Please keep in mind when balancing

[Yet Another Stupid Noob]

Please do not take offense to these comments; they are simply a general reminder/request.

I do not know if this was the trend in the past (started playing in 0.7), but it seems (to me) that most of the balance discussions lately have been referring to zigs/extended endgame. Judging by personal experience and a few months of lurking the forums, I would say that 95% of the people spend 95% of their time pre-rune. Therefore I suggest that the game be balanced around the 3-rune victory.

I agree that spell casters cap out in power higher than melee, and hybrids cap out higher than either, but I am not sure that is a problem. In a 3-rune victory, even melee classes generally will have not reached their cap. If you can win before any of the fighting styles peak out, how much does it matter the one peaks out higher than another? By way of example, I when I started playing I spent a long time trying to get naga casters off the ground because I read in the wiki, (yeah, silly me) that they can be the strongest endgame casters. It took me a lot of splats to realize that a little power early, (read: normal speed and a larger mana pool) is often better than a lot of power at a time I will never reach because I die too soon. Similarly, balancing it so endgame casters cannot easily throw out 5 firestorms and annihilate everything is not optimal if said balance keeps people from having the power to clear the vaults.

At the risk of damaging my credibility, I would like to make an analogy to MMOs. Many MMOs I have read about/played use high end raiders as a suggestion/balancing point. They have the most through knowledge of game mechanics, make the most lucid suggestions, and are willing to do the often hard work of actually testing (coding) ideas. While this is frequently the only feasible way to actually realize ideas, it has the side effect of tuning the game to be optimal for said high end raiders. Sadly this can have the side effect of leaving the casual players who make up the bulk of the player base out in the cold. A MMO casual who feels bored and disenfranchised will stop playing (and paying). This is less of an issue for crawl, but even crawl depends on casual players becoming involved in order to replace people who become burned out or leave due to other issues.

If one is to balance the early game, I would go back to the early spells. Perhaps I don't optimize my characters well, but I consider mephitic cloud to be more unbalanced than haste or controlled blink since I can get the cloud online before D5, and it trivializes most combats until around D17, including the early branches. Conversely, by the time I have the spell of haste or controlled blink and have made it reliably useful (i.e. have the mana/other spells and/or other combat skills to exploit the benefit), I am generally getting ready to start collecting runes and victory is pretty close. Judging by my own win percentage, I probably spend 70% of my time in the former situation and only about 15% of my time in the latter.

I guess my point is to make a request to keep at least a little focus on the casual player. I am not saying the devteam is not doing that (the training revamp and perma-buffs are ideal in this regard), but dismissive comments that the food mini-game is irrelevant bother me a bit. I agree that post-lair food generally is not an issue, but given the amount of time that most players spend pre-lair, I still consider it rather relevant. Likewise, in my experience, if I am able to spam high level spells and ignore the food/mana cost I am either about to win, or could have already and am working on the extended endgame.

[ElectricAlbatross]

I agree with your overall point, and would like to add that this—

95% of the people spend 95% of their time pre-rune

is particularly a problem that should be tackled. The early game is the most boring part of the game (imo), so it's not good that a large majority of players have to spend most or all of their time there. In general the early game should be easier (and not just in Crawl Light).

[dpeg]

Noob: I don't know if by "discussions" you are referring to this forum or internal discussions.

In any case, we are aware of the fact that most players never see Zot. Neither content nor balance is focused around the endgame (let alone the extended endgame). A quick look at the changelogs should show this.

[njvack]

Note that KoboldLord and minmay and the other experts here, despite their encyclopedic knowledge of Crawl, aren't devs -- the devteam's names show up in orange. And I've never gotten the sense that the devs give more weight to people who can play near-optimally than people who play like, say, me.

AFAIK, most of the dev team doesn't hang out in the Tavern much, if at all; most dev discussion happens in other places.

[ebarrett]

And here I thought the game had been getting a lot easier since 0.6! Besides, the extended endgame has barely been changed at all, ever, when compared to what goes on during a standard 3-rune game. High-level spells might get changed more often than anything else because this kind of thing is always harder to balance in any game (there's a fine line between "worth the investment" and "broken", and boy are the best players quick to take advantage when stuff gets broken), but that's it.

[roctavian]

Personally, I just want to point out for everyone that YASN's post is awesome and respectful -- far too many posts recently have been inflammatory or outright hostile. This is a much better way to advocate player interests.

[Sealer]

I'd like to point out that I play, and was playing this game for the last 3 (or 4, don't remember) years BECAUSE of its difficulty. I believe it is one of the goals of the game, an original and ambitious one to make the game ridiculously difficult, but very rewarding if mastered, and that is exactly how I feel about it. Note that in all that time I've ascended 2 3(or 4) runers.

[KoboldLord]

I wouldn't read too much into the series of Ziggurat-related threads recently. Once a thread on a particular topic crops up, it gets everybody thinking about that topic. Once everybody's thinking about that topic, all the people who have a half-formed idea in their head that's somewhat related to that topic will be more motivated to finish forming that idea, and at least a few of those people will decide they like that idea enough to post it. So really, it's only to be expected that thread topics will come in streaks, and we shouldn't read any particular motivation or philosophy into that quirk.

As far as balancing for a 3-Rune game goes, I can see the argument to a certain extent but you can also take it too far. Honestly, a 15-Rune game is not all that much more notable than a 3-Rune game, since the difficulty curve flattens out a LOT after getting your first Lair branch Rune. Mennas might kill the heck out of you the first time you see him, and he might catch you unawares even after you've got a few ascensions under your belt, but learning to deal with him isn't going to fundamentally challenge your understanding of the game like learning to deal with Sigmund or orc priests did. In the end, the balance problems that you encounter in 15-Rune games are going to be related to the problems that you encounter in 3-Rune games, and a fix that screws up 3-Rune balance was probably a bad fix for 15-Rune balance too.

I'd also like to point out that there's an unfortunate history of using 3-Rune balance as a copout. A few versions ago, Kiku was quite a bit different than she is now. You got books that were pretty similar, powerful then just as they are now, but by the third Rune you had redundant copies of all those books and she didn't give you any useful abilities at all to compensate for the fact that she'd been doing nothing but loafing around since the end of Lair. You didn't get piety for demon kills at that time, either, so your piety with Kiku would slowly degrade through Tomb, Hell, and Pan and there was nothing you could do about it. At that time, some people argued that the status quo was fine because Kiku was balanced for a 3-Rune game, and 15-Rune balance is comparatively unimportant. Fortunately the other side won, and we now have a version of Kiku who is useful, powerful, *and* interesting for the entire game without any negative impact on 3-Rune balance at all.

[greepish]

I wouldn't read too much into the series of Ziggurat-related threads recently. Once a thread on a particular topic crops up, it gets everybody thinking about that topic. Once everybody's thinking about that topic, all the people who have a half-formed idea in their head that's somewhat related to that topic will be more motivated to finish forming that idea, and at least a few of those people will decide they like that idea enough to post it. So really, it's only to be expected that thread topics will come in streaks, and we shouldn't read any particular motivation or philosophy into that quirk.

As far as balancing for a 3-Rune game goes, I can see the argument to a certain extent but you can also take it too far. Honestly, a 15-Rune game is not all that much more notable than a 3-Rune game, since the difficulty curve flattens out a LOT after getting your first Lair branch Rune. Mennas might kill the heck out of you the first time you see him, and he might catch you unawares even after you've got a few ascensions under your belt, but learning to deal with him isn't going to fundamentally challenge your understanding of the game like learning to deal with Sigmund or orc priests did. In the end, the balance problems that you encounter in 15-Rune games are going to be related to the problems that you encounter in 3-Rune games, and a fix that screws up 3-Rune balance was probably a bad fix for 15-Rune balance too.

I'd also like to point out that there's an unfortunate history of using 3-Rune balance as a copout. A few versions ago, Kiku was quite a bit different than she is now. You got books that were pretty similar, powerful then just as they are now, but by the third Rune you had redundant copies of all those books and she didn't give you any useful abilities at all to compensate for the fact that she'd been doing nothing but loafing around since the end of Lair. You didn't get piety for demon kills at that time, either, so your piety with Kiku would slowly degrade through Tomb, Hell, and Pan and there was nothing you could do about it. At that time, some people argued that the status quo was fine because Kiku was balanced for a 3-Rune game, and 15-Rune balance is comparatively unimportant. Fortunately the other side won, and we now have a version of Kiku who is useful, powerful, *and* interesting for the entire game without any negative impact on 3-Rune balance at all.

This. Not to mention, when you get down to it 3 runing doesn't need a HUGE amount of blancing out left. Specfically about your mephitic cloud comment consider the starting class books it's in. Wizard: a risky class since you're screwed if you don't get a decent book eventually. Venom mage: A class completely unable to deal with many types of monsters without spamming a rather inferior level 5 spell. Really only conjurer is perhaps undeserving of it but I'm not good enough to really know for sure :/

[dondy]

i've only had about 3-4 decent runs for a 3 rune win (v0.9.1 - so take my comments with a grain of salt) and i must say that it bugs me how, pre-lair is based on pure luck and somewhat random decision making (early temple/altar, rpois, curare kobolds, ...).

while one could argue that the early game should be allowed to be alot harder and random - since you didn't spend alot of time in it - i think that is quite a bogus argument, since ideally you'd want a linear progression of both the character and the game difficulty

i feel like lair and somewhat later branches are alot more reasonable to react to (besides getting your ass kicked by a unique cluster fuck) since you had time to gather some tools to handle some situations.

and finally, slime/hell/pan/tomb seem to be really tough again, combined with the terror that is orb of fire from zot (the only thing that killed one of those was my quickblade vampire with rf+++)...
at least that's what my impression was.

(not trying to get off topic here... but a decent ammount of information does kinda change the difficulty of the game)
additionally, i'd like some better (clearer) ingame descriptions - like how every game i HAVE to have access to the knowledge bot (even if it's only for precise monster descriptions) - which kinda seems wrong to me, since it makes the game quite impossible to play unspoiled.

generally, the information is there but i have to access it through a different interface (because number-ranges are mapped to words which can easily be misinterpreted) - rather than having it directly at hand.

good examples are weapon speed/skill - unspoiled you might simply go for lv27 in your weapon skill which is a HUGE miss-investment, if you're spoiled you'd most definitely won't do that.

magic resistance: and how it affects your chance of successfully applying your hex, the monster and "hit" descriptions are WAY too hard to evaluate correctly besides immunities (immune is immune and unaffected stays unaffected), thus you might just give it a go and get splattered.

shield skill: similar to weapon speed/skill

etc...

and after THAT people come up with the idea to map even more actual numbers to "words" leaves me confused.

meh, my two cents - though i guess my view may be a little off :s

[greepish]

while one could argue that the early game should be allowed to be alot harder and random - since you didn't spend alot of time in it - i think that is quite a bogus argument, since ideally you'd want a linear progression of both the character and the game difficulty

Sure, ideally you'd want the game to get harder, not easier. However, in crawl this cannot be the case because it's just too darned long. If you've ever been frustrated that you've spent 5 hours on a character only to have him die because of a stupid mistake, multiply that frustration by ten if the game got harder as you went. However, in the beginning of the game, it's really not frustrating at all if you splat 5-100 times since you do it all very quickly. This is really just how long permadeath games have to work.

TBH, I think I'd enjoy crawl more if it were shorter. 1-rune tournaments would be pretty fun

[nicolae]

i feel like lair and somewhat later branches are alot more reasonable to react to (besides getting your ass kicked by a unique cluster fuck) since you had time to gather some tools to handle some situations.

The fact that you have no tools is the reason why the early game is so difficult. Would you rather have a bunch of good scrolls and potions and items just all over D:1?

(not trying to get off topic here... but a decent ammount of information does kinda change the difficulty of the game)
additionally, i'd like some better (clearer) ingame descriptions - like how every game i HAVE to have access to the knowledge bot (even if it's only for precise monster descriptions) - which kinda seems wrong to me, since it makes the game quite impossible to play unspoiled.

There's a project undertaken by some of the players down in the Contributions forum to help improve the text descriptions, go have a look. Some of them are already in. One of the points of the game philosophy is that you should be able to play unspoiled just from the help descriptions. A general impression of the monster's toughness, abilities, and hazards is enough: you don't actually need to know a monster's specific number of HD or HP or MR -- or, at least, you shouldn't, which is why some descriptions are being updated -- in order to know whether to fight or flee.

and after THAT people come up with the idea to map even more actual numbers to "words" leaves me confused.

The devs don't include numbers for magic resistance or stealth because the numbers would be utterly meaningless to the player. What does it mean to have a stealth or MR of 100? It's not a percentage, it's just a variable with no obvious meaning. Numbers can turn out to be misleading, too: a +0, +1 weapon with base damage 6 and accuracy 5 is not equivalent to a +1, +0 weapon with base damage 7 and accuracy 4.

[absolutego]

i've only had about 3-4 decent runs for a 3 rune win (v0.9.1 - so take my comments with a grain of salt) and i must say that it bugs me how, pre-lair is based on pure luck and somewhat random decision making (early temple/altar, rpois, curare kobolds, ...).

this is just not true, and has been discussed many times before. here is the last one:
viewtopic.php?f=8&t=2467

there is luck involved (less than before: more hp, no early high-tier wands), and you won't be able to win 100% of your games even if you play optimally, but i understand that's not a design goal, because it'd probably make for a duller game in the long run.

[dondy]

i feel like lair and somewhat later branches are alot more reasonable to react to (besides getting your ass kicked by a unique cluster fuck) since you had time to gather some tools to handle some situations.

The fact that you have no tools is the reason why the early game is so difficult. Would you rather have a bunch of good scrolls and potions and items just all over D:1?

(not trying to get off topic here... but a decent ammount of information does kinda change the difficulty of the game)
additionally, i'd like some better (clearer) ingame descriptions - like how every game i HAVE to have access to the knowledge bot (even if it's only for precise monster descriptions) - which kinda seems wrong to me, since it makes the game quite impossible to play unspoiled.

There's a project undertaken by some of the players down in the Contributions forum to help improve the text descriptions, go have a look. Some of them are already in. One of the points of the game philosophy is that you should be able to play unspoiled just from the help descriptions. A general impression of the monster's toughness, abilities, and hazards is enough: you don't actually need to know a monster's specific number of HD or HP or MR -- or, at least, you shouldn't, which is why some descriptions are being updated -- in order to know whether to fight or flee.

and after THAT people come up with the idea to map even more actual numbers to "words" leaves me confused.

The devs don't include numbers for magic resistance or stealth because the numbers would be utterly meaningless to the player. What does it mean to have a stealth or MR of 100? It's not a percentage, it's just a variable with no obvious meaning. Numbers can turn out to be misleading, too: a +0, +1 weapon with base damage 6 and accuracy 5 is not equivalent to a +1, +0 weapon with base damage 7 and accuracy 4.

--- first point

and well, ofc i don't want to start out with a complete "win-game-set", but (i didn't generally check the code, but) isn't it possible to make sure to not spawn 2-3 non-sleeping orc priests on a entrance to d2, f.e. or what sicko thought up grinder ... *grmbl* hahaha - having these HUGE difficulty bumps in the early game when it's not even clear that you'll get a basic set of equipment seems kinda weird to me... (similarly when the RNG just gives you the finger - like when you pretty much having a decent chance to do something yet the RNG sez no and simply kills you)

--- second point

numbers may be meaningless (without proper context), but they are in fact harder to missunderstand (depending on the size)...
same with words, if you'd abstract both from symbols - yet words are easier to missunderstand than numbers when in proper context (because spoken language is not as exact)

so people who'd try to play somewhat optimally would obviously always prefer to get the numbers rather than some textualized approximation of a actual number (i guess this game is actually about playing optimally rather than playing sloppy and still beeing able to win).

the problem is, you can't reasonably expect somebody to create a textual representation that looks decent, isn't easy to missunderstand and is somewhat as finegrained as the actual numbers.

percentages would be helpfull though, see "brogue" for example - they give you a indication of how tough a monster is by the chance you'd survive a melee confrontation (not sure if they have resistances... i didn't bother too much with it).

meh, TL;DR i like numbers, good numbers that is - bad ones suck :p

[dpeg]

dondy: I told you several times on IRC to read that section of the manual. It explains what OOD monsters are good for.

Regarding numbers: No, it is not okay that you take your approach (optimal play -- although it is unclear to me how optimal that actually is, your comments seem somewhat misguided) and extrapolate it on the rest. There is absolutely no need to look up monster entries in the database. I never do this and that's definitely not because I know them from heart. The game is winnable without that, and we mean it.

[nicolae]

so people who'd try to play somewhat optimally would obviously always prefer to get the numbers rather than some textualized approximation of a actual number (i guess this game is actually about playing optimally rather than playing sloppy and still beeing able to win).

If you need to crunch your own numbers to play optimally then something has gone terribly wrong. This isn't Nethack. A player should be able to evaluate which spell is more worth learning, for instance, without having to look up tables and charts. If a spell has some side effect that isn't obvious from reading its description, or if it's presented as being better than it really is, then either the spell or the description needs to be fixed, and that's usually what's done.

Similarly, a player should be able to look at a monster's description and get a good idea of what tactics will work on it. Simply being told that some monster throws fire should be enough of an incentive to wear rF equipment as being told that a monster does XdY fire damage. Even when I do look up something on the database bots, I don't do much more math on the numbers beyond "I can deal with that now" or "aw hell no".

Situations where some mechanic turns out to be non-intuitive without spoilers are usually situations that the devs are eager to change. For instance, one of the main reasons, if not the only reason, why they want to redo how weapons skills work is because most new players don't know there's only negligible benefit to weapons skills after you get minimum attack delay.

percentages would be helpfull though, see "brogue" for example - they give you a indication of how tough a monster is by the chance you'd survive a melee confrontation (not sure if they have resistances... i didn't bother too much with it).

Okay, I'll bite: How the hell would they manage to calculate the chance of surviving a melee confrontation with any degree of accuracy?

[ElectricAlbatross]

I'm sure the calculation is simple enough if it's assumed that both the monster and the character swing at the other each turn. That might be a pretty big if, though I haven't played the game enough to know.

[dpeg]

The output is also not very useful, a monster's HP ranges between X and 2X, for example.

[dondy]

well the description is like:
yadda yadda, thematical description

monster Foo can kill you in n hits with a chance of p.
you could kill the monster Foo in m hits with a chance of q.

yadda yadda, resistances

that does give you a decent idea on how tough a monster is, esp. if you keep watching monsters.
(even though it doesn't take abilities into consideration)

--- bbiab, then i'll properly reply D:
--- EDIT: attempt at a more elaborate (proper) reply

dondy: I told you several times on IRC to read that section of the manual. It explains what OOD monsters are good for.

Regarding numbers: No, it is not okay that you take your approach (optimal play -- although it is unclear to me how optimal that actually is, your comments seem somewhat misguided) and extrapolate it on the rest. There is absolutely no need to look up monster entries in the database. I never do this and that's definitely not because I know them from heart. The game is winnable without that, and we mean it.

OOD Monsters: meh, it's ok - i just thought that stuff like that creates the difficulty bumps that makes the early stages of the game impossibly hard at times and thus really hard to balance; the generator could possibly be "optimized" (as in 2 uniques never spawn in LOS of eachother) to generate somewhat friendlier dungeon layouts, i think at least the idea could be taken into consideration - dunnoh why it's so fundamentally wrong (okok, i will read the manual again! :p) :[

so people who'd try to play somewhat optimally would obviously always prefer to get the numbers rather than some textualized approximation of a actual number (i guess this game is actually about playing optimally rather than playing sloppy and still beeing able to win).

If you need to crunch your own numbers to play optimally then something has gone terribly wrong. This isn't Nethack. A player should be able to evaluate which spell is more worth learning, for instance, without having to look up tables and charts. If a spell has some side effect that isn't obvious from reading its description, or if it's presented as being better than it really is, then either the spell or the description needs to be fixed, and that's usually what's done.

Similarly, a player should be able to look at a monster's description and get a good idea of what tactics will work on it. Simply being told that some monster throws fire should be enough of an incentive to wear rF equipment as being told that a monster does XdY fire damage. Even when I do look up something on the database bots, I don't do much more math on the numbers beyond "I can deal with that now" or "aw hell no".

Numbers: dpeg, i find taking yourself (and pretty much anyone on the dev team or who's on the thingy-players list) as example for a player that doesn't look up monster descriptions on the knowledge bot kinda meh as "point statement". additionally it's not clear to me, what seems misguided about my statements D:

since you lurk around on irc - you could take alot of (statistical) information about lacking documentation or bad descriptions on the number of lookups on the knowledge bots - interpreting them as indication d: (easy to overview on a "tag cloud" type of view), sounds like a nice idea?

on the other hand i agree with nicolae - actively crunching numbers isn't necessary (and way too tedious for a game, you could simply write a bot to play for you then, which would simply ruin the game for you - YAY) - yet you somewhat do it when simply looking it up, you make a approximation if you can defeat the monster or not - but i'm greatly opposed to create approximations of approximations (as is needed when checking in-game monster descriptions), unless you already have a "idea" of the "toughness" of a monster - which is what i'd say everyone has for most monsters he encountered a bunch of times (over the course of a bunch of games, etc. etc.)

on another note, you probably wouldn't only get one pip of fire res if something has a chance at doing more than 50% max hp fire damage... that's where numbers DO help (and where i like the nDm descriptions) :p

which is also a nice example for how easy to misinterpret words - what is "a bunch of times"? is it more than alot of times or is it not quite as many times? etc.


woah, so much text - i hope i could clarify my point and didn't leave anyone out

cheers, dondy