Armor and strength advice

Hurkyl · March 17, 2025, 10:49pm

Hi. I’m trying to relearn dcss again from having played it in ancient times, and I have two specific strategic questions.

For context, my favorite used to be a SpEn that transitions into a caster type (air mage) over the course of a 15 rune game for when stealth fails, and am mostly starting with that today.

First is about armor. In the old days, because of the way GDR and encumbrance worked, it was extremely effective to put on Swamp Dragon Armor, and even upgrade to Pearl Dragon Armor (or Fire/Ice) late in extended.

But… my impression is that things have changed in a way there is not so effective anymore. It doesn’t seem like you can mitigate casting penalties nearly as effectively, and GDR has changed in a way that this upgrade isn’t nearly powerful as it used to be. Maybe even EV changes play a role here too.

I’ve even seen people suggest things like it’s not even worth upgrading to Swamp Dragon Armor (from Troll Leather Armor) until it’s fully enchanted. (or you really need rPois)

So do I have the right impression about things? Its it no longer strongly recommended to transition to medium dragon armors?

The second question is about strength. In the old days, there was a minimum value for Str (that I’ve forgotten) below which it was very dangerous to go about the extended game due to risk of getting stat-drained to zero.

Is that still true these days? What is the number and the reason for it?

And because of the modern encumbrance mechanic, is there some sort of Str target for someone who intends to cast lvl 8/9 late in a 15 rune game while wearing Swamp Dragon Armor? Or maybe Troll Leather Armor? (or even Pearl Dragon Armor if that’s feasible)

stoatsoup · March 18, 2025, 7:48am

How ancient are the times? I don’t think the spell success formulae have changed much (if at all) in the time I’ve been playing (since 0.18) save the change to shield penalties.

Stat zero is less dangerous now because it doesn’t kill you, and I think stat drain is going away altogether in 0.33.

Ignore any rule of thumb that relates Str and ER linearly.

My impression is that in all but the most extreme cases (where even attempting to get the spell online is kind of futile), a point of Int does you more good than a point of Str. I don’t take Str in order to wear armour; I see what armour I can wear with the Str and spell skills I have got.

PaperRat · March 18, 2025, 10:06am

For a low Str combo like SpEn a point of Str would give more spell success than a point of Int, if you’re wearing stuff heavier than a robe (though Int of course also gives you spell power). So if your plan is to cast high level spells, while wearing medium scales in extended, then getting Str on a couple of your levelups is perfectly reasonable.

That said, robes (or steam dragon scales) are a fine end game armour choice for a sprig. You just gotta lean heavier into EV (high EV chars are quite durable in current version), be aware of the stuff that ignores EV (explosions mainly) and cover your resistances diligently.

Personally I don’t value troll leather on sprigs, cause it gives bad defences (its stats have been nerfed slightly since you last played) and regen is not super worth on a -30% HP non-melee chars, IMO.

If you play trunk, you can press ‘!’ when inspecting a piece of gear and it’ll show you how your spell failure would change, if you equip it, very handy for getting a feel for what stats/gear you’d want on your char.

Hurkyl · March 18, 2025, 12:05pm

Last played 0.19, I think: I remember clouds vanishing when out of sight being relatively new, and am pretty sure mottled dragons still existed. But I learned the game back when armor had an “EV penalty” rather than an encumbrance rating, and I think I was remembering some of the thresholds relating to that.

I used to really dislike TLA, but getting back into the game I feel like regen is super useful in one particular failure mode: when you are in “run and hide” mode but can’t safely escape to somewhere truly safe. Regeneration makes it more likely that you can recover before something forces you to find another temporary hiding place.

And to a lesser extent, heals while maneuvering and repositioning add up over time.

That and it seems like regeneration is more helpful than a couple extra points of AC. Although I’m not entirely certain because of GDR.

Regarding stat zero, I know it’s not instant death any more, but it’s still a fairly serious condition. I think 6 Str is the breakpoint I was thinking of, because of the mystical force?

I was hesitant to lean excessively into EV because of stepdowns that existed when I learned, but I do get the impression it’s more effective than I’m used to so maybe I should be less afraid of that.

zikada · March 18, 2025, 1:06pm

github.com/crawl/crawl

Remove all stat draining from the game

committed 05:23AM - 12 Jan 25 UTC

DracoOmega

+83 -529

While stat drain could still be caused by numerous different sources, its positi…on in modern Crawl felt fairly vestigal. The largest sources of int drain have been gradually removed, such that very few sources left pose a realistic threat of causing a player to hit stat zero. And even for the cases that remain, it's unclear that "Does very litle to most players, but in specific circumstances to very specific characters can rarely cause immediate paralysis that maybe kills you" is actually good. While it is theoretically possible to make a stat drain system that has a bit more impact in general without being as spiky in the worst case, I think that it would be better not to be beholden to the historic nature of this system (and the small-yet-often-complex effect stats have on a player) when designing any new medium-term 'debuffs'. And so, after some discussing, I am removing stat drain entirely (and stat zero itself will be simplified in a subsequent commit). The following changes have been made: -Stat drain attack flavors have been removed without compensation. The only monster whose stat drain could be relevant more than once in a blue moon were ghost moths, who already have plenty going on without it. Eidola and putrid mouth stat drain already did 'basically nothing' and they may recieve something else in a future commit. -Vampire batform loses the stat damage with no compensation (they're not long for this world in any case). -Klown pies lose stat damage without compensation. (They can already do a half-dozen other things which are usually more impactful anyway.) -Degeneration clouds are now clouds of alcoholic mist, which inflict Vertigo (to match the change to the potion). This is very silly, but the clouds were already rather silly and only exist in Alistair's wizlab, in a room where you likely never step on them. -Chei wrath loses stat drain without compensation (that wrath effect already generated massive noise). -The weight of Kiku's torment and HP drain effects absorb the weight of what used to be stat drain (and the HP drain is reduced in strength a little, in compensation) -Trog's str drain is replaced with long-duration Weakness -Sif's int drain weight is split between existing MP drain and confusion -Undead being malmutated now have their HP drained instead -Zot traps now blind instead of drain stats -Xom stat drain has had about half of its weight moved into pseudo-miscast Placeholder changes: -Hell effects has the old stat drain weight split evenly between HP drain, temp badmut, and doing nothing (and the HP the drain amount lowered a bit in compensation). This is temporary; there were already some awkwardnesses with hell effects, regarding going elsewhere to farm it off (and its interaction with the Persistant Drain mutation), and temp mutations don't last more than a couple kills in Hell and are largely irrelevant (their timer behavior really wasn't designed with this sort of system in mind). I already planned to change how they worked when Hellpan routlette is implemented, but this should hopefully be at least 'functional' until that point. -Summon horrible things now has a chance to give a temp Weak-Willed mutation instead of int drain. This is... slightly silly and low-impact (if thematically appropriately), but just a placeholder for a new type of effect I plan to implement later.

github.com/crawl/crawl

Simplify stat zero

committed 05:23AM - 12 Jan 25 UTC

DracoOmega

+74 -172

With stat damage removed from the game, hitting stat zero has become even more u…nlikely. While inopportune mutations and equipment that reduce stats can still do this, the former are even more situationally unlikely to make you hit zero than stat damage was, and the latter are mostly a momentary misstep by the player that can be immediately corrected. Given this, I think it is appropriate to greatly simplify the effects of having 0 of a stat. Instead of immediate paralysis, permaslow, additional stat-specific effects, and only gradual recovery, having a stat at 0 or below just gives the player permaslow so long as it is below 0. Once it is above 0 again, they are completely fine. This allows us to eliminate all checks for swapping form or equipment which might reduce a stat to 0 (being briefly slow is not the same sort of emergency that paralysis is, and there are no lingering effects.) This is good, because the code was somewhat complex and already full of corner cases that it didn't catch - and permaslow already seems quite sufficient to dissuade a player from simply powering through having -8 int from their gear. Whether even *this* level of effect is worth having compared to just capping stats at 1 is a little unclear to me, but it does keep players from considering artefact stat penalties to their dump stat to be completely free, and maybe once in a blue moon it makes a negative stat mutation be more impactful? Some of the removed status effects may find a home again as part of future debuff effects not tied to stat damage. (The stat_colour option is still around for now. Technically, there is still some value in making it obvious that the player is close to 0 in a stat, but it feels a little too... emphatic for the current danger level, in my opinion. I may also cut it in the future, if other people are in favor of doing so.)

petercordia · March 20, 2025, 5:39pm

I think the EV step downs (as well as the 5% guaranteed hit chance) have been removed.

I do agree having some extra regen is very nice, and on Sp there’s less chance you’ll find some randart equipment that gives Regen+. I’d probably take archmagi over Regen+ most of the game, but I could see myself using Troll armour in the mid game.

Hurkyl · March 21, 2025, 8:20am

I think I’ve literally seen the Archmagi ego twice since I started playing crawl again: once on Robe of Folly, and once the ring of the Mage dropped when I was taking a break and running an SpBe for variety. So so far it’s been Regen+ versus resistances and/or AC.

Regarding EV… how much have enemy ranged attackers been nerfed? In ancient times, if my SpEn encountered its first Centaur at mid-range without cover to hide behind, it was a serious risk of game over if I didn’t have a consumable to get out of the situation.

But today I feel like I wouldn’t be too scared of charging at my first Centaur from max range. (although I still pull in around a corner since that’s a lot safer, or use Tukima’s Dance)

Is this just an “EV is more effective” thing or have things like this been seriously toned down?

PaperRat · March 21, 2025, 9:35pm

The game displays chances to be hit and max damage the enemy can do in the xv screen, so you can just check that. EV is better when you stack lots of it than in the past, but I don’t think enemy damage was toned down. Some spawn tables have changed (no more D:1 halberd gnolls for instance), but if anything the damage has gone up, cause of the changes to enemy weapon delays (and attacks of opportunity). Basically weapons don’t affect enemy attack speeds, it’s always 1.0, unless the enemy is fast/slow or special cased (like for example ogres and juggernauts are normal speed, but have slower attacks).

Charging at centaurs as a sprig is still ill advised btw :), cause they do a lot of damage per hit with ranged attacks, and cause as a sprig you are faster than them anyway and can kite them to a corner almost always.

I’ll also note that minutia of stats and gear choices and such doesn’t make that much difference for winning a run. Your decisions around positioning and consumables, and threat assessment are much more important.

Hurkyl · March 23, 2025, 8:13am

The thing I was trying to convey with the example was a level of concern thing, not actual tactics. In ancient times, being in a situation where I was midrange against a Centaur and would eat 2-3 ranged attacks whether I closed to melee or fled LOS was something to escape at all costs. Today it’s not alarming at all. Sure there’s still significant pressure to close to melee or flee LOS, but it’s definitely not an emergency.

I’m content to chalk it up to a dozen updates worth of things that have influence on accuracy meaning that my SpEns today are just that more resilient against going splat.

And I don’t count “you can get randomly paralyzed for 2 turns completely out of the blue” to be minutae. It doesn’t really matter until you get to extended, of course, but I do intend to start doing that again once I reacclimate to the game.

But, I guess, it won’t matter at all next patch.

PaperRat · March 23, 2025, 4:42pm

And I don’t count “you can get randomly paralyzed for 2 turns completely out of the blue” to be minutae. It doesn’t really matter until you get to extended, of course, but I do intend to start doing that again once I reacclimate to the game.

Oh, haha, I didn’t mean that, knowing about which mechanics exist in the current version is pretty important!

What I meant by minutia was fiddling with your skills too much, or weighting advantages of +2 robe versus +1 leather and such. It’s stuff that sometimes players really like to obsess over, cause it’s very… quantifiable, I guess. “it’s optimal to train Fighting to 6 straight away”, “always wear a shield”, “troll leather is BiS on such and such”, this stuff is not very important for winning, IMO. What I think is important is knowing where your particular approach is strong and where it’s weak and taking encounters accordingly, but all the hard rules and prescriptive stuff is mostly bunk, in my experience. Although of course having a sort of a plan can be useful to lessen the mental overhead a bit :).