The Tavern

and what makes it a compelling gate on spellcasting when MP, fail rates, and spell slots are already mechanics that gate high-level spells

Because spells are the strongest player option in this game. Conjuration spells are basically ranged weapons but have way larger potential average damage per aut and have infinite ammo because MP is infinite. Hunger is currently too lenient because of how much chunks and permafood drop in one game regardless of luck

ONIchinchin wrote:Because spells are the strongest player option in this game. Conjuration spells are basically ranged weapons but have way larger potential average damage per aut and have infinite ammo because MP is infinite. Hunger is currently too lenient because of how much chunks and permafood drop in one game regardless of luck

This chart will help explain the problem in your logic:

As you can see, by the time conjurations becomes "ranged weapons but have way larger potential average damage per aut", spell hunger has long stopped mattering.

So you agree with what I've said, good to know that.

Yeah, I mean I appreciate that this is one of few RPGs that tries hard to make magic not the most OP thing ever, but spell hunger just feels redundant and irrelevant most of the time.

Carryover from Nethack probably? As far as I know, pre-DCSS crawl was mostly inspired by Nethack. So basically, tradition.

ONIchinchin wrote:Because spells are the strongest player option in this game.

actually that would be heavy armor melee.

tabstorm wrote:
ONIchinchin wrote:Because spells are the strongest player option in this game.

actually that would be heavy armor melee.

Obviously spell hunger shouldn't exist. I think it's a carryover from nethack.

But since this is CYC, I don't feel guilty about derailing the thread a bit. I am interested in this question. How would one go about showing that one playstyle is the "strongest"? What's the domain where one is arguing? (Obviously DEFi is weaker than DEFE, I think, so it can't be the whole universe).

For the sake of argument, let's assume that the player reaches D:5 or something, so the random early game isn't weighted too much. Or we could start every character with a potion of curing and scroll of teleport, as I suggested in another thread.

One hypothetical experiment is to have a few seeded games with temple on D:5 (so you have the choice of different gods based on the playstyle - we could probably rule out Trog if people want), and then look at the winrate?

But is it possible to answer this question without running an experiment of this kind?

No, It should exist. Spells need more drawbacks actually but seeing how a lot of people are already complaining at spell hunger... :lol:

Why do spells need more drawbacks?
To limit power? Vampiric melee is more powerful (I am not even talking about battleaxe with Makhleb)
To limit number it can be cast in a fight? MP exist.
To limit total number it can be cast in a game? Chunks still exist, after removing them spells will be very similar to wands.
To make players train more Spellcasting? They do it for MP reason alone.
To make players try to get more Int? They already do it for spell power, it's often optimal to put all 9 bonus stat points into Int for casters.
To make characters run out of food if they play too carefully? If it does not work for melee, why should it work for casters? Make out of depth monster summons already.
To make casters more annoying? Yes, it works indeed. When I want a mindless game, I play Be.

Spell hunger isn't totally irrelevant for spriggans. I mean, I guess it is in that you're a spriggan so you don't need big conjurations early, but the thing that really lets you not think about spell hunger is using magic to turn monsters into chunks.

njvack wrote:Spell hunger isn't totally irrelevant for spriggans. I mean, I guess it is in that you're a spriggan so you don't need big conjurations early, but the thing that really lets you not think about spell hunger is using magic to turn monsters into chunks.

This is what I meant in last sentence of my post, I am not even going to try SpFE/Cj/IE/EE because it is extremely annoying to check spell hunger before casting spells.
It's like "I can easily kill that Hill Giant with 2 Fireballs but I'd better kite it for 30 turns with Throw Flame because hunger matters".

Presently the only purpose it actually serves is to let you know if you have been just flat out training incorrectly.

If you literally put 0 points into spell casting, but (inefficiently and not very effectively) dump enough xp into spell skills to succeed at casting a high level spell (not many times, obviously, because your mp will suck) you will get additional negative feedback in the form of spell hunger becoming a problem.

This isn't helpful or useful to anyone who knows how to play at all, or like has read all the help files or spent any time at all learning b about the game, but it does serve newbies as a hint that you are doing it wrong.

Of course the drawback to that is that then novices tend to overreact and try to get all spells to 0 hunger cost before using them.

As to what purpose it actual *should* serve, if any, i don't have any opinions.

I've seen several times players having like Fire 20 and Conjurations 10, do we need to introduce "conjurations hunger" to help bad players notice this type of error?

Siegurt wrote:Presently the only purpose it actually serves is to let you know if you have been just flat out training incorrectly.

If you literally put 0 points into spell casting, but (inefficiently and not very effectively) dump enough xp into spell skills to succeed at casting a high level spell (not many times, obviously, because your mp will suck) you will get additional negative feedback in the form of spell hunger becoming a problem.

This isn't helpful or useful to anyone who knows how to play at all, or like has read all the help files or spent any time at all learning b about the game, but it does serve newbies as a hint that you are doing it wrong.

Of course the drawback to that is that then novices tend to overreact and try to get all spells to 0 hunger cost before using them.

As to what purpose it actual *should* serve, if any, i don't have any opinions.

This is not actually true. You can get noticable hunger with castable spells even if you have equal investment in Spellcasting and the relevant schools. This is most visible for specialists (who may just be working with a book drought) or for people using a source of Wizardry, which does nothing to combat spell hunger.

A 22 int human with 17 Spellcasting and 17 Conjuration will suffer 326 hunger every time they cast Orb of Destruction (nearly a third of a chunk for a single-target spell that might not hit anything!) at 2% fail rate. If they want to cast more powerful and consistent OODs, they should invest more in Conjurations, but if they want to cast it less annoyingly they want to invest Spellcasting...

Unfortunately even going to 27 Spellcasting will only partly help this poor human; they need to find an Int-boosting item to get less than 106 spell hunger. There's not actually a choice to be made; they just invest Conjurations or waste their experience.

A couple of weeks ago, I did an analysis in DCA which tried to show that hunger is basically never practically relevant, assuming one is not totally reckless. In particular, I looked at the hard cases (where the player themselves thought that they were tight on food, due to things like inedible corpses in Snake - inedible corpses are silly, but I digress), and found that actually, the game generated 180% of the food that they had consumed, so their worry was mostly psychological.

If anyone wants to repeat that experiment for something like Spriggans, it's very easy to do: just take a few random Spriggan winning characters (do "!lg * recent sp-- won -100" or something like that in Sequell), look at their action table and compare it to objstat data. You need to look at the herbivore nutrition only. It will take 10 mins max. My guess is that you'll find that the same general conclusion would hold: hunger is never practically relevant.

In any case, if hunger is supposed to balance spells, it's not doing a very good job of it. This is probably knowable just from experience, without doing any math.

So why can't the devs remove food chunks again? It's not like corpses would get removed and permafood functions better as a food medium so I don't really see why it can't. I must have missed it at some point since it's an obvious solution to this problem (?) of resource overabundance. My guess is that it's just genuine leniency, probably.

The devs are beings of supreme cruelty, cunning and deceit. No mere mortal should attempt to comprehend their esoteric ways. Ask only for the strength to endure the torment they have devised for you. To ask for anything more is hubris.

Doesnt wrote:
Siegurt wrote:Presently the only purpose it actually serves is to let you know if you have been just flat out training incorrectly.

If you literally put 0 points into spell casting, but (inefficiently and not very effectively) dump enough xp into spell skills to succeed at casting a high level spell (not many times, obviously, because your mp will suck) you will get additional negative feedback in the form of spell hunger becoming a problem.

This isn't helpful or useful to anyone who knows how to play at all, or like has read all the help files or spent any time at all learning b about the game, but it does serve newbies as a hint that you are doing it wrong.

Of course the drawback to that is that then novices tend to overreact and try to get all spells to 0 hunger cost before using them.

As to what purpose it actual *should* serve, if any, i don't have any opinions.

This is not actually true. You can get noticable hunger with castable spells even if you have equal investment in Spellcasting and the relevant schools. This is most visible for specialists (who may just be working with a book drought) or for people using a source of Wizardry, which does nothing to combat spell hunger.

A 22 int human with 17 Spellcasting and 17 Conjuration will suffer 326 hunger every time they cast Orb of Destruction (nearly a third of a chunk for a single-target spell that might not hit anything!) at 2% fail rate. If they want to cast more powerful and consistent OODs, they should invest more in Conjurations, but if they want to cast it less annoyingly they want to invest Spellcasting...

Unfortunately even going to 27 Spellcasting will only partly help this poor human; they need to find an Int-boosting item to get less than 106 spell hunger. There's not actually a choice to be made; they just invest Conjurations or waste their experience.

'noticable hunger' and 'spell casting hunger being a problem' are very different thresholds. If you're attentive, you will notice spellcasting hunger even if it's 1/10th of a chunk, or even 1/20th.

bel wrote:
tabstorm wrote:
ONIchinchin wrote:Because spells are the strongest player option in this game.

actually that would be heavy armor melee.

Obviously spell hunger shouldn't exist. I think it's a carryover from nethack.

But since this is CYC, I don't feel guilty about derailing the thread a bit. I am interested in this question. How would one go about showing that one playstyle is the "strongest"? What's the domain where one is arguing? (Obviously DEFi is weaker than DEFE, I think, so it can't be the whole universe).

For the sake of argument, let's assume that the player reaches D:5 or something, so the random early game isn't weighted too much. Or we could start every character with a potion of curing and scroll of teleport, as I suggested in another thread.

One hypothetical experiment is to have a few seeded games with temple on D:5 (so you have the choice of different gods based on the playstyle - we could probably rule out Trog if people want), and then look at the winrate?

But is it possible to answer this question without running an experiment of this kind?

Because it's obvious from actually playing the game that it's the strongest option in the game for most races that aren't elves, tengu, octopodes, or felids unless you declare the game to not include anything past D:8 on the grounds that it is so utterly trivial as to not count as a game. Since the developers have not yet deleted Lair, its branches, D:9-15, Orc, Elf, Vaults, Depths, or Zot, we can assume that this is not actually the case, at least for most people.

The trouble with spells is that just about anything that can be done with spells can also be done with god powers, evocations and ranged weapons, while spending less XP and wearing heavy armour.

I'm curious what tabstorm thinks of ranged combat.

VeryAngryFelid wrote:
njvack wrote:Spell hunger isn't totally irrelevant for spriggans. I mean, I guess it is in that you're a spriggan so you don't need big conjurations early, but the thing that really lets you not think about spell hunger is using magic to turn monsters into chunks.

This is what I meant in last sentence of my post, I am not even going to try SpFE/Cj/IE/EE because it is extremely annoying to check spell hunger before casting spells.
It's like "I can easily kill that Hill Giant with 2 Fireballs but I'd better kite it for 30 turns with Throw Flame because hunger matters".

When I finally managed to win a mummy for great player I wanted to win another unwon race, but I also wanted to streak, so I picked SpCj. I hadn't won a spriggan yet, and it's downright hard to die on a spriggan conjurer if you are playing super tedious streak safe tactics. I didn't have any trouble with spell hunger. I ended the game (3 runes) with 16 rations and 16 royal jellies.

What I will say about the combo is that it's horribly boring kiting everything. I got my first win offline, then played 3 naga games - my third win and first 15 rune game was under 142k turns. I was a naga monk who got necromutation. It was by all accounts probably a horrible game, and still got all 15 runes. I took 145k turns to get 3 runes on a spriggan. It is still the highest turncount game I have ever had - I typically run a bit low on turncounts (for someone who isn't usually speedrunning) and have never ever gone over 150k.

I am glad I was able to continue the streak (got up to 4 wins before dying), but trust me: SpCj, never again. This has nothing to do with hunger though, which isn't a problem.

TL;DR: As a newbie I played a 15 rune naga monk of chei with necromutation in fewer turns than as a veteran trying to streak SpCj of Vehumet for 3 runes.

There is no practical difference between ranged and melee in Crawl once you've reached the point where you are in no threat of running out of ammo. It is melee at a distance. Ranged is practically a subset of melee (though actual melee is still better than conj). In short, it's the heavy armor that is good, whether you want to melee at a range of 1 of 6 squares is up to you...

Personally I don't like ranged starts because of earlygame ammo management.

tasonir wrote:I didn't have any trouble with spell hunger.

Sorry, I cannot trust you on that.
Your SpCj

Code:: Iskenderun's Batt | | | 4 | 7 | 18 | 5 | 5 | | || 39

Your DgCj

Code:: Iskenderun's Batt | | | 1 | 85 | 153 | 67 | 4 | 13 | || 323

I ended the game (3 runes) with 16 rations and 16 royal jellies.

I was not writing about SpCj starvation. I was writing about SpCj playing in a very special (and annoying) way and your games and impression confirm that.

tabstorm wrote:Because it's obvious from actually playing the game that it's the strongest option in the game for most races that aren't elves, tengu, octopodes, or felids unless you declare the game to not include anything past D:8 on the grounds that it is so utterly trivial as to not count as a game. Since the developers have not yet deleted Lair, its branches, D:9-15, Orc, Elf, Vaults, Depths, or Zot, we can assume that this is not actually the case, at least for most people.

It's not obvious to me, therefore, I asked what kind of evidence would settle it one way or another. I've already said, for the sake of argument, you can assume that D:5 (or D:8 in your post) is reached by the player. Trog is probably a special case (probably for beginner reasons?); I am willing to concede that Trog is the strongest god and probably any decent character ^ Trog is perhaps strongest (though this is arguable). For the rest, I am not at all sure that heavy armour melee is the strongest playstyle, either in theory, or for "most people", by which I assume you mean "in practice".

For instance, I could channel my inner Berder and run a Sequell query for greatplayers (broken down by class; I removed races Op, Fe, DE, Tr, Gh, Mi and Te)

Code:: !lg greatplayers recent race=(hu|ce|dd|dg|ds|fo|gr|ha|he|ho|hu|ko|mf|mu|na|og|vp|vs) s=class / won o=% <Sequell> 2021/27173 games for greatplayers (recent race=(hu|ce|dd|dg|ds|fo|gr|ha|he|ho|hu|ko|mf|mu|na|og|vp|vs)): 88/685x Artificer [12.85%], 209/2021x Gladiator [10.34%], 76/763x Assassin [9.96%], 61/621x Fire Elementalist [9.82%], 76/812x Skald [9.36%], 125/1343x Berserker [9.31%], 58/637x Ice Elementalist [9.11%], 83/924x Hunter [8.98%], 67/748x Earth Elementalist [8.96%], 47/527x Conjurer [8.92%], <Sequell> 72/882x Necromancer [8.16%], 153/1906x Wanderer [8.03%], 56/736x Venom Mage [7.61%], 70/928x Enchanter [7.54%], 41/598x Air Elementalist [6.86%], 55/807x Summoner [6.82%], 233/3436x Fighter [6.78%], 50/830x Arcane Marksman [6.02%], 49/835x Warper [5.87%], 123/2233x Monk [5.51%], 54/991x Transmuter [5.45%], 56/1051x Wizard [5.33%], 68/1562x Chaos Knight [4.35%], 51/1297x Abyssal Knight [3.93%]

.
I don't see any obvious pattern of "heavy armour melee best". Fire Elementalists have about the same winrate as gladiators; enchanters, conjurers and all elementalists have higher winrates than Fi.

Yes, handwaving Sequell queries aren't definitive proof of anything; treat it as only a piece of the whole.

VeryAngryFelid wrote:I was not writing about SpCj starvation. I was writing about SpCj playing in a very special (and annoying) way and your games and impression confirm that.

The main point of my post was actually how annoying SpCj was, and how it took the most turns to win of any game I've ever played, because of the kiting and running away which spriggan speed enables, so yeah, I agree with you on that

I'm not entirely sure that the casting of battlesphere really shows a fear of spell hunger, though. More complete cast charts from those two games:

Spriggan:

Code:: Cast: Magic Dart | 144 | 199 | 298 | 106 | 39 | | | | || 786 Searing Ray | 5 | | | | 1 | | | | || 6 Iskenderun's Myst | | | 103 | 534 | 1124 | 418 | 64 | 1 | 6 || 2250 Iskenderun's Batt | | | 4 | 7 | 18 | 5 | 5 | | || 39 Airstrike | | | | 1 | | 5 | | | || 6 Venom Bolt | | | | | 431 | 888 | 366 | | || 1685

And end game spells were bolt of cold and iron shot, both cast over 1000 times, with 500 casts of freezing cloud.

Demigod:

Code:: Cast: Magic Dart | 99 | 154 | 341 | 390 | 295 | | | | || 1279 Searing Ray | 4 | 16 | 38 | 7 | | | | | || 65 Iskenderun's Myst | | | 66 | 314 | 1094 | 397 | 21 | 38 | || 1930 Iskenderun's Batt | | | 1 | 85 | 153 | 67 | 4 | 13 | || 323

I later moved into bolt of cold (~600 casts) and freezing cloud (~500 casts), however I cast them much less than on my spriggan, because I had the unrand +1 eveningstar "Brilliance" {holy, rN+ AC+5 Int+5 SInv Stlth-}, which made me go somewhat melee heavy for a conjurer.

Seems like my spriggan cast much more overall, my demigod shows I finally learned how to use searing ray (I seriously didn't know you had to rest to keep casting it for an embarrassingly long time), and my spriggan had a slight preference for casting mystic blast over battlesphere/magic dart, which the Dg used for much longer. Not sure why I didn't use battlesphere more on the spriggan, but I was mainly casting mystic blast/venom bolt on those levels, so it wasn't really over hunger costs...

I cannot find "Brilliance" in your SpCj game (I checked lst file also). You are right, maybe you used Mystic Blast more because Sp can kite easily so it does not need firepower and MP-efficiency of battlesphere that much.

I just escaped with a hunger's blade runner DsNe^Fedhas wearing an artifact amulet of rage. The highest hunger spell was deflect missles ( #### ) which i was forced to cast pretty often despite wearing a shield but EV just 24 when escaped, Fedhas Reproduction also eaten lots piety in vault:5, which forced me to eat only permafood until Zot. Berserker rage was casted often all over late game while carefully hesitated abuses of bolt of draining and agony.
IIRC in Zot:2 the bad news: i had just one bread ration, panic. Started the hunt for chunks out of draconians while managing to survive bitchy layouts with carefull use of best abilities.. until i win with a meat ration and one fruit in the bag. Btw the first time having both RMSL and DMSL in the spell list maked sense.
The tension made the game really satisying

VeryAngryFelid wrote:I cannot find "Brilliance" in your SpCj game (I checked lst file also). You are right, maybe you used Mystic Blast more because Sp can kite easily so it does not need firepower and MP-efficiency of battlesphere that much.

The brilliance mace was on the demigod, and hence he ended up casting less than the spriggan. But really late game casting isn't the issue, because by then it's easy to have 20 spellcasting if you want it, and pay no hunger costs.

Back to the big picture: maybe the issue is that spell hunger seems bad early on, and then can be utterly trivial later, to the point of having literally zero spell hunger on your spells? What if it was lower at first, but also couldn't be fully removed, so that spells would always cause some amount of spell hunger? That would make it at least meaningful, although I think the better design is the whole 'remove food' instead, which is completely in the opposite direction

The Tavern

Question: why does spell hunger exist

Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist

Re: Question: why does spell hunger exist