What could I have done better? (And two questions about playing casters)

Argyros · July 28, 2026, 2:47am

I’ve been playing the game for a few weeks now and feel like I can usually get MiFi^Oka to Lair/beyond. The furthest I’ve gotten to is Depths:2.

I like playing other things for variety, and have also had some success with different types of Deep Elf caster (notably fire elementalist and conjurer.) I’ve tried Djinn and didn’t like it as much. So, questions 1 and 2: how are the various types of caster usually ranked in terms of difficulty? And when playing Fire Elementalist, should I branch out much into other schools of magic?

Question #3: This recent death was kind of irritating because I couldn’t figure out a particular thing that I did wrong. Does anyone have any insight into how I could’ve done better?

Nephiyo · July 28, 2026, 12:20pm

you’re playing pretty fast! personally I don’t even reach lair in 2 hours in many runs :D and I finish most of my runs in ~9 hours. it is fine to play faster if you know what you’re doing, like the world record streaker finishes their runs in 5 hours on average because they’re much better than I am at the game and have already made all the connections to quickly tell what they should and shouldn’t do. People who don’t mind sacrificing their win % to spend much less time on their runs even finish in under 2 but I don’t learn from them. I try to take as much time as necessary to find the most correct line of play in all situations, and only speed up when the encounters are trivial to handle and I have spare resources to pay for my laziness with. I don’t always manage keep that discipline and I still rush a bit from time to time, but the attempt is made x)
so I’m not telling you you should play 9 hour long runs, but I’d recommend to at least slow down a little, take some time to correctly assess if the monsters you’re seeing are dangerous or not, mind your noise, don’t get too lazy on pulling enemies back into explored territory or positioning and all that

unfortunately from the morgue we can’t tell how the turns leading up to your death looked like, we only see what the screen was on the turn you died and a short message history leading up to it, but you can’t deduce that much from it. IMO crawl is a somewhat (somewhat) shallow game strategically, it’s a tactical game first and foremost. It’s not really about knowing which god to pick, which gear or spells to choose to use, how to skill, where to use your enchantment scrolls, how to identify things (TBF that one is probably the deepest strategic element in crawl though) and all that. there is a small learning curve to the strategy, but the ceiling is not high and once you’ve figured it out you almost always know what to do, you only deviate from the “good play” for the sake of variety, for fun. Crawl is the kind of game where strategic/build variety comes not from the need to adapt to the situation, but from you doing whatever you feel like doing even if it’s less likely to succeed than the ol’ standard road you’ve walked a dozen times already. You even get to choose how easy or difficult the run is going to be before it begins. Crawl doesn’t force you to do things differently, it just gives you options so you can do it differently if you wish to. The real difficulty is tactical. It’s about not pressing one movement key or tab too many when something comes into view, being patient enough to pull enemies back into explored territory one by one (not without use of tt tt tt tt tt tt tt tt tt tt sometimes, I don’t like you, elf:3 end vault). Being especially rigorous with pulling things back if you’re making a lot of noise. It’s evaluating the threat of whatever comes onto screen correctly, deciding if it’s okay to simply tab the thing, or if you can (or if you even should) run from it and how, or what’s the cheapest way to kill it is with your wands, potions, scrolls and god abilities. Not sleeping on the wheel as you’re clearing the snooze middle and endgame. All of those micro things are what crawl is really about, I think

anyway sorry for the last paragraph it’s not actually fully relevant to the topic here, I just wanted to say it somewhere because this game’s nature is something I think about often xD my point was supposed to be that the morgue doesn’t fully have the most important information about the way you died. The ending screen and the last messages do look pretty grim, but it’s hard to say what you did wrong there. Can only say that you didn’t do enough things right, you wouldn’t have died otherwise. I see that you have zero scrolls of blinking, fear or even teleportation, that must mean you understand to use those. But having zero of them also speaks that maybe you played sloppily pretty often. It’s correct to use them to prevent chance of death if there is one, but you should try your best to prevent getting yourself into situations where you have to use important consumables in the first place if you don’t have a decent bank of them accumulated already. If you don’t have at least 3 scrolls of blinking, slow down and make sure you don’t have to use them until you get more. Be especially slow and very very very careful if you have none

the morgue doesn’t tell us (or maybe I don’t see it) what buffs you had when you died, but if you didn’t use at least finesse to deal with a dangerous pack of spiders then you must’ve misevaluated the danger of the situation. If that doesn’t seem like enough, then you can also consider potion of haste (to make non-swinging your axe actions fast), might, invisibility (if the spiders can’t see invisible, I don’t remember if they can but I think they can’t, you can always quickly check in-game with xv), even lignification + torment. There are things you can do, you just have to know if you need to bring out your nukes or if it’s okay to tab the monsters

anyway, we don’t have information in the morgue to tell what you did wrong tactically speaking, but we can safely say not enough right xD you can always assume that when you die, there’s something you could’ve done better. It’s a useful assumption to have, because it makes you think about what you can improve upon and it’s usually true anyway. Of course in 0.1% of cases it’s not true, but until you’re a top player probably don’t give yourself the privilige of ruling what runs were winnable

there are some stategic mistakes

use your scrolls of enchant armor, you get 1 AC for each enchantment level on your gear. Generally the best targets to use them on are cloaks, helmets, boots and gloves, because they have a low maximum enchantment level and are a bit less likely to be replaced soon than body armor. It can be a waste sometimes as you’ll later find gear for the same slots that have brands and would have been better to be enchanted, but don’t worry about it IMO. Prioritize living and not having to use more valuable consumables now than your enchantments in the endgame

using automatic skilling instead of manual. Auto skilling is just not very smart and doesn’t train you optimally. You should begin learning how to manage your skills manually. In this particular run the skills weren’t that egrigious, but I think you had a bit much of armor and too little throwing. Evocations are also something you should train a little of in general, at least 6 by the end of lair. Shapeshifting made no sense

the jewelry you were wearing was slightly dubious. The ring of protection is completely fine but the ring of willpower wasn’t doing a lot in your spot, it would be better to be wearing the ring of slay +4 or strength +6. The will ring is good to have, you just don’t need willpower in spider nest. Keep it as a swap ring for when you need willpower (either immediately or you know the area you’re clearing has many will-checks), swapping rings only takes one turn. Also I think I’d prefer to be wearing the amulet of regeneration over that artifact with guardian spirit. Regeneration goes well with high armor, and guardian spirit may deny you god abilities if you get to low hp

smaller one - don’t willingly pop potions of mutation before you know what type of orb you will have in zot. Those potions are important to undo the mutilation orbs of fire do to you, but if you know you have another type of orb then it’s probably okay

I think this is all I have to comment on, I hope I don’t come off as overly critical, I really just want to help and give a good answer to the (third) question asked xD so hope that it does help!

Argyros · July 28, 2026, 12:37pm

You definitely don’t come off as too critical! That’s what I wanted!

Those are a bunch of really helpful tips, thank you. RE rings: is there a general priority ranking for which to use? Not having enough willpower always makes me nervous.

Didn’t know that about mutation, that’s really helpful.

I should probably use tt more, too, I wasn’t doing that and kind of going quickly…

Nephiyo · July 28, 2026, 2:04pm

I actually can’t quite answer this one xD for some characters it’s much more clear, like casters love int and wizardry rings and if they have to wear those to be able to reliably cast their spells then there’s less choice to be had. Dex-scaling characters (mostly users of ranged weapons or long blades) get a lot of value from dex. Str fighters are the least clear, I don’t know 100% if I prefer to use protection, evasion, strength or slay on one. Maybe it’s a “you can’t go wrong with any” situation, probably lack of experience on my end to an extent. I’d probably use rings of slay, I feel that’s probably the strongest, but I can’t defend the choice

also on the caster questions:
deep elf is not as easy to play as it may seem. They are kind of like the glass cannon species, they have huge int and aptitudes for magic skills and can get high level spells online much sooner than other species can and then use their innate mp regeneration to sustain their use, but they have 20% less max hp than average and a bad fighting apt. Double edged sword, not necessarily easy, but can be very fun! I’d recommend draconian or gargoyle for an easier caster run, just because they can get decent defenses without having to use heavy body armor

branch out yes, if you need to. some starting books have weaknesses and can’t handle every kind of enemy you come across, so you’ll have to find a solution against them outside the starting book. Like alchemists, they kill things by poisoning them, and if the thing can’t be poisoned then they better have a solution to that problen. It can help to think of spells as solutions to different problems, for example stone arrow provides decent damage against single targets and is very convenient to quiver, mystic blast answers “what if they for some reason manage to get in my face?”, mephitic cloud can cheese many things that aren’t immune to it, especially those that have spells and abilities that are dangerous but will be disabled when confused, fulminant prism can fulfil your AoE needs. If there are problems that your current spells don’t have an answer for yet and there’s a spell in your library that does it then consider getting it

the other reason to branch out is that you don’t necessarily get endgame-viable spells of the same school that you start with. Fire elementalists won’t always see their plasma beams or something like that. Pivoting to different schools for the later portions of the run is pretty common, especially for alchemists and necromancers I think. Look at your library regularly and look at what good high level spells if you have in there and if there’s a “path” of intermediate spells of its school(s) that you may use before the big one is online. Deciding when to pivot schools is a bit of a balance, you shouldn’t do it immediately because you may find spells that form a better “path” or even spells that align with your current schools soon after, but you also shouldn’t wait forever because you may not live to see those either. If it’s not already clear, I think it’s generally a good idea to wait at least until you see the shops in orc caves, it’s likely-ish to see spellbooks for sale there, and once you see what spells are offered there (if any) you may make a more informed decision

I think the difficulty of the different mage starts is about how self-sufficient and straightforward they are. Conjurers have no straight weaknesses so they are probably easiest. fire elementalists have large and pretty efficient damage, nothing in the early game resists fire (not that many things do in general in a 3 runer), but it is a bit quirkier to play than conjurers. earth elementalists have irresistible big bonk spells, can do some stabbing with passwall and petrify if you want to. they are a bit clumsier in some scenarios due to not having real AoE, but they are still great. Ice elementalists have massive output with all of their “setup spells” on, but cold resistance in monsters is not that uncommon to see, in particular all undead have cold resistance. ozocubu’s armor is often a solution to the cold resistant enemies already (just hit them with a weapon :p), but you probably want to find something for that anyway. The starting book of air elementalists feels strange to me, it gets the job done but it feels like there’s no “juice” in it. They are carried by the airstrike spell and I think they want to find spells that work well where you can’t maximize airstrike’s damage and spells that do real aoe. Alchemists poison monsters and are super effective against those that they can poison and almost helpless against those they can’t. Alchemist is probably where finding solutions outside your starting book becomes a huge theme theme. Necromancer is a weird background, I don’t know how to play them, the best I’ve figured out so far is to simply not be a necromancer and pivot to ranged weapons or other schools, depending on which you find first. I never played hedge wizard yet

take what I say with a grain of salt in general, I’m not actually a great player xD I also haven’t even used every spell in the game yet so my experience is limited, but I think this should still help

petercordia · July 28, 2026, 7:44pm

Question 1, how are various types of cast ranked:

People tipically differentiate “blaster-caster” “summoner” “hexer” and “hybrid”.
Blaster-casters are relatively easy to win. Pick Vehmut or Sif. (Dith, Ash and Ru are strong too, and that’s probably not an exhaustive list, but Vehmut and Sif keep the strategy you need to win simple.) Always lure enemies to stairs (easy with your loud spells). Always retreat when there is any risk at all of running out of mana. Because you’re disciplined by your mana, you’re less tempted to risk your health, and therefore less risk of dying by oopsy.
summoners are easy to win. Train enough fighting. (And dodging.) Pump Int. So long as you have mana, you can always summon some help and use it to separate yourself from the enemies. Sif results in the simplest gameplay.
hexers are weak. But for a player like me, being weak forces me to pay attention, which can increase my winrate. And hexers are fun.
hybrid characters are weak.

Question 2, should you branch out of your spell school:
No*
(*) by which I mean, it depends.

If you’re with Vehmut and you’re a fire mage, and the bastard offers you an ice spell (as he is wont to), don’t learn it. Veh casters especially have trouble running out of spell slots, and a character that learns both ice and fire is always going to be weaker than a specialist.
However, it is worth making sure you have non-elemental damage, which often comes from learning a lot of Conjurations or a bit of Earth. There’s also often random useful spells that you can pick up by learning 5 levels of some random school (like summoning, hexes, ice, etc) which is often worth doing.
And if you’re not with Sif or Vehm, you’ve got a level 5 spell at 1% failure, and the only level 6 and 7 spells you have are in another school, it can often be worth switching school. (“switching school” meaning that now instead of Fire being your highest-trained skill, for the rest of the game it’s actually Air. And you never train Fire again.)

Argyros · July 29, 2026, 3:01am

Interesting! Thank you for the advice, really helpful.

Argyros · July 29, 2026, 3:02am

What are the main advantages of Sif? Ash definitely makes sense, and Vehumet has the obvious advantages. Sif is mostly just a big spell library?

petercordia · July 29, 2026, 6:37am

there’s the mana-over-time that helps you recover when you strayed too far from the safe zone and your mana is getting depleted. And in situations like Vaults:5 and Slime where retreat may be impossible.
there’s the spells, which pretty much guarantee that you don’t need to learn spells outside of the school you’re planning to specialise in at the start of the game.
there’s the amnesia, which means you can afford to learn whatever low-level spells make the early game easier, without gimping your late game.
and there’s the big panic button, which you can use to cast things like Death’s Door or Summon Dragon Swarm (or Discord for lols). (It’s poorly designed imo, but it is situationally strong.)

Argyros · July 29, 2026, 4:52pm

Gotcha. I’ll have to try that.

stoatsoup · July 30, 2026, 11:02am

However, we very much can from checking the ttyrec. Before we get to that, though, I notice OP has no Dodging skill at all (and hence 4 EV), a lot of unspent gold (and some things worth considering in shops like a +2 pair of boots to replace their +0 boots), and 7 scrolls of enchant armour. A bit more EV and AC might have meant squeezing through…

On T:35773.8 Argyros encounters a torpor snail, a jumping spider, and a tarantella. They’re in a 2-space wide corridor with a chokepoint behind them, but the monsters are in a wide open area to the southwest. That said, they do arrange to fight in a chokepoint at the end of the corridor, not in the open area.

They immediately move forward. Now, we’ve all done it, but this isn’t the right move - they are slow and already we have monsters that can confuse and hold. Worse yet, they don’t move diagonally towards the monsters (not that closing down the range is great, but it’s better than spending 1.5 turns moving south, which doesn’t close them down but does move further from the line of retreat). And, as you surmise, they play quite fast.

This move reveals a moth of wrath, which should also be a bit alarming. Next they move west, completing the diagonal move and encountering an entropy weaver. They attack the tarantella, missing, and a ghost moth reveals itself. This is pretty bad too because they will lose any benefit from their amulet of guardian spirit quite quickly and - well, at least they don’t stat drain anymore but they still hit pretty hard.

Second attack, another miss, corroded by the entropy weaver, and the moth of wrath berserks the torpor snail. Third attack, hurting the tarantella, and a second jumping spider appears. By the time the tarantella is dead, a sun moth has appeared, they’re at Corr -6, the ghost moth is meleeing them from somewhere, and a berserk jumping spider is behind them.

Soon enough the jumping spider webs them twice. At 104 HP (and 0 MP), Corr -10, and Held, they consider using a consumable for the first time, looking at scrolls (they have no teleports or blinks, although there was a blink in a shop) then quaffing their last heal wounds. This doesn’t help very much - 87 HP and they look at divine abilities, but of course they have no MP because of the ghost moth. I’m not actually sure what action they do take that turn: on their final turn they have 29 HP, disentangle themselves, and immediately get killed (and re-webbed, to add insult to injury).

So, given the above, I think the main mistake is not to recognise that you can be on full HP and still have a serious problem on your hands, which I think you did at the least from the point where you had a torpor snail, entropy weaver, and moth of wrath visible - certainly when the ghost moth starts eating all your MP. Backing up might have let you control LOS better - the torpor snail only slows you if it sees you, the entropy weaver only corrodes you if it can see you, and the moth of wrath only berserks things if it can see you.

As soon as the ghost moth started to drain MP you should have considered Heroism and/or Finesse. You’re kind of short of consumables (which is good, you can’t take them with you) but a might or even a haste would have been a boon here - or the humble scroll of fog would have restricted LOS more - or even a Hail Mary berserk rage might have carried you through. And, of course, buying that scroll of blinking might well have let you escape.

I’m afraid the thing not to do is to very quickly make a series of melee attacks and then think about what to do next.

Nephiyo · July 30, 2026, 5:15pm

50% of winning at DCSS is having the patience to do the right thing even when it comes at a cost of real time :x

erf · August 1, 2026, 11:48am

IMO crawl is a somewhat (somewhat) shallow game strategically

hybrid characters are weak

These are probably both true, but as someone who hybridizes characters because it makes for more interesting strategic choices, I thought I should weigh in here.

Question 1 (spell caster difficulty):
Crawl is about doing two things: killing and not dieing. Different species/backgrounds have a different mix of those two, and you’ll generally have the easiest start if you pick something that’s good at both.

Some examples:
Conjurer is great at killing and ok at not dieing once you get Mystic Blast online (level 4).
Fire Elementalist is great at both at level 1 (foxfire is kinda OP), but the rest of their starting book sacrifices not dying for more killing
Hedge wizard has a solid set of spells for not dying, but killing power is low
Deep Elves have killing set to 11, but have not-dieing set to -11
Gargoyles, Draconians, and Oni are above-average at not dieing
Fast races also tend to be good at not dieing (including spriggans, though their low HP makes sudden death a very real risk.)

Gargoyle Conjurer is (IMO) the easiest combo: conjurer basically solves the “kill dudes” problem for the whole game, and gargolye gives you strong defense even with no armor at all. (Admittedly I can’t achieve a win-streak with them, but that’s a player skill issue, not a problem with the combo.)

That said, since I am a hybrid player, I will throw this curve ball at you: brigand can be a strong start for a future spellcaster. Stealth, dex, and dodge are the core defenses for any spellcaster, so you get a head-start on your defenses, powerful offensive consumables, and an enchanted dagger as a fallback weapon. In exchange, you’ll have to sneak and stab your way to your first spellbook and what your actual build looks like will be entirely RNG dependent.

Question 2 (should you pivot):
I personally always do so. I don’t know if that’s actually good, but it’s where I find my fun. For fire elementalist specifically, I treat foxfire as an auto-win button that will eventually fail, and immediately look for something else to pivot into. Found a good sling/bow? Guess I’m a ranger now. Conjurations book? Kill-dudes is set for the game. Alchemy/Summons/Necromancy? Get a enough levels to cast one or two spells, then pump spellcasting so it’s easier to pivot again when I need to.

Strong aptitudes can make pivoting to a new school easier, so a Spriggan learning translocation is trivially simple, while a Gargoyle trying to learn air spells is fighting an uphill battle.

Regarding multiple schools, your spellpower is based on the average of the related schools (plus one-fourth of your spellcasting). This means dual-schools can make it easier or harder to pivot, depending on whether you’ve already trained one of the schools: if you’re starting at zero in both, you’ve got twice as much work to do compared to a single school. This is a major factor in why I personally prefer strong, single-school spells as late-game pivot targets (Ignition is the best example of this.) It is also why I tend to over-train spellcasting: it’s only 1/4 as effective, but guarenteed to be useful regardless of what spells I find (intelligence also affects all schools, so I pritize Int boosts for the same reason.)

For pivots that completely change archetype (ie. fire elementalist to ranger), the decision is a much more complex one, with aptitudes, stats, god choice, and more all having an impact on how difficult (and rewarding) it is to make that change. This is where I find the most fun - though it’s one of the many things I do that makes my winrate cry.

Question 3 (what went wrong):
I don’t disagree with any of the advice you’ve been given up until now, but I do want to point out some strategic issues I see with your build:

You’re using FrostBite. That thing is a death trap: *Noise and 2-handed each require playing extra carefully, and that weapon has both.
You’re using a 2-handed weapon with 4 EV. That is terribly dangerous - there’s so many bad effects that you need to block or dodge, and you can’t do either. Considering you’ve already got 12 points in shields, swapping to a 1-hander with a shield is probably the best upgrade you could give this character.
Your body armor is +1 plate mail. That’s… not great. If that’s the best you’ve got by the time you’re doing a rune branch, search (ctrl+F) for “armor” and try to find something with better AC. Chances are good you’ve overlooked a lighter armor with high enough enchantment to be an upgrade. (On that note: using that +2 cloak with the swamp dragon armor is 1 less AC but gives rCorr. Considering entropy weavers, it might be a better choice for spiders.)
Spirit and infusion are anti-synergy. You’re effectively sacrificing your HP to deal (a tiny bit) more damage.
In this situation, I’d personally do a floor or two of Vaults before doing the 2nd S branch. You don’t have rPois on any of your core gear so you’re effectively handicapping yourself, and a floor of Vaults will yield much more XP.
@stoatsoup already pointed out automatic skilling, but I’d like to particularly point out how you have barely trained throwing. Setting throwing to 100% focus for a floor or two would give you a reliable way to hurt enemies at a distance, making them come to you rather than wasting turns (and getting hit) trying to run towards them.
You have zero blink scrolls, zero teleport scrolls, and the spell Blink unmemorized in your spell library. Learning that spell was basically your only way to escape at that point, and it’s so low level that it should be trivial to get enough XP to cast it safely.
As a tie-in to that: translocations is probably the best spell school for a melee glass cannon to learn. IIRC, I won with a character using a (normally ill-advised) heavy triple-sword by using translocation spells to close on enemies and execute them before they could react. (Although at the end of the game I had Manifold Assault which let me execute enemies anywhere on the screen.)
One last note on skilling: as your skill level increases, it takes a lot more XP to gain another skill level. IIRC, it takes the same amount of XP to go from skill level 20 to 21 as it takes to go from skill level 1 to 18 (or something like that - it’s been a while since I looked up the numbers.) As a result, it’s often better to get a new skill up to level 10 than put that XP into a skill that’s already over 20.

Hopefully these thoughts are useful. As I sorta alluded to here, I’m not great and my playstyle is pretty weird, but hopefully this is good feedback nonetheless.