Overflow Altars before D:4

[4Hooves2Appendages]

If I understand this correctly, overflow altars (i.e. altars in the dungeon, rather than in the temple) were introduced because the number of gods exceeded the number of altar spaces of many, especially older, temple vaults. (Another solution to this, which I am not plugging here, might have been multi-level temple.)

I don't quite understand how this lead to allowing altars to spawn earlier than D:4.

The recent, and previous, and probably forever on-going, discussions around god selection before the game, made me wonder about altars on D:2.

Typically, I do try to win when I play. For example, I explore manually in parts of the game, I use summons, necromancy, evokers, invisibility, luring, etc. Although occasionally the tedium of certain things gets the better of me, like managing Vp hunger status, or eating lots of mutagenic chunks, and I deviate from what I believe maximises my chances of winning. Auto-explore falls into the category of course.

Naturally, wanting to win means picking strong gods. For me that means gods that give immediate, or near immediate, help when I need it most: early. The obvious candidates are Trog, Kiku, Fed, Gozag, Hep, etc, and not Xom, Chei, Ash, etc.

(As an aside, If I was allowed to choose a god before the game, then I would treat that decision not as part of trying to win the game. Similarly, I tend to play weaker combos deliberately.)

The problem I have is that I end up in situations where I win a game with Fedhas, then in my next game the first altar I find is Fed on D:2. Now, in order to win, I should join the mushroom cult. But come on, two times Fed in a row? If I found the temple first, then I could just choose another useful god. In this case I'm left with leaving the altar, and handicapping myself by carrying on. The difference between converting now and waiting for the temple on D:4+, or hoping for another useful god, is quite large in terms of power.

I don't quite know what the solution is, or even if it would generally be regarded as a problem, but keeping overflow altars a bit lower in the dungeon might be an idea.

Finding a strong god close to the first D:1/2 stairs is approaching starting with a strong god in any case.

Mainly, I would like to ask, are overflow altars before D:4 good? How many people actually regularly choose gods at the temple these days? I usually don't enter it anymore.

[njvack]

I think the design intent is that you might find a weaker god early and need to decide if you want, say, Vehumet on D:2 (who is maybe workable but suboptimal for your character) or wait for Kiku or Fedhas or Ru who might show up later.

Nowadays there are enough gods who will be strong for a variety of character builds, and enough altar vaults (maybe these are different from just getting an overflow altar?), that there's usually a very strong choice available before I see the temple. I may wait until temple because I really want a particular god, but I have no illusions that this is the best choice for winning.

Quick idea: what if the distribution of altars was weighted such that subjectively "weak" gods were more common earlier, and "strong" gods tended to be at or below temple depth?

[Majang]

Oooh, I would love this. Simply for the reason that common lore (at least here) calls the gods weak that give me my surest wins and funnest games, and therefore I could expect to see them a lot earlier than usual. And to know that I would never, ever see an ally-god's altar in the early game would make this game much nicer!!!

[4Hooves2Appendages]

I would not be opposed to that, but it might be difficult to agree on the comparitive power levels of the gods.

[njvack]

Doesn't have to be exact; that's why I said subjectively weak/strong. But I think there is broad general agreement that, say, Fedhas is stronger than Xom.

[Factorialite]

Another idea is that only faded altars can appear on floors before temple. A bonus to that is that when you find an altar to a specific god, you know that the ecumenical temple is on that floor.

[4Hooves2Appendages]

Only faded altars before D:4 is a good idea in my view. I wouldn't usually use them, unless I was really desperate, but I also wouldn't feel bad about passing on Fedhas or Kiku.

[amaril]

I think altars outside of temple could stand to be rarer, maybe slightly less rare than shops. Alternately, they could remain as common as they are and the temple could be removed. If god choice is supposed to be an *actual strategic decision* all the altars need not appear (or a temple with all unseen altars could appear on d15 or something).

IMO a worse problem than "find fedhas on d2 and be compelled to join" is "find dith or oka on d2 and feel no need to join because altars are so common and gozag or hepl on d5 are stronger anyway." So the real problem is god balance. If there were no overflow altars, your god choice (if playing optimally for speed or for win %) would still be limited to roughly 3 gods, so the "oh boy another fedhas game" dilemma would persist.

Gods should be balanced roughly around okawaru imo, but that would be difficult to do without removing ally gods entirely.

[johlstei]

Doesn't playing to win compromise novelty in almost every sense, not just god choice? If you're using optimal tactics, if you want to play to win you have to just keep using them. You're generally punished for novel branch order, tactics, skilling, etc. I think god selection is one of the cases where this is the least true, most of the time (95% or more) your first altar is going to be different than it was last game. If maximizing win rate is your true goal, shouldn't you be excited to get a real strong god like fedhas twice in a row? Should novelty between runs be an explicit goal in the design notes? It's there already in terms of level/monster/item generation. I could think of other ways of increasing it if that's what we want but I'm not quite convinced.

[amaril]

The reason I would not like fedhas or kiku twice in a row is b/c now I'm stuck dealing with the terrible corpse/ally management interfaces in multiple consecutive games. I think part of why dynast went gozag every game in his notorious streak (this included waiting to late d until the altar spawned, which was definitely not optimal) was b/c gozag is both powerful and not particularly micromanagement dependent. Having less micromanagement (with no cost to power) means you can pay more attention to the core tactics of the game and play at a quicker pace and a higher average level.

I think that if crawl has any merit whatsover (I think it does) tactics are situational and skill-testing. At a high level of play, playstyles diverge. This creates novelty. I agree with the gist of your post, I think. Strategic novelties (gods) tend to homogenize tactics. A crawl-like game with no micromanagement and no strategic game insofar as strategy is distinguishable from tactics sounds like a good direction to me.

[4Hooves2Appendages]

If maximizing win rate is your true goal, shouldn't you be excited to get a real strong god like fedhas twice in a row?

It's not a well-argued point, but in essence that's not how my brain works. For me there is an internal struggle between wanting to do everything I know will help me win and not doing things that I find annoying. I knew that picking Fedhas would help me win, but in that case I couldn't convince myself to hard mushrooms again. I left the altar, but I didn't like leaving it. Ironically, I ended up with Kiku...

Ideally, that internal struggle would never be required. Everything that helps winning should also be natural and straightforward and free of tedium. Most of these concepts are subjective of course. Poor designers...

[njvack]

One other potential solution would be to open more of the game to overflow altars, but place the same number of altars in total.

[Yknaar]

From my perspective as relatively new player, overflow altars serve three purposes that do a lot to areas of the game other than power balance:

• they allow new players to get acquainted with gods one at a time without being overwhelmed by naturally spacing the altars out (that's how I started in 0.19.1),
• they make the first few levels more varied - especially since some of the altar vaults are ingenious (that one with Yredelemnul and human being zombified, or Hepliaklqana being stuffed in a 1-tile closet),
• they are an early, important factor that makes a run feel lucky or unlucky - especially for players who already know which god they will pick.

I felt a need to point that out since the discussion seemed to be getting quite focused on balance only.

[yesno]

I like early altars in that they often influence my selection of deity. I used to always go to temple to choose my god and so I usually had a good idea of whom I would worship. With early altars I often think, not my first choice, but at d:2 I'll take it. For me it makes for more interesting games with gods I might not otherwise have chosen. If I really don't want to take it because it's a boring choice for me, I just don't do it, same as when I discover a powerful item that I should use but don't feel like using.

[mattlistener]

"Only faded altars before D:4" seems boring. If they were close to break-even in value (early religion almost being worth downgrading from choice-of-god to random-god) it'd be more interesting.

Weighting gods by strength and depth would be really good. It's a lot more interesting to win a game because you took a non-preferred god early, or survived to take a preferred god late, than to win because RNG served you up your preferred god early and you snowballed from there.

[4Hooves2Appendages]

To take it in little steps, perhaps there could be a maximum of 1 overflow altar on D:2 and a maximum of 2 on D:3.

[chequers]

I like the idea of restricting the number of overflow altars before temple. Perhaps an average of 0.5 altars per floor from D:2 to temple's floor.