God balance

[sanka]

I know that currently gods are not very balanced. It seems straight contradicting to the design goals as it creates a bad choice inside the game for the theoretical player who wants to win: should I choose the powerful but really annoying or the very weak but fun god? Some people claim that gods are "difficulty setting", but then they should not be an in-game choice. Also, dpeg always mentioned in god design threads that balance is not an issue, because it is easy to adjust later. But nobody seems to adjust them later.

I propose to start a great god balancing plan, gradually at least try to make them competitive with each other.

Since there are gods that clearly for challenge like Xom, I propose to divide gods into two groups. Those who are worse than not worshiping anyone (for the purpose of increasing win chance) need not to be balanced, but should be visibly distinct for the rest, so they are not newbie traps anymore.

Second, pick a god which offers a particulary nice play, for example Okawaru, and try to balance other (non-challenge) gods to them. Of course I do not want every god equally good for every character, but at least they should be a good choice for some characters and there should not be such a huge differences in power.

Third, please cut back the number of gods that offer allies.

[Majang]

Third, please cut back the number of gods that offer allies.

YES!

[Siegurt]

I think the solution to powerful-but-annoying gods is to make them not annoying (possibly by getting rid of them) of course getting consensus on what is "annoying" might prove harder than you imagine.

Also "worse than athiest" is pretty much just xom, and even that is arguable (some would claim chei, but some would debate that) and xom's description pretty clearly lays out what you are getting yourself into.

[Majang]

Chei asks a lot, but he also offers a lot (except allies).

[Aean]

With regards to balancing the gods, I see it a lot like balancing weapons within a type. Is a whip of venom supposed to be balanced against a giant spiked club? The whip is better early on, the club is better later on. And then there are options like the morningstar which may be worse early on than the whip, and worse later on than the GSC, but are useful throughout. They're not balanced because they're not competing in the same field. The same is true with gods. Trog may be devastating early on, but falls off hard by the time you get to extended; TSO isn't very useful early on, but one of the strongest late-game gods. Trying to balance gods that aren't meant to do the same things at the same time isn't logical.

That said, let's look at your proposal: separate gods into those that are worse than not worshiping anyone (as ones that don't need to be balanced) and balancing the others to "a god which offers a particulary nice play."

For the first, as Siegurt said, Xom is probably the only god that could fairly be called worse than not worshiping a god; even gods with powerful handicaps, like Chei, are amazingly powerful if played right. And even Xom can give you some serious benefits, if you're lucky, and you play carefully - my last CK was gifted 2000+ gold before D5, for example. So this doesn't seem to be much of a category.

For the gods to use as a baseline for balance, this comes back to my original point - what are you balancing against what? And how do you measure their value in a given situation? Okawaru is generally useful throughout, but his utility is probably highest in the midgame, where you have access to and benefit from all his abilities and are receiving his gifts. Trog is far more powerful early on, but far weaker later on - does that mean that Trog needs more early weakness and more late-game power? Does TSO need an early-game buff and a late-game nerf? How valuable are Okawaru gifts vs Sif gifts vs Trog gifts vs, say, Yred gifts? Depends on where you are in the game, as above. Depends on your character background, and your species - a TrFi benefits much more from an artefact GDA than a DECj. Depends on how you're trying to specialize - does your character rely on spells for damage, or defense, or escape, or hexing, or allies, or melee buffs, or not at all? Depends on your playstyle - are you oklob-farming with Fedhas, or just pulling out abilities in emergencies? Are you saving Oka piety for maximizing gifts, or spamming heroism? Etc., etc. As you wanted, they are already all good choices for some characters (with the arguable exception of Xom); any differences in power depend on so many other factors.

Finally, when it comes down to cutting down the number of ally-gifting gods...what do you suggest instead? Seems that the things gods provide are as follows:

passives (e.g. mutations, resistances)
offensive abilities (e.g. damaging, incapacitating enemies)
defensive abilities (e.g. damage reduction, regeneration, healing)
escape abilities (e.g. blink, enter abyss)
gifts (e.g. allies, spellbooks, weapons, armour)

If not gifting allies, what should gods do? Importantly, what should gods do that's consistent with the design of the god, useful without being overpowered, and doesn't tread upon another god's design space?

[duvessa]

The majority of other gods beat TSO throughout the entirety of a 3-rune game and Kiku and Gozag are probably better throughout extended as well. The only god that "falls off" significantly in 3-rune games is Fedhas and that's only because Fedhas is absurdly powerful in D/Lair/Orc, not because it's weak in Depths and Zot (rain is broken). It's true that the majority of gods don't do much in extended, but that doesn't strike me as a good thing.

And then there are options like the morningstar which may be worse early on than the whip, and worse later on than the GSC, but are useful throughout.

A morningstar is worse than a GSC at both low and high skill.

Trog is far more powerful early on, but far weaker later on

You have got to be kidding me.

As you wanted, they are already all good choices for some characters (with the arguable exception of Xom)

I cannot think of ANY character where any of the following gods is the best choice: Ashenzari, Beogh, Cheibriados, Dithmenos, Elyvilon, Lugonu, Nemelex Xobeh, Okawaru, Qazlal, Ru, Sif Muna, Vehumet, Xom, Yredelemnul, Zin, the Shining One.
It is, as far as I can tell, better in all situations to pick one of Fedhas, Gozag, Hepliaklqana, Kikubaaqudgha, Trog, Uskayaw, Makhleb (for deep dwarves), or Jiyva (if you somehow get an early altar). Eventually, if you want to get 15 runes, Zin and TSO become slightly relevant, Vehumet is relevant for ziggurats, and Lugonu is relevant for converting to it in Abyss. That's still about half of the gods in the game rendered almost irrelevant by a small pool of gods being ridiculously strong - the only time you should worship the others is if you can't find an altar to one of the best gods. To continue with the weapon analogy - which is kind of silly since weapons have way less effort put into them, and difference between them, than gods - this would be like daggers and spears being better than all the long blades, axes, maces, and staves, so you'd only wield any of the last 4 when you can't find a dagger or spear.

[amaril]

It is worth noting that zzxc and elliptic's streaks made use of vehumet, an oft memed-on god, probably due to availability/character synergy. Is it a bad thing that players might have to make the decision of taking a decent god early vs an optimal god later?

I think balance is good, and sif for example could really use a buff. But making gods balanced is less of a priority than making them playable.

[watertreatmentRL]

Optimal and early are pretty much the same thing. Playing for streaks is mostly about the first 30 minutes of the game. I don't know much about zzxc, but it's pretty clear elliptic is not always bringing his A game in the famous super streak. The game just isn't that hard.

[amaril]

Vehumet lets you build a certain type of character consistently (if your background is appropriate) and this type of character can steamroll the game at a certain point. In a streak in particular where after the early game u are mostly playing a game of 'don't fall asleep while driving,' vehumet is valuable, maybe even 'fun' or 'less unfun' or whatever. Except for the spell-gifting, which everyone knows sucks, veh is a reasonably well-designed god i.e. minimal & intuitive and that makes him less of an issue (even if he is weaker) than ely or beogh who, regardless of power, are unplayable.

edit: I don't think the state of difficulty/tedium in the mid/lategame is good, but that is an entirely separate topic

[Shtopit]

Some days ago there was a discussion about Trog. I almost always play melee characters. When choosing a god at the temple, I can't but feel that I am actually doing a mistake whenever I don't choose Trog. Gods should have a mechanical reason to go for them, not just the desire to try something different. This is one of the many reasons why balance is important. A few months ago, it was a point of contention.
Trog in particular is an important god for new player. My first win was a DsBe. So he should remain in the game, and he has a role. But I think that he should be only selectable from the character creation screen, or be an outsider deity like Logunu. You should have a god powerful, simple and fun enough to allow new players to push to a win, but also with mechanical reasons not to choose him each time you play that archetype.

(Powerful, fun and simple mainly in opposition to Fedhas, who I am told is powerful, but I could never bring myself to play with because it bores me.)

Cheib is a god with enormous drawbacks, and still very much loved. This is because Cheib is very simple (no scrolls collecting, no spells needed), very straightforward, allows for unique playstyles, has a passive power that makes single decisions more incisive and the game faster, and active powers that are useful and with many different applications, and also are different from most spells/abilities/powers.
Cheib could actually do without part of its active conduct: allow use of potions of haste and berserk (although the difference with renewable sources may cause confusion), quickblades (that's a strange one), weapons of speed...

The "gods with allies" made me laugh a bit, it's like having going to a pie shop and finding that, out of 25 different pies they have, 10 are cherry pies. When you ask what the deal is, they start saying "this one is eagle eye cherry, this one is peak cherry, this is Sargent's cherry, this one...", they're still all cherry. Allies are a convenient way to give an always valid power to gods, especially because heavy meleers probably can't cast much in the way of summoning, and allies are universally useful (even though I don't like them). The problem is that too many gods are centred around allies: Hep, Kiku (in theory playable as blast necromancer), Beogh, Hydrolemur, Fedhas have them as shtick; TSO, Makhleb, Lugonu, Nemelex, Qazlal, Trog all grant allies as a more-or-less-secondary aspect.
There was an idea some time ago of turning Beogh into Orc Hep with respawn.

[Hellmonk]

It's hard to talk about god balance independently of the weird backwards DCSS difficulty curve. When the difficulty is so heavily frontloaded, a natural consequence is that gods that don't do anything very early become bad choices for winrate/streak play. It's no coincidence that every top god selection does something immediately upon joining or at *..... piety. Gods are still not balanced even if you consider lategame, of course, but I think that if early D was easy and depths/zot was pants-shittingly terrifying you'd have more consideration given to some gods with powerful abilities that only come online later in the game. If you want to keep the DCSS difficulty curve and also balance gods, you either have to give every god a strong early ability or take the strong early abilities away.

[bel]

About the OP, I do think Gods should be roughly balanced. If you want players to choose a difficulty level, it should be at the start of the game, not afterwards.

However, I do not think the analysis of God power by winrate/streaking is a good metric, for reasons I have spelled out elsewhere. Using that metric simply means that 90% of the game is meaningless. Moreover, it does not reflect the subjective feelings of the players, because "optimal play" (by this measure) is so boring that few people want to do it.

It's better to use something like score, but unfortunately, the current way of scoring is pretty bad. I have some ideas on how to create a better scoring mechanism, but it will have to wait.

[mattlistener]

When I choose between weapons, I'm trying to make the best choice for my character based on what I have available.

It's not the same with gods. I want to have varied gameplay experiences so I will choose not-maximal but reasonable gods for my character and have fun trying to win with them. Put differently, the gods I've won with feel "solved", so now picking them in order to win feels boring to me.

When someone argues "X god should be removed because there is no character for which X is the best choice" I cringe. In my opinion that is not a valid reason to remove a god. Other reasons may well be -- I never got to play Pakellas, but if there was no way to render her without opening the door to degenerate game-breaking behavior, okay.

[tasonir]

Chei asks a lot, but he also offers a lot (except allies).

If chei is ever given allies, that's the version I'll refuse to upgrade to, seriously! I mean it! Fortunately, I think this has almost no chance of happening.

I think that Qazlal's allies could be removed pretty harmlessly, they aren't really a central part of the god. Just give the passive cloud effect, upheaval, and apocalypse or whatever the final nuke is (I mix it up with Ru's final nuke all the time). You could come up with another minor ability or just leave it entirely. Imho the clouds and upheaval are 99% of the god, upheaval is just that strong of a nuke. It's amazing for characters who have trouble otherwise with ranged offense - picture a low int character who's bad with ranged weapons, maybe a troll or ghoul or something. Trolls do have large rocks of course, but upheaval is both aoe and smite targeting, so there's a valid case for it. You could do both. My first ever Qaz game was a ghoul who I won on my first try.

Overall though I think hellmonk hit it on the head, it's the high difficulty up front which makes it nearly impossible to balance gods - getting a good power early is more important than overall design or final strength. I love chei and I'm willing to put up with 15-30 minutes of more challenging gameplay to build the character I want, but chei isn't really good until you have at least 4* piety, and is much better with 6* piety.

If we were willing to really tackle the "early ability power level", I'd much rather see those early powers moved to higher piety levels or otherwise nerfed, and not all gods getting OP powers at 1*. I wouldn't mind trog's berserk remaining available for newbies, and it does have downsides that can get you killed if you use it poorly, so it's a decent learning tool. Mushrooms don't really have any downsides aside from the piety cost, but even if you stay at 1* piety on fedhas you still have access to mushrooms...