not lured, does luring remain problematic in the same way? Or does it become problematic in a qualitatively different and far less harmful way? I think it's quite obviously the latter.

To use an analogy here, luring is to noise what dropping your items before each fight was to item destruction. And the proposed solutions in this thread thusfar seem to be along the lines of, "Make the monsters get angrier and smarter when the player drops all his items." Perhaps removing noise itself would be the more elegant solution to this whole mess.

that analogy can be broken down further, letthediver. Noise and item destruction were both solutions. But...

It looks complicated because there are problems, and solutions to problems, and problems with solutions. There are multiple solutions to original problems, and multiple solutions to problems with the solutions to the original problems. It's a tree of problems giving rise to solutions which give rise to problems. And you have to decide which path to take, what combination of problem and solution to settle on, whether to change the solution or fix the problems spawned by the solution. By the way, I think it's easy enough to see original problems, and a little harder to see why solutions have problems, and I find that it's often unnecessary for the solutions to have those problems, problems which I shall liken to loopholes. And while it may seem modest, I think closing loopholes can sometimes be more effective than scrapping the solution to look for another one, at least because the particular solution creates interesting gameplay. Sometimes, not.

So in trying to understand the example of item destruction, I think we can do a breakdown like this:
Issue: Consumables can be avoided through patient/tedious play; people hoard consumables; people get too many consumables available to them at any one time.
Solution: Give some monsters the ability to destroy consumables. It destroys some %, rather than #, of all your consumables. So theoretically, all characters maintain some reasonable equilibrium between linear gain/expenditure of consumables and their exponential decay via destruction. People are thus encouraged to use consumables, potentially even to use those consumables to prevent their very destruction, instead of letting them go to waste, and instead of tediously grinding the game down with renewable resources.
Loophole in solution: Players can drop scrolls and potions and carry only 1-2 of any one type as a matter of course, and drop everything before knowingly entering an item-destroying encounter.
Closing loophole indirectly: Make dropping each consumable take time individually. Make consumables get destroyed even when they're on the ground. Set the rate of destruction really high so you don't even have time to disengage, unless you wear a certain item that is otherwise useless, which you'd rather not do by default. "Make the monsters get angrier and smarter when the player drops all his items."
Closing loophole directly: Goldify consumables - take them out of the inventory, make dropping consumables impossible, and make consumables get picked up immediately when found anywhere in LOS.
Altering original solution: Ignore the issue. Reduce the generation of consumables to compensate. Mimic the mechanic with -Potion/-Scroll statuses.

I think we can do the same for luring.
Issue: Fighting a monster one on one is, and must be, survivable + you can fully recover from a fight by resting + it's fully feasible just fight one monster and rest = boring game.
Solution: Noise. Monsters make it + fighting makes it + monsters are attracted to it = you're not fighting monsters one on one anymore, removing an essential ingredient from the above recipe for a boring game.
Loophole in solution: Monsters make noise only when they first notice you, and other monsters don't necessarily notice you, so you can relocate, just yourself and select monsters, to a cleared area, re-establishing the boring game recipe.
Closing loophole indirectly: Most of the solutions in this thread. Or, making combat/shouting noise so loud that it propagates through the entire level.
Closing loophole directly: Monsters don't suddenly stop making noise after they first notice you (yeees). Or, shouts convey the player's location to everything that hears the shout. Or, levels are so small that combat is heard throughout the level and the time you buy by luring a monster is about the same as the time it takes to kill the lured monster.
Altering original solution: Substitute or augment the noise system with improved monster co-ordination (works too). Or, remove noise and make 1v1 interesting. Or, play DD.

Here's an imaginary scenario for fun:
Issue: The game has too much piety, gods are OP, their abilities have no cost, whatever.
Solution: The game has unavoidable creatures (dancing runes?) that drain your piety on hit.
Loophole in solution: The game lets you abandon your god, and re-join with the same amount of piety.
Closing loophole indirectly: The piety-draining creatures go berserk if you have no god, and drain your XP instead.
Closing loophole directly: You can't abandon and re-join with the same amount of piety you had before
Altering original solution: nerf piety gain, give piety cost to active abilities.

[Ceann]

The things I have seen described want to essentially mesmerize or effectively root the player and force them to fight anything they see, be this by swiftness or putting monsters behind them etc etc. These actions aren't really dungeon crawl though, like its more like the old D&D goldbox games where your party gets into a fight with a group of monsters and you don't have a route to flee, so you just have a battle to the death. That is not RNG "sorry Dynast", RNG is getting LCS'd for 92 and dying in one shot as a Sp, or getting Throw Icicle'd by Frances for 64, falling down a shaft into the middle of a group of monsters, or a teleport trap doing the same, that is RNG. Getting almost killed by an Ogre hit on D:4 and running away from it is not "avoiding RNG, or refusing to roll the dice", it is the correct tactical decision. Forcing fights removes tactics from the game and makes it into an RNG right cross fest, whoever connects the most times first wins, is not how the game should be played.

I think that the real problem here is the separation of degenerative luring from other tactics.

My understanding thus far of optimal luring is.... any time you see monsters, whether it be a single monster, or a pack or an entire room etc, your optimal way to play is pull as small a number as possible the amount of visible monsters back into a previously cleared room or area, or far enough back that you have confidence that any noise generated will not be able to draw additional monsters. Essentially reducing your exposure to events that can kill you by the largest amount.

If that is an inaccurate description of optimal luring, let me know.

If I am a Spriggan or a Centaur or a Felid, Boots of Running it is difficult to separate optimal luring "to be further known as OL" from kiting, moving to a kill hole,grouping, positioning to maximize AOE of some kind etc.

My problem with a solution for OL.

As a race with an inherent speed bonus, isn't a player suppose to use it to their advantage?

Should not the solution provided, not have impact on other valid tactics?

Turn based movement, movement speeds and combat are so intricately linked, how do you target the specific kinds of movement related to OL without affecting other movement?

Personally I have maybe a handful of times that I perform actions that could be interpreted as OL, I would say that most combat you really don't even need to consider OL is any fashion. While I understand that it is an option, in my experience from the games I watch, I don't think the number of people that perform OL in a dedicated fashion "like every pull" is very high.

Maybe there are some statistics that I am unaware of but this seems akin to prohibition to me in a fashion... essentially sure everyone lures a little bit but only some take it to a degenerative level that can be classified as OL. So it seems that everyone must be punished because of the actions some few take, is the impression I am getting.

Summarily so far, all the solutions I have seen probably introduce more harm than benefit, which I am assuming is the main problem, of this problem.

[lethediver]

Great overview HG.

Here's my issue with "improved monster coordination" - i think most attempts to use AI to "outwit" players using exploitative tactics just opens up a different set of exploits. Players are always smarter.

For instance, what happens when the player begins to anticipate melee attackers taking up positions behind corners and uses AoE spells to hit them from out of range? Or he notices smiters/archers getting into nice spread out formations and camping in the rooms he previous fled from... so he simply places exclusions and avoids those rooms making the level more trivial to clear? To use the borderline reductio ad absurdum example: monsters trying to flee back if they find themselves isolated = player will simply pelt them from range.

The only real effect of all this would be to make the game more spoiler-y. Vets would quickly pick up on the new AI behaviors and how to exploit them, whereas noobs would die more frequently due to not anticipating them. Therefore, the people the problem was originally aimed at (vets abusing luring) would be unaffected. (That's not even mentioning the field day one could have with Ashenzari or DetMon.)

Now, on to my issues with shouting. One, not every monster shouts. One of the most luring-encouraging areas in DCSS is the slime pits (due to lack of choke points and corners ie acid walls), but very few (none?) monsters here shout. Swamp is another light on shouters IIRC. You might call these edge cases but I think the fact that this solution has absolutely no impact on them is... troubling.

Two, continuous shouting effects melee chars a lot -- however ranged/magic chars will likely still be able to defeat it, just with more tedium. A ranged char or blaster can snipe a single enemy within a few turns usually. This is essentially the equivalent of a single shout - the player still has plenty distance to run unless A) other monsters were very close to the first, allowing them to get into range starting an endless chain of shouting. AND B) the player cannot outrun these other monsters, which becomes less likely as the game goes on given the preponderance of translocations spells/haste/swift.

Three, even characters who cannot outrun monsters can still go up the stairs. Now, lots of shouting around stairs MIGHT eventually result in areas near stairs having too many monsters nearby for luring to be effective, but it's not clear that it MUST.

In essence, the new exploit becomes "pick off a few enemies then translocate or swift/haste away OR go up the stairs and come back down a new one." In the end, therefore, luring still exists, it's just even more tedious than before. Given that the original goal was to block off a tedious-yet-optimal tactic so players wouldn't *have* to use it in order to play optimally, simply making that option more tedious does not feel like a win.

[HardboiledGargoyle]

My understanding thus far of optimal luring is.... any time you see monsters, whether it be a single monster, or a pack or an entire room etc, your optimal way to play is pull as small a number as possible the amount of visible monsters back into a previously cleared room or area, or far enough back that you have confidence that any noise generated will not be able to draw additional monsters. Essentially reducing your exposure to events that can kill you by the largest amount.

I think that's right. Only, I would rather call it Degenerate Luring rather than Optimal Luring, where Degenerate just means "optimal in a bad/unfun way". Unlike Siegurt, I believe making the optimal less bad and less unfun is good enough. And I would say optimal luring becomes degenerate when "the amount of visible monsters" is decreased by pulling them back. Just making it not decrease would take the degeneracy out of optimality.

Should not the solution provided, not have impact on other valid tactics?

You should still be able to do the other things you mentioned ("kiting, moving to a kill hole,grouping, positioning to maximize AOE of some kind etc") if the situation allows it.

Turn based movement, movement speeds and combat are so intricately linked, how do you target the specific kinds of movement related to OL without affecting other movement?

By targeting not the movement related to OL, but the other stuff related to OL. Good question.

Summarily so far, all the solutions I have seen probably introduce more harm than benefit, which I am assuming is the main problem, of this problem.

Well, you think that stealth is useless unless monsters shut up 1 turn after they notice you and shout, so this statement has to be taken with a grain of salt.

Here's my issue with "improved monster coordination" -

Those issues are valid. Sorry, when I wrote "improved monster coordination" I had something simple in mind like "if a monster notices you, it makes nearby monsters notice you too"

Now, on to my issues with shouting. One, not every monster shouts.

More can be said about silent enemies, but let's focus on the vast shouting majority.

A ranged char or blaster can snipe a single enemy within a few turns usually. This is essentially the equivalent of a single shout - the player still has plenty distance to run

Yes, and in this case you're not luring monsters, because they don't follow you. You kill them on the spot, or run before they notice you. I don't think that "killing monsters very quickly" and "running from a monster and letting it wander" qualify as tedium. But it would be harder for melee chars to pull this off.

too many monsters nearby for luring to be effective

I don't understand this part at all

In essence, the new exploit becomes "pick off a few enemies then translocate or swift/haste away OR go up the stairs and come back down a new one." In the end, therefore, luring still exists, it's just even more tedious than before.

That's not luring, that's... roughly what play without luring looks like. It's not "more tedious" or exploitive, it's just... gameplay, you know? Victory!

[lethediver]

too many monsters nearby for luring to be effective

I don't understand this part at all

In essence, the new exploit becomes "pick off a few enemies then translocate or swift/haste away OR go up the stairs and come back down a new one." In the end, therefore, luring still exists, it's just even more tedious than before.

That's not luring, that's... roughly what play without luring looks like. It's not "more tedious" or exploitive, it's just... gameplay, you know? Victory!

Point A: Basically if each stair has monsters in LoS, you can no longer effectively stairdance to isolate monsters. If a player tries to kill one monster near stair + ascend + go down other stairs too often, it could happen, and could be a hard limit on how much players can abuse stairs... however, I don't think it's likely to happen. The fact that it's so convoluted to explain maybe proves how unlikely it is to happen.

Point B: Is there a real difference though? If you kill a single monster, then run from the reinforcements (so you only had to fight 1v1) is it qualitatively different from seeing the monster, running, then fighting it 1v1? Personally, I don't think there is a huge difference. You're using the same resources (already cleared retreat path, possibly a stair, possibly some swift/haste/other source of speed) to achieve the same effect (1v1 fights only) by defeating the same mechanic (noise).

We would need a branch with continuous shouting added to know for sure, but my gut tells me that players will still work out a tedious method to achieve 1v1 fights and avoid the intended effect of noise/shouting. They'll just use stairs + speed + translocations more to pull it off.

[HardboiledGargoyle]

Point A: If it's something that would very rarely happen then we can ignore it, I guess.

Point B: To fight a monster 1v1 with an escape option, you need a corridor, which have been diligently removed in many places, and even then, monsters from the back can swap with the monster in front, so that suddenly the monster in front of you has full HP. If your corridor isn't twisty like shown below, then monsters can buff allies, summon, smite you, blink behind you, stuff like that. And you're bracing yourself to run away, which adds a tingle of anticipation that is missing from the situation where you've lured a monster into an actual 1v1 fight. I actually don't think you need them to do the thing that you're talking about, but speed and translocations are usually/often limited, whereas Degenerate Luring is universally accessible.

  Code:

##y#
#y##
##@#
#.##

[crate]

Since I'm coming in late, would someone like to summarize the actual problem being discussed here in, let's say, three sentences or less? I have a few ideas about what you're discussing, but there are too many words in this topic for me to be sure.

My guess is that the problem in question is summed up as: "Moving toward monsters is almost always an unambiguously bad action", and the desire is to see that it is as often a good action as a bad action. Is this correct?

If this is the case I do not think this is a fixable "problem". The player has limited knowledge of unexplored terrain, and near-perfect knowledge of explored terrain. Especially for characters with limited range, it will always be better to move the fight to the location with the best terrain possible for the fight, and doubly so against enemies that have ranged attacks. Note this is true even before you account for the possibility of enemies in unexplored areas; removing noise entirely or such does not do much here. You could artificially restrict player motion, or turn crawl into a giant arena-fighter (like meatsprint but without the pillars or edges), but those games are pretty clearly not crawl. I guess you could also alter monster generation so that it emphasizes placing monsters specifically in places the player has already explored, but I really don't think I would enjoy playing crawl much if that change were made.

(I also don't personally think this "problem" is actually a problem. Positioning is interesting.)

If you really want to make players charge in against monsters then the only real option I see is to implement a very harsh clock of some sort for each individual monster. (Because crawl regen is so slow, it must be per-monster; otherwise retreating to better terrain will prevent enough damage that a global clock actually only encourages luring.) But any implementation of such a clock (it must be very harsh to encourage any sort of forward movement) is sure to punish weak characters severely, which surely causes greater problems than it solves.

I would support making the most common movement speed for monsters 11 instead of 10, but that does almost absolutely nothing about most forms of "luring", since you have about fifty tiles before the monster actually catches up with you in most situations (plus an additional nine or ten every time it hits you as you move to more advantageous terrain). Going much past 11 changes the current crawl paradigm of choosing which fights you take into one where you must fight the enemies you encounter. This is okay, but it's a pretty radical change.

If the real problem is that you are bored of how similar "luring" is all game, I suggest implementing the lair level generator "decay" or whatever in all of crawl, since lair terrain is far and away the best in the game in terms of producing interesting movement paths during combat. It's really a shame it's wasted in a branch that has so few ranged attackers.

[sanka]

crate:
I think that some players think that it is optimal to lure every single monster across half the level, and that it is more boring than charging the monsters and pressing tab or something, so it should be make not the optimal tactics. I think this was the original problem, altough most proposals do not address this.

Thanks for mentioning the lair level generator, that's a really good idea.