Armor for Hybrids

[MainiacJoe]

So far I've had success with heavy-armor melee (HOAK and MiBe) and with no armor casters (SpEn and GrEE though the latter doesn't really count as no armor). I've not done well with hybrid characters, and part of it is getting the armor right. I don't know whether I ought to be putting skill into Armor or Spellcasting or spell skills to best overcome armor penalties to spell success. I don't even know whether I ought to be choosing to gain INT or STR! When I get past the early game, I'm dying in late mid-game sometimes to poor defenses and sometimes to poor offense (and often to "learning experiences" that have nothing to do with either).

If I want to aim for a character that is melee with support spells, what armor and skills ought I shoot for?

[radinms]

Worship Chei and wear Gold Dragon Armour.

[greedo]

Dying in midgame I find is usually about poor god choice or lack of defense. If your character has a strong book or setp (caster with mephitic, or miniotaur melee based), you have freedom to do whatever. Otherwise you should choose a god that will help in the beginning of the game.

Hybrids are generally best with high dodging annd fighting aptitude and very light armor, with some kind of dragon armor as end game. Some good hybrid races would be merfolk, halfing or something. General rule of thumb for hybrids is never go strength, and consider dex only in early game.

Edit: And yeah if you plan on heavy as a hybrid go chei wouldn't recommend it

As far as spell skills, I would focus on the early game spells only for a long time as a hybrid. Generally speaking, mephitic cloud, swiftness, repel missiles, low level summons, spectral weapon, animate skeleton, confuse, blink, conjure flame, regeneration, cause fear, things like that. Primarily meph, con flame, repel and summons. Get one or two of these, and then focus on defense and weapon skills.

[radinms]

Chei is highly recommended if go hybrid whether or not you wear light or heavy armour.

[Berder]

Armor on hybrids:

• Use whatever armor you find that lets you cast your spells, until you have your endgame armor. Leather or ring mail is fine. Artifacts may be nice.
• You probably want Fire Dragon Armor for your endgame armor. nothing heavier. In FDA you can cast a level 7 spell such as Orb of Destruction, but probably don't want to go for level 8s.
• Have like 10-15 STR (compare your STR to the armor's encumbrance rating). STR does a lot to help reduce the spell success penalty of armor.
• Training armor skill does little for spell success. Instead armor skill mostly gets you AC. Train spell skills if you want to improve spell success, or wear wizardry.
• Get a lot of dodging as your major defense. Also get nice defensive spells: repel missiles/deflect missiles, regeneration, shroud of golubria, phase shift, ozocubu's armor.

[tedric]

re: mid-game problems, a major hurdle for me was learning not to rush to "sexy" higher-level spells just because they are available (whether through book finds or Veh/Sif). I find I have better survivability when I let my starting spellbook do the heavy lifting until I have decent defense (through Dodging and Fighting, plus 5 Shields when I find a buckler). Stealth is also a crucial defensive skill, and I try to have at least 5 levels by Lair/Orc and bring it up to around 10 before doing Lair branches.

Making defense a priority going into the mid-game can mean thinking a little bit more creatively about how to get the most effective use out of low-level spells, and using consumables/evokables/abilities liberally to fill in any tactical gaps. I'm currently running a HEWz^Sif who didn't learn any spells higher than level 3 until I had finished Orc and Lair, and never felt underpowered -- though that was certainly made easier by finding an early rod of clouds and short sword of elec (along with a handful of enchant scrolls), it's mostly due to the fact that I adapted my tactics to the spells I had on hand, and took time to develop solid defensive & melee skills.

[Sandman25]

I am not sure stealth is good for hybrid unless you are a stabber. Hybrid is already XP-starved and it does not mind running out of MP. Hybrid does not mean you have defense, hybrid means you can kill dangerous monsters without MP.

[tedric]

A little bit of stealth is great for any character that's not wearing heavy armour, especially in the early game. It's not about stabbing, it's about being able to back away from things you don't want to fight right now.

[MainiacJoe]

I have already discovered through experience that, at my current skill level with Crawl, XP invested in Stealth early improves my survivability a great deal regardless of character type. I recognize that as I get better at the game I'll be able to recoup those XP and spend them more efficiently, but for now at least, I'm almost always turning on Stealth for a while at the beginning unless I find early plate mail on a melee dude.

[Sandman25]

A little bit of stealth is great for any character that's not wearing heavy armour, especially in the early game. It's not about stabbing, it's about being able to back away from things you don't want to fight right now.

Yes, I know how stealth works
I train stealth for Na early game but I avoid it with normal speed hybrids, early game I need attack skills and later I need to hybridize which is dangerous so I cannot afford wasting XP. I do train stealth all the game with casters when they can't become hybrids (Op/Fe is typical here).

[mps]

Worship Chei and wear Gold Dragon Armour.

A wise man once said: "Get big or go home."

In all seriousness, though, compromising on defense for spells is bad unless the spells are summons. If you want a "hybrid," get two runes and either FDA, storm DA, or shadow DA, then learn some good level 5 or 6 spells -- i.e. stuff that's not conjurations.

[Utis]

I train stealth for Na early game but I avoid it with normal speed hybrids, early game I need attack skills and later I need to hybridize which is dangerous so I cannot afford wasting XP. I do train stealth all the game with casters when they can't become hybrids (Op/Fe is typical here).

You can get decent AC even with MDA, in which case you can have decent stealth (for the purpose of picking fights) already for as little as 6-10 stealth skill. I. e. depending on race, practically nothing by late mid game. It's a safety investment that is in my opinion well worth it for chars that wear armour with low encumbrance. I don't think I would want to raise stealth in the early game, though.

Otherwise what Berder wrote. Except that I like to have AC and EV somewhat close together (modulo what's reasonable according to armour finds, spells I want to cast etc.), so I tend to emphasize dodging only slightly over armour skill.

In all seriousness, though, compromising on defense for spells is bad unless the spells are summons. If you want a "hybrid," get two runes and either FDA, storm DA, or shadow DA, then learn some good level 5 or 6 spells -- i.e. stuff that's not conjurations.

At higher skill levels you get diminishing returns for your XP investment. If you diversify your portfolio wisely, you might deal oh-so-very-slightly less damage per aut and receive oh-so-very-slightly more. But you gain a lot of options and, depending on what exactly you do, you are both safer and deal more damge over all. The archetypical case would be a character with Freezing Cloud and strong melee. But situationally I wouldn't even rule out Bolt spells, e. g. if you find good armour and a good weapon (and thus can afford to invest in spell schools) but few wands of fire/cold.

That makes sense already pre-lair or even early D. A melee-heavy wizard in ring-mail won't be as good a brawler as a fighter, nor won't he be as good a caster as one in a robe. But, boy, he's an meleer with a lot of options!

I honestly regard so-called "hybrids" as the default case for book starts. I mean, unless you want lvl 9 dual school conjurations castable in a 3-runer, you will probably get strong melee anyways at some point. If floor finds make that reasonable, you might as well start early.

[Sandman25]

Yes, I train stealth when training it becomes more efficient than everything else (spells, fighting, dodging, weapon at min delay). I love stealth in extended, it's awesome there.

[Utis]

Yes, I train stealth when training it becomes more efficient than everything else (spells, fighting, dodging, weapon at min delay). I love stealth in extended, it's awesome there.

Ah, sorry. I see that I missed that you were talking about early game. I agree with you, then.

[mps]

Yes, diminishing returns on physical skills tends to start after 2 runes. That's what spells like phase shift and shadow creatures are there for. Cloud spells are conjurations, hence bad.

[Utis]

Yes, diminishing returns on physical skills tends to start after 2 runes.

Let me rephrase my original statement: The higher your skill, e.g. your weapon skill, the less you gain in terms of survivabilty for every skill point. That starts early: Raising your weapon skill from 3 to 5 is huge. From 8 to 10? While still very good, not quite so huge. In addition to that the XP investment you need is much higher. If you use your XP instead to get e. g. Mephitic Cloud to ~25% fail rate, you are somewhat worse in most melee situations, but on the other hand you suddenly gained an escape option and have an edge when meleing certain foes.

I'm not saying something silly and stupid like: "Dude, you should always do that, whenever!!1 It's an epic strategy, whenever!!1". It depends a lot on aptitudes and what kind of equipment and consumables you find. I might do it with a Tengu, I won't do it with a Troll. For anything in between it depends on what I find. Excellent equipment means I can keep my edge in melee easier and can afford to diversify. On the other hand, lots of the right consumables means there might be little gained in doing so.

Cloud spells are conjurations, hence bad.

Uhm, so ... "Conjurations are bad for melee chars." is literally an AXIOM for you rather than something to be argued for and against? I'm sure that makes a lot of discussions easier for you.

[Sandman25]

I wonder about Spellforged Servitor when you learned something like Airstrike and the rest of spells are Summonings so Spellforged Servitor does not cause any problems to your summons. Is Spellforged Servitor good because it is Summonings or is it bad because it is Conjurations?

[Hurkyl]

Raising your weapon skill from 3 to 5 is huge. From 8 to 10? While still very good, not quite so huge.

You should check your assumptions. The most important effect of training weapon skill is usually the reduction in delay, and the relative benefit of that usually increases as you get closer to minimum delay. Wielding a +0 flail against a yak, raising skill from 3 to 5 might get 10% extra damage per aut, but 8 to 10 is more like 17% extra damage.

[Utis]

Raising your weapon skill from 3 to 5 is huge. From 8 to 10? While still very good, not quite so huge.

You should check your assumptions. The most important effect of training weapon skill is usually the reduction in delay, and the relative benefit of that usually increases as you get closer to minimum delay. Wielding a +0 flail against a yak, raising skill from 3 to 5 might get 10% extra damage per aut, but 8 to 10 is more like 17% extra damage.

No offense, but speaking about assumptions: You're assuming that I was talking about raw damage output. What I actually wrote was (emphasis added):

The higher your skill, e.g. your weapon skill, the less you gain in terms of survivabilty for every skill point. That starts early: Raising your weapon skill from 3 to 5 is huge. From 8 to 10? While still very good, not quite so huge.

This is about staying ahead of the killing curve while at the same time adding options for surviving or avoiding threats above the current average threat level. In other words: It is not about raw damage output in the abstract, it is about what helps a character at a concrete point of dungeon exploration. You need need to be able to kill the majority of monsters you encounter at your particular point of dungeon exploration, while never significantly draining your health bar. Everything else you need to be able to avoid.

Case in point: At the time ("time" here as a function of "what threat level you have to cope with") you raise your weapon skill from 3 to 5 on a melee heavy char, you are still dealing mostly with kobolds, hobgoblins, adders -- all of which need to not affect your HP bar at all at this point. When you encounter adders for the first time, they're quite dangerous -- with deadly spikes in the damage they deal to you. They need to get to "not very dangerous" status fast where unlucky spikes are just a nuisance.

On the other hand, when you encounter Yaks, that 17% (going with the numbers you provided) increase in damage output is meaningless if even under favourable tactical circumstances you don't have enough staying power for the encounter. If a Yak is an out of depth monster at the time you encounter it, you avoid it. If Yak packs are the normal threat level at the time you're attacking with a +0 flail at weapon skill 10 (and melee is your main offense), you quit and start a new character.

(That being said, I doubt that 17% is a realistic number. Oh, I didn't check, but I trust that this is the number you get from fsim if you raise only weapon skill. But in real dungeon exploration you would also raise fighting, in which case the the difference gained by raising from 8 to 10 is smaller. That however would raise a discussion of how much fighting one should train, which I'd rather avoid. The point I'm trying to make here is valid even in the numerical worst case.)

I have to emphasis again that I'm not talking about an abstract recipe that's supposed to apply to every situation. Raising weapon skill from 8 to 10 might be the only reasonable thing to do for a particular character at a particular point of dungeon exploration. Even if this just means (I'm pulling the numbers out of my but here) that encounters at the top tier of your current threat level leave you on average at 80% of your HP rather than on average 75%, this is already an important gain! I meant what I wrote: If you decide to diversify, you do become worse in melee. If you do make that decision, the option gained needs to significantly affect your chance of survival, otherwise you're crippling your character. (This is why the general advise to newbies to stick with your main offense is fundamentally sound!)

The reason that accepting this kind of trade-off can be worth it for particular characters at particular points of dungeon exploration is the time-honored principle of "Don't put all your eggs in one basket!" As we all know, the way randomness works in crawl, there are huge spikes from the average. So it's reasonable to do what every sound company and every good poker player does: You diversify and spread your risk. Where and when it makes sense, of course. If you encounter a pack of blink frogs, a 17% increase in damage output is meaningless, unless you could have defeated them already in the first place (though with a more depleted HP bar). If you don't, having Mephitic Cloud means you might survive, even if that just means you can escape. Or, in late game: A melee character with Freezing Cloud is not as good in melee as one of the same race-background-god combination who wears heavy armour an raises only melee skills. But even the latter won't always be able to avoid fighting multiple monsters at the same time, in which case the former does more damage per aut. The latter will on average receive significantly less damage per aut, but the encounter will take longer and he might get unlucky and have a huge spike in damage received.

Again, it all depends on aptitudes, dungeon finds and the decisions you made up to that point.

EDIT:
tl;dr: Reasoning about the relation between skill and damage in the abstract is limited. What's matters for this particular point of discussion is the relation to the current threat level. Diversification does make your character worse in your main offence. If done incorrectly, it can be crippling. If done well, it can help spreading the risks and increase chance of survival. There is no abstract recipe for this, it all depends.

[gammafunk]

OP, Berder's list above is the basic idea for how you make a melee character that also casts damaging magic spells. You can disregard any useless meme posts in this thread of "you can't use conjurations or damaging magic depending on spell power." You also don't at all need Cheibriados to do this type of character well, although that's certainly something you can do if you don't mind the way Chei works in terms of dying randomly. It should be said that to many of us, "melee with support spells" means you actually don't cast damaging spells and instead use charm-type spells that are mostly low level, in which case you do wear the heaviest armour available.

If you're beginning these chars as a mage, you'll want to prioritize your offensive magic and your basic defenses (fighting and dodging) at first and then move into using a melee weapon reasonably fast. It sounds like you're already getting this far, so your poor defenses later might come from not having at least 15 dodging or from not taking any armour upgrade from your starting robe. You shouldn't be taking more Str than what you need to deal with the ER of your armour, and if you only cast summons or animated remains spells you don't need a lot of Int.

For damaging magic, pure conjurations spells can be nice since they're irresistible and scale well up to OOD, and cloud spells are good since they don't depend too highly on spell power. Nothing prevents you from using e.g. elemental bolt spells since you can e.g. get 8-10 in the elemental school and save putting more than 10 levels for conjurations. An enhancer like a ring of ice/fire or a stave as a swap can be very helpful to get decent spell power there. Just try to avoid getting those 8-10 levels in many schools for purposes of casting damaging spells as this really starts to eat into what you need for good fighting/dodging/armour.

[MainiacJoe]

It should be said that to many of us, "melee with support spells" means you actually don't cast damaging spells and instead use charm-type spells that are mostly low level

To be clear, this is what I had in mind with the OP.

in which case you do wear the heaviest armour available.

i.e. plate, storm dragon, etc.? My first thought is, my spell success for even lvl1 spells is abysmal! But perhaps I just need to be cognizant that it will simply take much more XP in spell skills before I begin to see an effect from that training, compared to robes and leather.

If you're beginning these chars as a mage

Typically lately I've been tryng to start as melee and move into spells, instead of the other way around. I simply have better survivability in the early game that way. What would you suggest for that gameplan?

You shouldn't be taking more Str than what you need to deal with the ER of your armour

I know the AEVP calculation as a function of EVP, STR and Armour skill, but what value should I be aiming for?

[ThreeInvisibleDucks]

Unless one is playing a high-strength character, ER 11 seems to be the breakpoint after which the spellcasting penalties start to ramp up quickly.

I generated the following tables to get a better grip on how much more spell training one actually needs to cast in heavier armour: viewtopic.php?f=5&t=17049

[Hurkyl]

This is about staying ahead of the killing curve while at the same time adding options for surviving or avoiding threats above the current average threat level.

And training two levels of skill does a lot more to get there for a character with 8 skill who is at/behind the killing curve than a character with 3 skill at/behind the killing curve.

It's easy to spin the scenario to conclude anything you want here. e.g. raising a weapon skill from 3 to 5 has a negligible impact on survivability for a character who is already ahead of the killing curve because of conjurations.

I'm not trying to refute your point about diversification; I more or less agree with it. What I'm trying to do is to oppose the propagation of a misconception about physical skills. Raising your Fighting skill from 8 to 10 has less of a relative impact on how long you last in a fight than raising fighting from 3 to 5 (the absolute impact is the same, of course). Depending on your current AC number, raising your armor skill from 8 to 10 might have slightly more or slightly less impact on the rate you receive damage than raising it from 3 to 5. A lot of people seem to believe the same thing about weapon skill, but in reality raising weapon skill from 8 to 10 has a much larger relative impact on the rate you deal damage than raising it from 3 to 5.

(I'm ignoring rounding in the above)

[Utis]

This is about staying ahead of the killing curve while at the same time adding options for surviving or avoiding threats above the current average threat level.

And training two levels of skill does a lot more to get there for a character with 8 skill who is at/behind the killing curve than a character with 3 skill at/behind the killing curve.

It's easy to spin the scenario to conclude anything you want here. e.g. raising a weapon skill from 3 to 5 has a negligible impact on survivability for a character who is already ahead of the killing curve because of conjurations.

True.

I'm not trying to refute your point about diversification; I more or less agree with it. What I'm trying to do is to oppose the propagation of a misconception about physical skills. Raising your Fighting skill from 8 to 10 has less of a relative impact on how long you last in a fight than raising fighting from 3 to 5 (the absolute impact is the same, of course). Depending on your current AC number, raising your armor skill from 8 to 10 might have slightly more or slightly less impact on the rate you receive damage than raising it from 3 to 5. A lot of people seem to believe the same thing about weapon skill, but in reality raising weapon skill from 8 to 10 has a much larger relative impact on the rate you deal damage than raising it from 3 to 5.

Fair enough. My apologies if my comments seemed to encourage misconceptions about weapon skill. I was about to make an attempt at rephrasing what exactly I meant in more statistical terms (relating to the average percentage of player HP saved per fight through raising damage output), but doing so I realized that this is moot without going into the actual numbers (monster HP, monster damage, average player HP at certain dungeon levels etc.). I don't have those and I'm not going to research them any time soon. (I mean: I might be wrong. And finding that out after hours of work would be rather embarrassing.)

Out of curiosity: How does the fighting skill scale in on that damage progression, exactly? I was always under the assumption that an increase in accuracy is more important (cf. the effect of Corona) than a slightly better delay?

EDIT: I'm tired and my brain is not working anymore. What you wrote doesn't affect accuracy and the merits of the fighting skill at all, since you were talking only about the effect of mindelay on the relative scaling of weapon skill. Any benefit from fighting carries over to that as well. (Right?)

[mps]

Yes, diminishing returns on physical skills tends to start after 2 runes.

Let me rephrase my original statement: The higher your skill, e.g. your weapon skill, the less you gain in terms of survivabilty for every skill point. That starts early: Raising your weapon skill from 3 to 5 is huge. From 8 to 10? While still very good, not quite so huge. In addition to that the XP investment you need is much higher. If you use your XP instead to get e. g. Mephitic Cloud to ~25% fail rate, you are somewhat worse in most melee situations, but on the other hand you suddenly gained an escape option and have an edge when meleing certain foes.

I'm not saying something silly and stupid like: "Dude, you should always do that, whenever!!1 It's an epic strategy, whenever!!1". It depends a lot on aptitudes and what kind of equipment and consumables you find. I might do it with a Tengu, I won't do it with a Troll. For anything in between it depends on what I find. Excellent equipment means I can keep my edge in melee easier and can afford to diversify. On the other hand, lots of the right consumables means there might be little gained in doing so.

Cloud spells are conjurations, hence bad.

Uhm, so ... "Conjurations are bad for melee chars." is literally an AXIOM for you rather than something to be argued for and against? I'm sure that makes a lot of discussions easier for you.

It's one of these things where there are people who have an interest in representing that there is a broad and interesting middle ground between conjurations focused characters and melee focused characters. For example, suppose you had friends who were really invested in the idea that crawl is brilliantly designed. Then you probably wouldn't want to say that people who make these kinds of dudes are just making bad characters, but crawl is so easy it doesn't matter. You'd want to say it's a reflection of the infinite variety and depth crawl offers and everyone who says otherwise is bad.

Now you want some kind of argument for why conjurations are bad for melee characters: It's not that complicated.

Melee provides a strong combination of offense and defense (in the form of heavy armor and xp availability for skills like fighting and dodging). These alone provide the simplest and most reliable way to win at crawl.

Magic and armor don't mix well until the endgame when the penalties can be largely ignored because of availability of relatively low ER armor, high str and armour skill, and the large amount of xp available to dump into magic schools. Until then, you're faced with trade-offs in defense, xp consumption, and casting ability. Then the question becomes whether the spells you want will justify the trade-off.

Now you consider conjurations: Their primary advantage is range. The problem is, standard crawl melee tactics make ranged attacks largely unnecessary. They're nice if you have them, but you don't need them often. Notice that "ranged attacks you don't need often" is a niche cheaply filled by evocables, the use of which is not affected by armor. On the other hand, conjurations, particularly the better ones, tend to be some combination of high level and multischool. This means they take a lot of xp to use. With armor, they take even more.

What do conjurations add to your combat effectiveness? The answer is: Not much. Good melee is all you need to kill everything in the game (comfortably, easily, optimally). You get a ranged option you don't actually need. At melee range, there's little to no added value.

That doesn't mean all magic is bad for a melee focused character. Summons have great synergy with melee, for example. The better ones are single-school. Spells that improve your defense or affect your speed are also great. Hexes are good in some cases, though they also have the problem of being overshadowed by evocations.

[radinms]

anyway conjuration is bad unless if you can cast Lv9 magic.
if you want ranged combat, go bow or xbow or throwing. to pick up conjuration is bad unless if your int is high.

[gammafunk]

Out of curiosity: How does the fighting skill scale in on that damage progression, exactly? I was always under the assumption that an increase in accuracy is more important than a slightly better delay?

It has a large impact on damage even if you don't consider accuracy. It's an N/60 increase in damage comparing a skill of N fighting to 0 fighting, so e.g. a 25% increase at 15 fighting, a 45% increase at 27 fighting. Only about a 1.67% increase in damage with each level, but it certainly adds up.

Your reasoning that you increase the flexibility of your character by taking advantage of damaging spells (within reason) is fine. I'd just ignore any posts with silliness like "cloud spells are conjurations so they're bad," especially if they depict developers like they're in some high-school drama conspiring to mislead people into "thinking their game is brilliantly designed." You can win characters whose pre-buff defenses slowly converge to ac+ev of 50-60 very comfortably and easily using some conjurations.

[Sar]

Last time I tried Bows I was very underwhelmed by their low-level performance. I would take a good race IE over a bow Hu anytime, survival-wise. Level 9 spells don't even factor in that.

[mps]

Generally speaking, you want to pay attention to people whose go-to models for human interactions are high school kids who thought he was a loser and the X-Files, especially when such people offer vague and equivocal affirmation of what you want to be true.

[Utis]

One thing that I would like to mention, though: I have no idea of how exactly ranged combate compares to damaging spells with regard to XP investment. I've played characters using either, but I don't know the numbers. Anyways: If winning were my only goal -- it isn't, since I get my kick out of crawl from exploring strategic options, and that secondary goal sometimes conflicts with what would be the better build decision -- but if winning were my only goal, then I wouldn't usually start training skills from 0 in the middle game when I happen to find higher-level conjuration spells like Freezing Cloud, let alone bolt spells.

If I'd do, then because I already invested in the respective skills for other reasons. So when comparing e.g. ranged combat (assuming you find plenty of bolts/javelins of penetration) with e.g. bolt of fire/cold with regard to their skill investment, this needs to be taken into account. The latter are part of an upgrade part that might involve things like Conjure Flame, MephCloud, Throw Icicle, Sticky Flame. (As well as ice beasts and, possibly, Ozocubus's Armour on the elemental side.)

I'm not going to engage in speculations about hidden motives or socio-psychological (if that's the word in English, I'm not a native speaker) mechanisms behind people's reasoning. Anybody who disagrees with me is mind-controlled by the media, anyways.