But seriously, remove hunger

[archaeo]

edited to add: Please see this spreadsheet for an easier-to-read/edit version of my proposal.

I make a lot of jokes about removing food, but it's something I seriously think is a good idea, which is why I ended up writing 1,700+ words on the topic, like a crazy person. On the plus side, I promise that I'll let it go after this thread.

From a player's perspective, food causes inventory annoyances and requires a few hundred presses of c and e per game, and that's when you're doing it right; if you've gotten absurdly bad RNG or you're playing sloppily, on the other hand, you get to enjoy one of the slowest, most obnoxious deaths Crawl has available. From the developers' perspective, hunger is wrapped up in so many mechanics and species that it's basically impossible to make the hunger clock matter for decent players without seriously impacting hunger-intensive options and species. So, instead, I think we should just remove it.

Here's a list (that I've collapsed because it's hilariously long) of things that are currently designed around hunger and my thoughts on how they might work in foodless Crawl:

Spoiler: show

"hunger clock"
For the most part, this needs no correction; Crawl really doesn't need a clock, though hunger is more of a very lenient parking meter. There would still be piety decay to keep players moving, at the very least.

If scumming is a serious concern, I have a few suggestions. First, and most obviously, anything that currently can benefit a player who just stands still, such as Xom gifts, is moved to an XP timer or removed. Next, all monsters that spawn after a level is generated provide no XP, using whatever flavor justification works best. I prefer a system that ends these spawns when you kill the last XP-bearing enemy. When you leave a level and monsters are in LOS of the stairs, all of those monsters will patrol that staircase as long as you are off the level.

No further changes are necessary for Pan or the Abyss; players who want to scum there already can, and it's miserable, and they should be left to their devices.

vampirism
on equip, the message says "The [weapon] thirsts for blood," and the vampiric effect kicks in fairly slowly as you kill enemies ("The blood of the [enemy] coats the [weapon]"). When you unwield, "The aura of blood disappears from the [weapon]," and you start from 0.

Trog & berserk
Making berserk cause hunger isn't really a meaningful limit on the number of times you 'zerk, as there's more food than enemies that require it unless you're a Spriggan or terrible at Crawl. I'd suggest replacing it with strength draining, especially now that stat drain is on an XP timer. Trog reduces the chance of str drain with higher piety, but it's still a good chance, and you can't berserk if str falls below some threshold (half str?). This can be a really modest str drain; the goal isn't to kill players via stat zero, but to keep players from 'zerking in every encounter.

starvation penalty
The wiki says it exists, but I doubt it ever, ever matters, so just remove it.

regeneration
Regen increases hunger, but right now that doesn't matter much unless you're wasting a whole lot of time while having less than full HP. I'm inclined to just let this one go. If there needs to be a better limiting factor to regen, it should be covered by the effect itself, and not the hunger cost it levies.

notable species (Sp, Vp, Gh, Mu)
Sp
I think no change is really necessary; if Sp would be overpowered in foodless Crawl, they're overpowered right now, and if a nerf is needed, the food conduct isn't really the place. You could, of course, retain fruit for Fedhas and turn Sp into a species that will die without consuming fruit, and then tune that "hunger" level so that the game generates enough fruit for Fedhas & Sp, but that's only if you really want to maintain Sp's status as the "hunger species."

Vp
I mean, remove Vp, but if you can't, then simply leave bottling blood alone until someone comes up with a better idea for Vp. The hunger clock on Vp can be calibrated to whatever Vp players think is a good idea, and just ticks down over time. You could always make an (a)bility like "Bite," which lets Vp control their aliveness without "hunger," too.

Gh
Just make it Makhleb: The Species, honestly. If it kills things that would give it chunks now, it gets a shot of healing; if it goes too long without killing, it begins to rot away. Again, the rot timer should be balanced around whatever works for Gh.

Mu
Mummies really only get screwed on hungerless channeling and casting, which are two major advantages it doesn't get in foodless crawl. If it's thought that they need a buff, they could have some kind of +wiz or something? "Fast channeling"? Whatever, they're mummies and they're supposed to be bad, and most of these changes only increases the degree to which they're relatively bad.

spell hunger
This is probably one of the trickiest things to replace, since I think a substantial number of people feel that this is a meaningful limitation. I think they're often mistaken -- it takes bad play, a bad combo, or terrible RNG to run into a hunger wall with casting -- but it also seems as though people think this is a good limitation.

First, I'd advocate for simply not having spell hunger anymore. MP is a meaningful limitation for the number of spells you can cast at any one time; if it's felt that players can cast too many spells before resting, reduce max MP instead. If the concern is that players are using high-level spells too early, increase the XP required for higher spell levels, or require that players achieve a certain XL to cast spells of a certain level.

However, if people really want a non-MP/XL limit on spellcasting, I would suggest int drain, in a similar fashion to the way berserk causes str drain above. As your spellcasting skill and intelligence gets higher, the chance you get int drained is lower. A staff of energy prevents int drain, while Necromutation moves down to level 6 and has no effect on this draining.

channeling
The hunger cost of channeling is virtually meaningless. Strategically, there's more than enough food for any caster to channel regularly for the whole game. Tactically, I find that the limited MP you get from channeling is a lot more impactful than the hunger cost, especially since it's always easy to just scarf down a fast food item.

If channeling doesn't seem overpowered to you right now, it probably doesn't need a replacement limitation in foodless Crawl. If it's overpowered, though, I'd suggest making each channel take longer than a single turn, with high invo/evo levels required to bring it down to 10 aut.

hungry ghosts
Can just go away, thanks.

(a)bilities
No abilities really need to cause hunger, as it's not currently a serious limiting factor. If people are worried about players spamming +blink or =tele, you can increase the contam of either. If +Inv spamming is a problem, turn it into an evocable item that goes inert instead.

rods and special staves
Rod hunger (and staff of Dispater hunger, lol) is pointless, and needs no replacement. Rods are fine; I guess evocable hellfire might need to do more HP damage.

Fedhas
If people like the idea of Spriggan as a "hunger species," you could always keep fruit in the game for them and then retain Fedhas' mechanics. I've also heard people suggest that, when you pray over corpses with Fedhas, you occasionally get fruit instead of mushrooms.

I suspect it would be simpler to just make Fedhas run off piety entirely.

mutagenic chunks
I really don't think there needs to be a replacement for this. If people want to play mutation roulette, they can quaff mut pots. If it absolutely can't be lived without, anything that previously provided mutagenic corpses create a mutagenic cloud over their corpse for a few turns after death (with clouds changed to give random mutations, if they currently lean toward bad ones, I guess?). But that's pretty dumb, so, sorry mut roulette lovers.

hides
Dragons/trolls can simply drop these instead of requiring butchering. Easy!

other gods
I'm pretty sure that it's basically just Trog and Fedhas that have to be seriously looked at for foodless Crawl. Ely already basically works off piety instead of hunger, and piety costs can be adjusted for her abilities if removing hunger makes them too spammable. I guess this is also a Zin buff, since now his food conduct doesn't matter, but it's pretty minor. All other abilities that only have MP and hunger costs should probably just have higher MP costs if they're especially spammable.

To put all that another way: if you've played Trunk and you are moderately good at this game, you've basically played foodless Crawl. It just doesn't feel like it, because nearly every action you take in the game has some trivial hunger cost associated with it that nevertheless merits interrupting play.

Most proposals I've read, including the devs' experiment with chunkless, just seek to solve the player-side problems of food. This would be better than what we've got, but it retains all of the mechanics that are difficult to balance without impacting something else. Likewise, I think calls for a tighter hunger clock fail to wrestle with just how many mechanics are tied into hunger; how do you make a clock that affects, say, a MiFi of Oka that never stashes without completely decimating average play with ranged or casting combos? Plus, and this is just my opinion, running out of food is a lame and boring way to kill players; if a clock is desired, there are more interesting things that could be put in place.

All that said, this proposal calls for massive changes, and I'm a well-known idiot who can't write code to save his life. I also totally recognize and respect the fact that food and hunger are quintessential roguelike mechanics, and that several people have good reasons to disagree with going foodless. But I still think it would lead to a better game, overall.

[greedo]

One problem is exactly what you say: food is very intertwined with crawl. From carnivores, to undead, to spell hunger, to berserking, removing it would change so much. It might be more trouble than it's worth to get rid of it. I think it'd be better to just abstract the part that the players deal with.

E.g.
When you pick up food, you just add it's nutrition worth to a stockpile number. When below full, you gradually take extra food to go up to full, otherwise you just slowly lose stockpile and hunger doesn't decrease. You only go above full by eating corpses whole which only carnivores can do (and just take out chunks altogether). Stockpile reaching 0 just means you stop recovering hunger and eventually starve.

^edited a bunch xD

[archaeo]

First off, I totally acknowledge my proposal is "more trouble than it's worth," probably. I still think it's a good idea, but I'm not the one who would have to actually do all that work, and I'm content with just spinning my wheels on the subject. Food isn't that bad.

Your "food 'stockpile'" idea fits right in with a number of other proposals that have "abstracted" or automated food in the name of removing player annoyances. It's not a bad direction to go, but I have a couple of problems with it.

First, you can already accomplish most of the goals of these proposals in Crawl right now. A few .rc file settings and some lua magic will have the entire butchering and eating process automated. If you can code a bot that can win the entire game, you can code one that will eat when you're resting or autoexploring.

More importantly, though, I feel that abstracting/automating food would mean retaining all the bad design of hunger while throwing out the only good part, the flavor. Sure, players wouldn't have to press c or e as much, but the game would still be saddled with a mechanic that only very rarely impacts play despite its pervasiveness. The only thing that redeems food as a design concept is the fact that it's iconic and immersive. Hunger clocks are standard in roguelikes, and in my opinion, it has a lot to do with roguelike players appreciating the "realism" of eating and hunger, however tenuous it may be. If we're going to get rid of food, and break with a roguelike tradition, we might as well actually get rid of it, instead of hiding it behind abstractions.

[duvessa]

It isn't very intertwined though. Carnivore, spell hunger, and undead food systems don't have any real balance implications, spell hunger's supposed tactical purpose is the same as MP's, and berserk has Exh. The important purpose of food is serving as a clock, except that chunks, carnivore, spell hunger, and undead food systems all ensure that it does a really bad job being a clock; you can still farm lair forever with Mu/Vp. Piety decay exists to at least try to stop most characters from mummystabbing or farming, which is nice, but unless it actually kills you (which it doesn't) it's not going to do the full job of being a clock.
Food is used as a clock in so many games because it provides a convenient way to have a clock (nutrition) that kills you when time runs out (starvation) that you stall by progressing in the game (finding food) while giving the appearance of being simulationist. Sometimes game designers* screw it up spectacularly, of course, like in NetHack where dying from eating too much is easier than starving, or in the the hit indie title Dungeon Master But Darker--er, sorry, I mean Legend of Grimrock-- where casting fireball spells at thin air gives you infinite nutrition. All the goofiness of Crawl's food system can probably be traced back to trying to imitate other roguelikes, when it should have just imitated Rogue and made it a simple clock.
I don't think you appreciate how difficult it is to do away with a clock entirely in almost any single-player game where you want optimal play to be fun. ToME4 tried it, and it made each level take forever when playing well (admittedly everything takes forever in that game no matter what, but still). DoomRL tried it, and most of the actions in survival-oriented DoomRL runs are the 'wait' command. Play one of the levels in Wesnoth with no turn limit and you can win even as an awful player just by stalling the AI for ages. Tetris wouldn't be such a great game if the blocks didn't fall down on their own.

*using the term loosely here

[mps]

Indeed, remove hunger. I don't see this being a big deal. Remove food by just not generating it and holding player nutrition constant (if this is actually a hard thing to make happen in crawl... well...). Maybe in trunk berder's "blaster casters" will be moderately more powerful until the next round of magic nerfs come around. Not a big deal at all. Characters that don't spam level 6 spells constantly, which is to say most good ones, will see little actual difference.

As for various ramifications mentioned by archaeo in his "all work and no play" list, only ghoul and Vp seem important at all. They could be special cased to maintain their interactions with corpses/chunks/blood, simplified to interact only with corpses (which would presumably be kept for their necromancy interactions) and adjusted to make lack of storage workable, or in the case of vampire removed entirely.

All that said, if devs are looking for a punt that would satisfy some food critics, I would recommend making chunks give more nutrition so that players can ignore food more thoroughly in reasonable play. The problem with hunger is not merely that it has no real gameplay impact. This is also true of other roguelikes I've played a fair bit, e.g. angband. The real problem is the nagging interface issues, where the player feels obliged to constantly stuff the character's face with chunks of monsters once or twice per minute of play. Limiting the player's ability to eat and derive nutrition from chunks does not meaningfully improve the gameplay impact of hunger (which, again, is essentially zero). It just makes hunger a far more irritating non-issue.

[Berder]

Here's a way to remove hunger while keeping the flavor. It's ridiculous to have characters almost starve to death while casting five spells. Normal people can comfortably go hours between meals and it takes them weeks to starve. Just adjust the hunger clock so that all costs are 1/10 of what they are now. Three meat rations and you're set for the game.

[archaeo]

It isn't very intertwined though. Carnivore, spell hunger, and undead food systems don't have any real balance implications, spell hunger's supposed tactical purpose is the same as MP's, and berserk has Exh.

I agree, that's true if you look at the game objectively. But, in talking to people and reading the Tavern, I think it's clear that many people do think these costs are significant, including some devs, and the game acts like they're significant, with those brightly colored status lights. I see people do things like rush to learn necromut or wear (and recommend!) gourmand despite having tall stacks of permafood all the time.

imo, games shouldn't fake the player into thinking there's a worrying time limit when one doesn't actually exist, unless it's a clever plot twist or something. If Crawl needs actual limits for these mechanics, there are far better ones than the overburdened hunger system.

I don't think you appreciate how difficult it is to do away with a clock entirely in almost any single-player game where you want optimal play to be fun.

I mean, my proposal doesn't really do away with all clocks; it retains an OOD spawn timer, it just removes the XP from the spawned enemies. Probably easily overlooked in that dumb wall of text I wrote.

But that's more about stopping people from scumming than really turning the screws with a simple timer. If we really need a timer, hunger would be the last way I'd do it, honestly. Using metabolism as a clock makes the player constantly have to mind the meter, and while the simulationism is a nice touch, it doesn't play very well. Instead, I've always admired the clocks in games like ADoM or Majora's Mask, where the timer is integrated into the setting.

All that said, if devs are looking for a punt that would satisfy some food critics, I would recommend making chunks give more nutrition so that players can ignore food more thoroughly in reasonable play.

Here's a way to remove hunger while keeping the flavor. It's ridiculous to have characters almost starve to death while casting five spells. Normal people can comfortably go hours between meals and it takes them weeks to starve. Just adjust the hunger clock so that all costs are 1/10 of what they are now. Three meat rations and you're set for the game.

I'm confused as to how these proposals wouldn't just encourage Mu/Vp-style scumming for all species. I assume you guys would pair it with monsters that gave 0 XP if they spawned after the level was first generated?

Regardless, I think the whole "shrink the hunger clock until it's small enough to drown in a bathtub" idea isn't very satisfying. From a code standpoint, I imagine these ideas would be easier to implement, since you're just changing some numbers, but as far as gameplay goes, if you minimize hunger to this degree, you may as well just remove it altogether.

[mps]

I don't think any resolution to the issues with hunger is likely to be satisfying to the sensibilities shared by you and I.

In my opinion, there's no scumming for xp in crawl without the monster spawns that supposedly prevent you from scumming for xp -- the solution is to remove the spawns.

We've been at a stage where you might as well remove the "hunger clock" (probably shouldn't actually use this language, since it concedes that hunger puts meaningful check on reasonable play) for a while now. It doesn't happen, even though you'd think given the amount of time you spend mashing buttons to eat in a typical game of crawl, it would be a real priority for someone at least. Take a look at another roguelike: angband. You see exactly the "3 meat rations and you're set" Berder suggests.. yet food's still in the game! I think the situation here is similar. Food hangs around for reasons of tradition, simulationism, and false belief in computer roleplaying. Food comes down directly from dungeons and dragons, along with identification and curses. People question whether the game would still be a roguelike without these things, even if their effects are primarily cosmetic in practice.

[archaeo]

We've talked about removing the spawn timer before, but I do see how it could be useful. One of the issues with hunger-free/diet Crawl is the problem duvessa brings up above and in that thread, where players just stand around waiting for enemies to come to them, or sit upstairs and wait for them to go someplace else. I like 0 XP monsters that stop spawning when you kill all/most of the initial enemies and some kind of hack to keep monsters alert and near the staircase if they were in LOS when you went up/down stairs. You could probably improve on that idea, but it'd prevent the worst scumming in foodless Crawl, I think.

And sure, food is mostly kept because it's a roguelike tradition, but that does matter to some people, and I'm cool with that. I know I just wrote an "all work and no play" list (lol) arguing to remove hunger, but this isn't the first time reforming food has come up, and it's never gone anywhere; dpeg's chunkless idea is the only one that's ever been supported, but I have a hard time seeing how it really fixes the issues people bring up. I don't think anybody's really happy about how food works in Crawl right now, but I'll be satisfied with any solution that actually works, even if it keeps food. For that matter, if everybody else thinks it's no big deal, and nothing changes, I still think current Trunk is a fun version of the game, and the whole butcher-and-eat routine is all muscle memory at this point anyway.

[mps]

Right, I think a lot of these things are tied together into a single ball of wax. I suspect that the fear that players will hang around doing nothing waiting for monsters to wander to them would be more usefully addressed by ... monsters not wandering to an idling character. The whole idea of branding this kind of cheezy tactic "scumming" was misguided in the first place, as was the idea of "fixing it" by introducing stronger spawns that would somehow kill the player wherever they've set up to kill wandering monsters. It's just a monster behavior exploit that could be eliminated by improving monster behavior. Generally speaking, if monsters are doing something stupid that gets them killed trivially, the solution should involve them not doing that anymore.

re: roguelike tradition vs. food removal, I can't address this topic without getting into a long rant about the ridiculousness of current ideas about what makes a roguelike, except to say that the idea that the details of the roleplaying rules implemented in a roguelikes, e.g. "do we care about food?", is part of some specification or definition is a pretty big part of the staleness in the genre for the past 15 years.

[bel]

There are very good reasons to remove food, and OP has given a relatively easy fix to any issues that arise. I don't really see any problems with it. Taking one point at random, I do think that hunger is relevant to spellcasting right now, but it is relevant in an hacky, tedious, and non-intuitive way (why on Earth does spellcasting cause hunger? I remember encountering this in nethack and thought it was weird then). I would be totally happy to remove hunger from spellcasting: it does not lead to any balance issues of any import.

When I first played crawl, the chunk mechanic seemed quite ingenious to me, and I thought that it would be very tactically relevant. Now, however, I think it does not fulfil its purpose.

I think food exists because of historic reasons, though the tedium is not so much that it is a high priority.

Spoiler: show

Also, how come you eat the whole dungeon and never need to go to the toilet?

[ontoclasm]

Removing hunger altogether would be fine, but I'd prefer to make it an actual clock instead of the absolute 100% non-issue it is now. How on earth can hunger stop you scumming when scumming literally generates food on its own? The chunkless branch removes chunks... And then has monsters drop rations, completely undermining the point.

There should be permafood, in small and strictly measured amounts - - enough to keep you alive if you progress slowly, and not one bit more - - and that's all the food you get ever. Stuff like spell hunger and channeling hunger should get removed, so that a ration provides the same amount of time to everyone. This is how Brogue does it and it's great (Rogue too, I guess).

[tabstorm]

If you get rid of the various hunger interactions that do almost nothing (Berserk, ability hunger, rods, regeneration, spell hunger etc.) and get rid of the food conducts of undead races (Gh no longer rots, Vp has the same resists as Gh but gets batform, if this is not "distinct" enough just remove one (I vote Vp)) and you can basically have a food system that serves only as a clock. All food would just be rations. This would keep the player pushing forward to some extent while eliminating the need for things like OOD respawns.

The only time I've ever actually noticed food affecting my character in a meaningful way has been Su of Sif with no melee, because it is probably the most food-intensive character you can play. It was pretty unbearable (probably because you are hungry so often, and the game is quite slow)

With that said, if you got rid of hunger for everything but channeling, spells, and Berserking you could have a "quite meaningful" food minigame to inflict on spellcasters and Berserkers (to a lesser extent), if you think they need to be regulated further.

[lessens]

I don't really have a strong preference for either hungerless or the current system but the one system I don't want to see is something like the current chunkless, mainly because of spell hunger but also the other hungering abilities. A clock is fine but it's ridiculous in my view to punish players for casting spells.

[duvessa]

the way brogue does hunger is actually awful because of how the item generation works (ever heard of the mango chart) but the basic idea is right, yeah

[Sprucery]

Instead, I've always admired the clocks in games like ADoM or Majora's Mask, where the timer is integrated into the setting.

In ADoM, though, you could still completely ignore the clock (as I always did). There are enough guaranteed scrolls of chaos resistance in the game.

There should be permafood, in small and strictly measured amounts - - enough to keep you alive if you progress slowly, and not one bit more - - and that's all the food you get ever. Stuff like spell hunger and channeling hunger should get removed, so that a ration provides the same amount of time to everyone.

Good enough, if that can be balanced. Different character combinations take different amount of time to advance (at least for me). For how many turns should a 3-rune game have food for? I am afraid that all my characters will then starve (I seem to play extremely slowly compared to most others).

[archaeo]

Yeah, to echo Sprucery, how low is the time limit in permafood-only Crawl? Without all the other hunger costs, how much time should the player get before starving in the 3-5 rune game? 100,000 turns? 120,000? 80,000? Will players doing extended have to spend time gathering food in Pan to do Hell/Tomb?

It seems like you'd either end up with a hunger clock that remains a mere nuisance to all but the slowest players, or you'd end up tightening the restriction enough to starve new players fairly often, neither of which seems great.

[greedo]

I can definitely agree with the above. The disparity in turn count between offensive spellcasters and meleers with a decent armor is ridiculous. There's no way to make a reliable clock in crawl without something like increasing mana regen with consecutive resting. Personally I find the given clock, sheer boredom after 3 hours or so, to be the prime motivator -.- By the time I get to zot and have like a zillion reliable panic buttons I just dive to 5 even though it makes ascension harder.

[Jeremiah]

Removing hunger altogether would be fine, but I'd prefer to make it an actual clock instead of the absolute 100% non-issue it is now. How on earth can hunger stop you scumming when scumming literally generates food on its own? The chunkless branch removes chunks... And then has monsters drop rations, completely undermining the point.

There should be permafood, in small and strictly measured amounts - - enough to keep you alive if you progress slowly, and not one bit more - - and that's all the food you get ever. Stuff like spell hunger and channeling hunger should get removed, so that a ration provides the same amount of time to everyone. This is how Brogue does it and it's great (Rogue too, I guess).

I'd be interested in playing something like this, but if the amount of food was strictly measured rather than random, is there any reason why it would need to exist as an item at all? Entering a new level could just add a certain amount of nutrition, then if you found yourself running low you may have to decide to take the next staircase down before fully explorig the current level.

Maybe then "expends nutrition to cast spells/use abilities" could be a bad mutation or species gimmick (Spriggan?)

[Bodrick]

You could tie the clock in with rune collection - "The dungeon will collapse / world will end in 20000 turns!" and "Collecting the slimy rune of zot has stabilized the dungeon for 5000 turns" etc. This would be fairly easy to balance, doesn't depend on the swingyness of item generation, and could lead to interesting cases where you have to try and dive for a rune quickly before you run out of time.

[lessens]

If you have a system like that it should be a fairly liberal clock, otherwise you're pushing towards the territory where you're penalizing players for pressing 'o' and '5' too much. The scoring system already does that and, critically, getting a low score doesn't kill you.

[Sprucery]

You could tie the clock in with rune collection - "The dungeon will collapse / world will end in 20000 turns!" and "Collecting the slimy rune of zot has stabilized the dungeon for 5000 turns" etc. This would be fairly easy to balance, doesn't depend on the swingyness of item generation, and could lead to interesting cases where you have to try and dive for a rune quickly before you run out of time.

I can't see how this would be easy to balance. Either it would be meaningless for some characters and players, or game-ending for others, or both.

I wouldn't like hard (or 'challenging') combos be made harder by a time limit.

[archaeo]

Regardless of what kind of clock you use, hunger or otherwise, you'll either end up keeping the broad lenience of the status quo, or you make it strict enough that it'll significantly change how lots of combos have to play the game. There's no real middle ground or "balance" to be found.

Personally, I think the only timer that really works in the context of Crawl is a spawn timer. Our current timer is very bad, especially in Lair, but I think you could reform monster spawning to be less annoying and more meaningful. I don't think the same is true of hunger clocks, and I'm skeptical if there are other "global" timers that could be made to work either.

[TeshiAlair]

Two paradoxical statements-

1. Having a clock is stupid. Speed/score is the reward for playing faster, playing at your own pace should never be punished.

2. I like the hunger clock not for long term impact (which it doesn't have) but short term tactical impact for people like me who are dumb and spam spells and berserk and then get hungry a lot. I have had to use food stores before, and that's really neat to me.


Solution- Going along with what some people have said of alternative costs (contam, statdrain, exhaustion), make a consumable that either cures these or temporarily removes the costs. Something like Potion of Vigor- you aren't exhausted/your stats don't drain after berserking/casting spells, at the cost of being a limited resource and the penalty being more meaningful when you don't use it.

[mps]

If you want to have a clock, please make it a clock, not a wind-up watch. Players should not be responsible for interactively helping the game with strictly time-related bookkeeping (e.g. stuffing your face with food carefully measured by the game to give you a certain amount of gametime). This is purely tedious and the only way to make it less so is to require less frequent eating -- obviously, then, the solution is never having to eat and being "timed" by some other passive factor.

The situation right now is a watch you have to wind frequently if you sensibly resist the temptation to eat chunks constantly and just eat your permafood and constantly if you can't bring yourself to "waste" an apparently (but not actually) limited resource and which in practice provides no actual limitations.

A good timer, imo, would look like some kind of map decay where earlier parts of the dungeon progressively turn into something like pandemonium or the abyss along with insane spawns that normal characters cannot take on until the endgame. If this kills early shops and altars... good!

[Sandman25]

Spawned monsters can leave corpses but give no XP. This way slow players like me will not starve but will not get any extra XP.

[lessens]

Spawned monsters can leave corpses but give no XP. This way slow players like me will not starve but will not get any extra XP.

That sounds annoying and frustrating. As you struggle with the game you're given increasing challenges that give you no reward.

[archaeo]

Sandman25, I'm pretty sure that any serious food reform would just remove chunks entirely.

lessens, the impression I'm getting is that people think Crawl either needs a timer or it needs spawning monsters, otherwise players will just stand around waiting for the monsters to come to them. I don't think this is true -- I'd love an experimental branch to check it out -- but if it's a real problem, it needs to be addressed. Since the best way of fixing it, changing Crawl's monster AI, sounds like it would be hard, and timers are very difficult to balance, we probably want to try fixing monster spawning instead.

[lessens]

I don't think it's actually that hard to balance a timer in a way that prevents mummy-level scumming for XP but still lets you press 'o' and '5' freely.

[mps]

Changing monster AI to prevent the form of "scumming" monster spawns are supposedly designed to prevent would be trivial. Assuming the player isn't using some kind of automation script, deciding whether they're waiting around for monsters for a large amount of gametime is straightforward. Once the player has entered this state, you simply prevent wandering monsters from entering LOS. You could even keep them out of a slightly larger buffer zone to make it difficult for the player to defeat this measure by walking around a little bit. Since it takes time to enter this state, more than it takes to rest as much as you like, the potential for exploitation seems minimal.

[lessens]

Changing monster AI to prevent the form of "scumming" monster spawns are supposedly designed to prevent would be trivial. Assuming the player isn't using some kind of automation script, deciding whether they're waiting around for monsters for a large amount of gametime is straightforward. Once the player has entered this state, you simply prevent wandering monsters from entering LOS. You could even keep them out of a slightly larger buffer zone to make it difficult for the player to defeat this measure by walking around a little bit. Since it takes time to enter this state, more than it takes to rest as much as you like, the potential for exploitation seems minimal.

Giving the player an invincibility barrier during waits to prevent scumming is like cutting off the nose to spite the face except the face doesn't care at all because scumming has nothing to do with monsters wandering into LOS during waits.

[WalkerBoh]

I firmly believe that Crawl is a much more fun game without the hunger clock. It became especially clear to me just how tedious chopping and eating is after winning a couple mummies recently. I agree in principle with the OP regarding the issues and solutions, especially the concept of not using hunger as an additional cost for various actions. I don't have a serious opinion on the specific implementation for removing hunger - like how to handle Fedhas fruit, for example - but nothing the OP suggested seemed unreasonable to me.

However, the whole question of timers is completely unrelated to hunger, in my opinion. The overall long-term timer is not tied to hunger in any way currently; right now the only true hard limit on how long a player needs to take to win the game is the end of the world (2 billion turns?). There is more than enough food available for any reasonable amount of turns (especially if you go into Pan or Abyss) and I think making that limit much shorter is an undesirable change. Especially since it really has nothing to do with removing hunger.

Of course, hunger also doesn't really function as a short-term timer to prevent behavior like sitting on a level for indefinite amounts of time. There's more than enough food to do any reasonable amount of scumming, so eliminating hunger doesn't really necessitate any changes to this system. The current implementation, the OOD spawn timer, is a very questionable way of implementing a short-term timer in my opinion, and there's a lot of potential ways you could change or improve the system (one idea off the top of my head is to have a slow exp drain kick in after X turns without exploring new tiles). But it's only at best tangentially related to removing hunger so I don't see the need to explore that in-depth in this thread.

TL;DR: I think you could implement a lot of the OP's suggestions without having to worry about adding new timers, since hunger doesn't function like a timer in 99.9% of games currently anyways.

[archaeo]

I don't think it's actually that hard to balance a timer in a way that prevents mummy-level scumming for XP but still lets you press 'o' and '5' freely.

But then why bother having a timer at all? Why not just remove the mechanics that enable mummy-level scumming in the first place?

[lessens]

I don't think it's actually that hard to balance a timer in a way that prevents mummy-level scumming for XP but still lets you press 'o' and '5' freely.

But then why bother having a timer at all? Why not just remove the mechanics that enable mummy-level scumming in the first place?

It's more like the opposite, if you can change the way the timer works and fix the problem without ripping out game mechanics then why not do it?

[sanka]

But then why bother having a timer at all? Why not just remove the mechanics that enable mummy-level scumming in the first place?

Some people like to play the game in an optimal way to winning, but still do not like tedious actions. I think they are the main target of game design in crawl. So many design considerations try to remove the options for tedious but optimal tactics - few players would do it anyway, but it's a good feeling that you do not play suboptimally.

Note that it's perfectly possible - even for non mummies - to scum Lair for a long-long time and level up, or to wait in an ideal fighting position till the monsters wake up and come one by one. Ideally I think we should remove chunks (so food drops from monsters), and balance the amount of food to be enough for a slow player, but not for somebody who stands on the stairs or a chokepoint and wait for the whole level to come.

[mps]

Giving the player an invincibility barrier during waits to prevent scumming is like cutting off the nose to spite the face except the face doesn't care at all because scumming has nothing to do with monsters wandering into LOS during waits.

You don't seem to understand what was being discussed in the quoted comment.

Unrelatedly, I totally agree with WalkerBoh.

[Sprucery]

Imo food is not a good long-term timer, but it could be a good short-term timer. But currently there's so much permafood that it is not even a good short-term timer. So how about leaving chunks as they are (goldified preferably though) but reducing the amount of permafood? This would then require some compensation for Spriggans I guess.

[Hurkyl]

Changing monster AI to prevent the form of "scumming" monster spawns are supposedly designed to prevent would be trivial. Assuming the player isn't using some kind of automation script, deciding whether they're waiting around for monsters for a large amount of gametime is straightforward. Once the player has entered this state, you simply prevent wandering monsters from entering LOS. You could even keep them out of a slightly larger buffer zone to make it difficult for the player to defeat this measure by walking around a little bit. Since it takes time to enter this state, more than it takes to rest as much as you like, the potential for exploitation seems minimal.

Easy way to collect most runes:

Make enough noise so that everything in the vault is awake.

Retreat somewhere cleared.

Wait to trigger this exclusion.

Take one step forward. Wait to keep the exclusion active.

Take one step forward. Wait to keep the exclusion active.

Take one step forward. Wait to keep the exclusion active.

Take one step forward. Wait to keep the exclusion active.

....

Take rune.

Take one step backward. Wait to keep the exclusion active.

Take one step backward. Wait to keep the exclusion active.

..

[mps]

Yes, you would need to take player movement into account in deciding whether the player is camping for wandering monsters...

[adozu]

when i first started playing food was a quite complex and occasionally punishing thing... i mean, nausea was a thing and you actually had to go out of your way to find "good" chunks and stuff, food also seemed more scarce in perma form.

it was a real bother but at least it was a "mechanic"

now it keeps getting parts of it stripped and what's left is pretty much a vestigial husk. imho we should either make food matter by ointroducing more severe limitations or it can just go for good.

[Sandman25]

Spawned monsters can leave corpses but give no XP. This way slow players like me will not starve but will not get any extra XP.

That sounds annoying and frustrating. As you struggle with the game you're given increasing challenges that give you no reward.

It is the least of two evils IMHO. Either that or some scumming. Also it means OoD spawns will be removed, players can continue fighting XP-less goblins on D1 if they need to rest after having problems with normal quokkas.

[Sandman25]

Sandman25, I'm pretty sure that any serious food reform would just remove chunks entirely.

Even without chunks those spawns should not give XP IMHO

[acvar]

I don't think it's actually that hard to balance a timer in a way that prevents mummy-level scumming for XP but still lets you press 'o' and '5' freely.

But then why bother having a timer at all? Why not just remove the mechanics that enable mummy-level scumming in the first place?

It's more like the opposite, if you can change the way the timer works and fix the problem without ripping out game mechanics then why not do it?

But it is not a mechanic. It is a kludge added to rogue to prevent scumming. It got kludged in nethack (eating coprses) to make it less severe which made it toothless. So now it is a kludge to fix a kludge. Unfortunately this is how programming often progresses. It no longers serves a unique purpose since it is not an effective clock and only really serves to limit berserk and high level spell spamming both of which have other mechanics which can/do/should allready limit them. There is no good reason to keep it.

So how do you prevent scumming? Easy make it so monsters are not experience pinatas. The method stated in the original post would work, heck I have even promoted it in the past, but there is an easier way to accomplish the desired effect. No monsters give experience. Simply tie experience to something else in the game. Perhaps leave little bits of experience lying around the dungeon for characters to pick up. We could call them gold pieces.

P.S. I sure hope somebody gets that joke!

[bananaken]

I think a different form of OoD spawning should be devised alongside hunger removal, were it to happen:

• OoD spawns should only occur in a preventable way (something like mps' suggestion where the Pandemonium/Abyss spreads, and there's visual indicators and messages of it happening)
• These spawns should no longer give XP
• To add to the "corruption" idea, it could be like Lugonu's "corrupt" but only swapping tiles and not destroying them; around the "center" of a corruption you can have a one-way exit from Pandemonium/Abyss that will complete once you have remained in a floor for too long, and monsters would start to come out at regular intervals.

Basically the idea would be to make pressing '5' over and over obsolete. Mummies can *try* to do this currently but they're so shitty that they're almost always too weak to take on anything OoD that spawns. The only remaining obstacle then, IMO, would be Xom. If you have faith, sustain attributes and various items to amuse xom during "teddy bear" piety he will shower you with items, gold and good mutations. A decent level 12 or so melee character should be able to survive his summons, and his items will boost your survivability over time. Without hunger clock obviously every character could try and abuse him, and I think I would try it just because it sounds funny to be ready for Depths by early game.

[byrel]

Why do we care about scumming monster spawns? If someone wants to go to that much effort, risk and boredom to get slightly ahead of the XP curve, why should we care as designers. In almost all cases it isn't optimal (because OODs spawn, and can kill you if unlucky.) It's tedious and boring. Yes, some people will do it, but those people will also scum abyss, save acquirement scrolls for troves, etc. There are already lots of clearly non-optimal scummy play that new players don't always realize are suboptimal. The correct solution is education (through in-game stuff, etc.) Maybe we could have a message if you've waited 500 turns with full health, mana, and no status effects:

  Code:

Terrifying monsters can be attracted to your presence if you spend too long on one floor. Are you sure you want to keep waiting here?

Edit: And I think it would be a good idea if the monster set for OoD spawns was expanded to include abyssal monsters after a significant amount of time. It would be both flavorful, and prevent someone who can kill ogre mages and warlords from just hanging around in Orc forever.

[bananaken]

Why do we care about scumming monster spawns? If someone wants to go to that much effort, risk and boredom to get slightly ahead of the XP curve, why should we care as designers. In almost all cases it isn't optimal (because OODs spawn, and can kill you if unlucky.) It's tedious and boring. Yes, some people will do it, but those people will also scum abyss, save acquirement scrolls for troves, etc. There are already lots of clearly non-optimal scummy play that new players don't always realize are suboptimal. The correct solution is education (through in-game stuff, etc.) Maybe we could have a message if you've waited 500 turns with full health, mana, and no status effects:

  Code:

Terrifying monsters can be attracted to your presence if you spend too long on one floor. Are you sure you want to keep waiting here?

Edit: And I think it would be a good idea if the monster set for OoD spawns was expanded to include abyssal monsters after a significant amount of time. It would be both flavorful, and prevent someone who can kill ogre mages and warlords from just hanging around in Orc forever.

Scumming monster spawns may not be "optimal", but something I learned is that the average player simply doesn't care about what is optimal, they just like to grind the extra XP before proceeding. Others like to get to XL27, train every skill to max, collect every unrandart in the game through Zigs, and finally end the game. This is not optimal either, but it's fun for them. Removing the XP on OoD spawns would just be an incentive to proceed, but you could certainly keep it and "punish" the player with boredom.

If I had a DEFE with Firestorm by Lair, I would scum for monsters without a hunger clock simply because you can. It's fun to try different things, specially if it's rewarding me with XP for doing so. I don't care if it's optimal or not, I will only care about optimal play if it's required for me to stop dying in the roguelike. I wouldn't try this every time, and I would probably learn about "optimal strategies" at some point, but in my mind I'm not trying to become a Hypothetical Optimal Man, I'm just trying to have fun. Even without XP rewards on OoD spawns I might try "scumming" just to experiment with the character and see if I can survive indefinitely by X point of the game. Likewise, I would probably stop favoring GDA over plate armour at some point on a character that could be casting spells.

[njvack]

The important purpose of food is serving as a clock, except that chunks, carnivore, spell hunger, and undead food systems all ensure that it does a really bad job being a clock; you can still farm lair forever with Mu/Vp.

I don't know that I agree that hunger makes a really bad clock in Crawl. Yes, there's undead cheese; I'm not of the opinion that those exploits are terribly important. I don't always want to play a mummy.

In most games, food keeps me moving in a sort of low-key way and occasionally makes me think a little bit about skilling and spell/rod usage. It isn't the world's most elegant system, but I don't see it as horrible, either.

Put me down for "I think there are bigger fish to fry."

[archaeo]

I don't know, njvack, it seems like one of the bigger uncooked fish in the game to me. Food reform is up there with extended reform, Tomb overhauls, charms fixes, and so on. For all of these frequent topics, the players and devs seem to broadly agree there are obvious problems, but finding consensus for solutions is difficult.

But that said, sure, the system does basically work, mostly by tricking the player a bit. It's annoying, but Crawl remains a totally playable and fun game. And I doubt anyone in this thread will quit in a huff if hunger stays the same. It would just be nice to replace our inelegant and silly food clock and hunger costs with something that adds to the game instead of just, you know, being a raw fish stinking up the joint.

[kroki]

What's wrong with scumming though? The game already punishes you for it - its boring and kills your score. I can't imagine anyone scumming in crawl past his first or second win, or maybe when trying to win a mummy.

[Siegurt]

What's wrong with scumming though? The game already punishes you for it - its boring and kills your score. I can't imagine anyone scumming in crawl past his first or second win, or maybe when trying to win a mummy.

The problem is that if scumming increases your chances of winning proportional to the amount of scumming you do, then scumming becomes the best way to win, therefore it becomes "the way to play crawl" (because not scumming reduces your chances) and then crawl is a boring, easy game that no one wants to play.

In order to avoid this, scumming has to decrease your overall chance of winning, or at least be no net change to it.

The problem is that scumming and just plain old inexperienced slow play look very very similar, and it is very hard to punish one and support the other.