Confirm or deny these skills and aptitude mechanics +I.D. Qs

[Pollen_Golem]

Reading some old posts by experienced crawlers made me doubt that I understand these things correctly, so please correct if I'm mistaken

1`The wiki is correct about how aptitudes work and scale: http://crawl.chaosforge.org/Aptitude.

2`Spell power and success rate depend on the average of the spell's schools. Whether you have 12 in one, 6 in two, or 4 in all three, it does not matter.

3`Maxing conjurations without touching elemental schools is an unintuitive but valid strategy for blaster casters, partially as a result of the above.

4`Specifically for ogres and two-school spells, training spellcasting (apt +1) is exactly as XP-efficient as training the relevant spell schools (apt -3), for the purpose of success rate and spell power.

5`The amount of XP needed to raise a skill by 1 depends on your XL, but not on how much training the skill has already received. If a 15XL centaur with 0 dodging and 25 armor starts training only one skill, he will reach 1 dodging or 26 armor in exactly the same time.

6`Draining affects each skill separately. An X amount of draining will reduce each skill by X, simply reducing the skill to zero if Skill - Draining < 0.

7`Draining and skill boosts by Ash/Oka/crosstraining do not care about your aptitudes. They affect your skills directly, bypassing the aptitudes. If for some reason you want an ogre to uses axes (apt -3) only, raising maces and flails (apt +3) to 25 and having a 25*40%=10 skill boost to axes is more XP efficient than actually raising axes to 10.

8`This one's not about aptitudes. The recommendation to have str>encumb is a mere rule of thumb. No penalty is completely removed when str=encumb, unlike how when shields hit 15 or 25, nor does any step-down function kick in after str>encumb.

[Elystan]

5 is definitely wrong. As far as I can tell the amount of XP taken to raise skills is static and depends on the skill level not the experience level. For example raising a skill from 20 to 21 takes more XP than raising from 19 to 20 regardless of XL. The extreme example you gave, I'd not be surprised if you could take a skill from 0 to 12 with less XP than it takes to take a skill from 26 to 27. I'm not sure of the exact mechanics but you're definitely mistaken there.

Also, as for Ashenzari skill boots - it seems non-linear, and I can't say how exactly. But at low skills you get smallish boosts, at medium skill levels you get high boosts (up to around 4 skill levels) and at very high levels the boosts drop off again in terms of levels - but pertaining to the above, I think the more XP in the skill, the bigger the boost.

As for crosstraining, again I think it's a flat % of the experience invested in one skill gets added to the other - definitely less than 100% in any case. Aptitudes affect this. If you want to train M&F you train M&F, not axes, even if your aptitude is better in axes you'll get less bonus to M&F because of your worse aptitude.

[ajon]

I might be wrong on some of these but Ill take a shot.

1- Im too lazy to read it
2 - Yes
3 - Nuanced I suppose but yet, raising conjurations will raise all conjurations You could put tons of Xp into it. It might be cheaper to put less into it though and raise the other skills a little bit each.
4- What do you mean XP efficient? Raising spellcasting has some small effect on success rate I think. It mostly affects hunger cost, raises MP, and gives you more spell levels to train spells with
5- False. As skills increase in level, they cost more.
6- Not sure, I have seen some skills still drained though when others weren't I think
7- Im not sure exactly how it works.
8 - Armour penalty is never completely removed.

[Arrhythmia]

1. Broadly true; haven't closely read the page so there's probably some dumb advice snuck in there, but apts mean it takes 2^(-apt/4) times as much XP to level up a skill.

2. True with an epicycle; spellcasting counts one-fourth as much as any individual school does.

3. False. Because skill costs grow exponentially, it will always be worse to get 12 in Conjurations as opposed to 6 in Fire Magic and 6 in Conjurations.

4. False, see 2.

5. Incredibly false. The amount of XP it takes to raise a skill increases exponentially.

6. False? I'm not quite sure what you're saying here, but my understanding is that draining hurts higher level skills more than lower level ones, i.e., you may drop from 20 to 18 in some skill, but only from 10 to 9 in another. The part about there being a floor at 0 is true though.

7. Okay, so, two things here. The first part is true; the skill boosts by Oka definitely doesn't care about aptitude, the skill boost by Ash probably doesn't but I'm not sure. The second part is false though; these days it will always be more effective to directly train a skill than by cross training.

8. True.

[Siegurt]

The amount of xp it takes to raise a skill increases *both* as the skill increases level, *and* as you train skills in general.

2 is correct, but recall that spells use different number of schools (from 1 to 3) and actually its ( (average of skills*4 )+spellcasting)/5

3. Is misleading, it is efficient to train conjuration only for conjuration only spells, but not for soells which use another school

4. Is only true if they are currently at the same level (see above)

5 is false

6 is true
7 is true
8 is true ( penalty is 2/5 of er squared, divided by str+3)

[Pollen_Golem]

5 is definitely wrong. As far as I can tell the amount of XP taken to raise skills is static and depends on the skill level not the experience level. For example raising a skill from 20 to 21 takes more XP than raising from 19 to 20 regardless of XL. The extreme example you gave, I'd not be surprised if you could take a skill from 0 to 12 with less XP than it takes to take a skill from 26 to 27. I'm not sure of the exact mechanics but you're definitely mistaken there.

So a medium-high level character who slays many high-HD monsters can super-easily get all of his skills to some low level like 4 or 5?

Also, as for Ashenzari skill boots - it seems non-linear, and I can't say how exactly. But at low skills you get smallish boosts, at medium skill levels you get high boosts (up to around 4 skill levels) and at very high levels the boosts drop off again in terms of levels - but pertaining to the above, I think the more XP in the skill, the bigger the boost.

Ash's boost is adjusted for how well-trained you are already, ok. But, you say, he adjust not for your SKILL but for how much XP you've invested in that skill (which are different things)?

As for crosstraining, again I think it's a flat % of the experience invested in one skill gets added to the other

So in other words crosstraining adjusts for aptitude in both the trained and the boosted skill. Eugh...

6. False? I'm not quite sure what you're saying here, but my understanding is that draining hurts higher level skills more than lower level ones, i.e., you may drop from 20 to 18 in some skill, but only from 10 to 9 in another.

I thought a certain amount of draining will take 27 to 24, 10 to 7, 4 to 1, 2 to 0.

You're saying it will take away 100 invested XP out of every skill, individually, and since it's exponential you will see a greater arithmetic difference among low-trained skills.

PS epicycle?

[CanOfWorms]

Ash gives up to +9 to the relevant skills (the amount is determined by piety/cursed equipment), which is mitigated by the amount you've already invested (the more you invest the weaker the boost is, in 0.14 the penalty was skill/4.) You only get the bonus if the skill is trained to at least 1.

[Hurkyl]

2. True with an epicycle; spellcasting counts one-fourth as much as any individual school does.


4. False, see 2.

You either overlooked the fact an Ogre spends half as many skill points to raise Spellcasting as they would any magic school at the same spell level, or you overlooked that a skill level in Spellcasting is equivalent to 1/4 of a skill level in all spell schools.

[duvessa]

1`The wiki is correct about how aptitudes work and scale: http://crawl.chaosforge.org/Aptitude.

the math stuff on that page is correct, except that 0 isn't the mean aptitude

2`Spell power and success rate depend on the average of the spell's schools. Whether you have 12 in one, 6 in two, or 4 in all three, it does not matter.

this is correct

3`Maxing conjurations without touching elemental schools is an unintuitive but valid strategy for blaster casters, partially as a result of the above.

this is stupid

4`Specifically for ogres and two-school spells, training spellcasting (apt +1) is exactly as XP-efficient as training the relevant spell schools (apt -3), for the purpose of success rate and spell power.

this is wrong because higher skill levels are more expensive

5`The amount of XP needed to raise a skill by 1 depends on your XL, but not on how much training the skill has already received. If a 15XL centaur with 0 dodging and 25 armor starts training only one skill, he will reach 1 dodging or 26 armor in exactly the same time.

this is wrong, higher skill levels are more expensive, and ENORMOUSLY so. skill point cost also increases with experience already earned, but not XL

6`Draining affects each skill separately. An X amount of draining will reduce each skill by X, simply reducing the skill to zero if Skill - Draining < 0.

not sure what the question is here...

7`Draining and skill boosts by Ash/Oka/crosstraining do not care about your aptitudes. They affect your skills directly, bypassing the aptitudes. If for some reason you want an ogre to uses axes (apt -3) only, raising maces and flails (apt +3) to 25 and having a 25*40%=10 skill boost to axes is more XP efficient than actually raising axes to 10.

oka and ash: yes; crosstraining: no, crosstraining uses skill points which care about aptitudes

8`This one's not about aptitudes. The recommendation to have str>encumb is a mere rule of thumb. No penalty is completely removed when str=encumb, unlike how when shields hit 15 or 25, nor does any step-down function kick in after str>encumb.

this is a stupid recommendation and people who make it are stupid

[Pollen_Golem]

The question in 6 is two-part.

When affecting a particular skill, does drain look at any of your other skills?

When affecting a particular skill, does drain look at the aptitude for that skill?

If drain brings a level-10 skill down to 8, does that same drain bring a level-20 skill down to 18, less than 18, or more than 18?

[ajon]

why is the strength to encumbrance rating suggestion stupid?

[duvessa]

because there's nothing special at all about strength equal to encumbrance rating

[HobGoblin]

I was just wondering about cross-training a moment ago.

From what I read, EXPERIENCE is carried over in a percentage to the cross-trained skill.
APTITUDE alters the amount of experience needed to increase a skill.

If this is true, the experience gained and passed on is always the same exact thing, no matter what aptitude any skill has.
The aptitude is then "applied" in that it affects the experience needed to actually raise that skill.

For example, say Aptitude 3 needs 20 exp to increase by 1, Aptitude 0 needs 50. Also, lets say x-training gives 50% exp.
If you put 100 exp into the Apt3, it goes up 5 levels, and say it x-trains the Apt0 at 50%(50exp), then Apt0 gets 50 exp gaining 1 level.
Train 100 into Apt0, it gains 2 levels, and x-trains Apt3 50 exp, gaining it 2.5 levels.

This means, yeah you cannot increase efficiency of a weaker apt skill by cross-training, but you CAN get a high apt skill up nicely without focusing on it.
I wanted to x-train to get a poor aptitude skill high too, but then I realized that would undo a lot of the species' specializations.
Is this the general consensus on how cross-training works?

EDIT: Don't forget the skill has to be the higher one to actually cross-train the other skill. This might lower the possibilities some.
EDIT2:From my testing lower skills do now xtrain higher ones, although I've only tried Slings and Throwing.

[Pollen_Golem]

Welcome HobGoblin, and I'd like someone to confirm that.

[Pollen_Golem]

5`The amount of XP needed to raise a skill by 1 depends on your XL, but not on how much training the skill has already received. If a 15XL centaur with 0 dodging and 25 armor starts training only one skill, he will reach 1 dodging or 26 armor in exactly the same time.

This was actually my most default assumption with the least doubt, LOL.

After all that's how skills work in just about all other RPGs. You level-up and spend 10 or so skill points on whatever skill you want, like the way you add a stat point every 3 levels in Crawl. I clarified a lot for myself by reading the Skill and Skill_point entries in the wiki. It ought to be required reading for all crawlers who want to manipulate anything in the m-menu.

[HobGoblin]

There is a slight problem. I am raising slings, I just raised it from 2.0-2.1, and it cross-trained throwing from 2.4-2.5.
There are multiple places on the wiki that talk about cross-training, but I think they are outdated.
I'm going to stick with my theory until someone says otherwise, but with the exception that now you can cross-train higher
leveled skills with lower ones.

[Sar]

Actually there are plenty of RPGs apart from Crawl that make skills more expensive to raise past certain amount.

[duvessa]

Crosstraining adds 40% of your skill points in one skill to the effective skill points of the skills that it crosstrains. That's it, that's all it does. So if you have 10 skill in a long blades with an aptitude of 0, you have 2800 skill points in it; short blades will get 1120 skill points added to its level; if your aptitude for short blades is also 0, this is enough to take it from level 0 to level 6.198, or from level 10 to level 11.626.

[Pollen_Golem]

Actually there are plenty of RPGs apart from Crawl that make skills more expensive to raise past certain amount.

OK, but I don't know any that hide the extra expense from the player.
For example in Fallout you have to invest 2x or 3x as many skill points to raise a skill by 1, past certain thresholds.

Maybe shields should crosstrain with armor. That would make sense.

[Elystan]

Ash gives up to +9 to the relevant skills (the amount is determined by piety/cursed equipment), which is mitigated by the amount you've already invested (the more you invest the weaker the boost is, in 0.14 the penalty was skill/4.) You only get the bonus if the skill is trained to at least 1.

This isn't my experience. I've seen skills I have no XP in get small boosts (less than a skill level), and then once trained larger boosts - the most I've seen at max piety and fully bound is around 4 or 5 skill levels. It seems to vary non-linearly depending on how much skill you have in it. For example I've seen skill level 1 boosted to about 3.5 and skill level 10 boosted to about 14. To hit a boosted 27 seems to take 24 skill or so, at least with a high aptitude.

I don't know the exact mechanics and I'm happy not to. All I know is he gives large skill boosts and training the skills further is always worthwhile.

[Siegurt]

Actually there are plenty of RPGs apart from Crawl that make skills more expensive to raise past certain amount.

OK, but I don't know any that hide the extra expense from the player.
For example in Fallout you have to invest 2x or 3x as many skill points to raise a skill by 1, past certain thresholds.

Maybe shields should crosstrain with armor. That would make sense.

I have played a LOT of RPGs, generally if the game lets you spend points on raising skills, it's either 1:1 or it's obvious how many points it takes to raise a skill level. Usually if the skills increase in the background as you level (either automatically, according to what you use, or interactively, where you select which skills you are going to focus on) the skills raise exponentially.

Typically in "you spend your skill points" style RPGs, you get skill points at specific levels, and the *levels* take exponenially more XP. (Which is similar it just moves the mechanic elsewhere)

[tedric]

average of the spell's schools. Whether you have 12 in one, 6 in two, or 4 in all three, it does not matter.

2 - Yes

2. True with an epicycle; spellcasting counts one-fourth as much as any individual school does.

2 is correct, but recall that spells use different number of schools (from 1 to 3) and actually its ( (average of skills*4 )+spellcasting)/5

this is correct

apparently none of you know how averages work

[duvessa]

"average" most commonly refers to the arithmetic mean, which is exactly what is used for effective spell skill
the question assumes a three-school spell of course

[Greyr]

Thanks to Duvessa I actually learned something from a thread.

[Pollen_Golem]

if the skills increase in the background

in "you spend your skill points" style

Oh yeah. Crawl awkwardly falls into a "neither of the above" category. I'm guessing this comes from its origins when it was designed to be a quintessentially opaque roguelike. Eventually they got rid of the XP pool and allowed manual training, but the substructure remained the same.

[Siegurt]

if the skills increase in the background

in "you spend your skill points" style

Oh yeah. Crawl awkwardly falls into a "neither of the above" category. I'm guessing this comes from its origins when it was designed to be a quintessentially opaque roguelike. Eventually they got rid of the XP pool and allowed manual training, but the substructure remained the same.

I would put crawl firmly in the "skills increase in the background" category. (The fact that you get to choose which skills will increase doesn't change that.)

The subcategories of "skills increase in the background" include "you choose which skills will increase" and "the skills which will increase are the ones you use" I've never encountered an RPG which uses either of these subcategories where skill increase was linear with XP. (I have encountered a RPG where there was simply an amount of XP directly *in* each skill (there was no mention of a "skill level") however in that case the *effects* took exponentially more XP to achieve, which translates to an exponential XP per skill level, where the skill level just isn't explicit)

[tasonir]

because there's nothing special at all about strength equal to encumbrance rating

It's a pretty curve. But you still have to pick somewhere along the curve to sit at. There's implicit assumptions underneath it, like you don't want your spellcasting to be completely disrupted, a reasonable dodging penalty, etc. If you didn't care about casting or dodging at all, you could wear the heaviest armor with low strength, but most people expect to still be able to cast in medium armors, and having strength >= to ER makes that reasonable.

A fighter who doesn't have 18 strength early on should still put on plate asap in general, but a fighter will probably have 18 strength most of the time. And raising it to 18-20 later on isn't a bad idea if you don't already have it. But you probably don't want to raise it to 25-30, by that point dex or int would have been better (if you're casting, int is probably better sooner, but now we're getting into a "how much do you cast" sort of question).

[dowan]

this is a stupid recommendation and people who make it are stupid

This is a rude and unnecessary statement and people who make it are rude and unnecessary.

Ignorance of a game mechanic does not make a person stupid, and knowledge of a game mechanic does not make one intelligent.

[njvack]

this is a stupid recommendation and people who make it are stupid

This is a rude and unnecessary statement and people who make it are rude and unnecessary.

Ignorance of a game mechanic does not make a person stupid, and knowledge of a game mechanic does not make one intelligent.

This is correct. @duvessa, calling an idea stupid is borderline OK (it bugs me mostly because it adds bad feelings and doesn't add any information), calling a person stupid is not.

[partial]

this is a stupid recommendation and people who make it are stupid

This is a rude and unnecessary statement and people who make it are rude and unnecessary.

Ignorance of a game mechanic does not make a person stupid, and knowledge of a game mechanic does not make one intelligent.

family feud

[Siegurt]

because there's nothing special at all about strength equal to encumbrance rating

It's a pretty curve. But you still have to pick somewhere along the curve to sit at. There's implicit assumptions underneath it, like you don't want your spellcasting to be completely disrupted, a reasonable dodging penalty, etc. If you didn't care about casting or dodging at all, you could wear the heaviest armor with low strength, but most people expect to still be able to cast in medium armors, and having strength >= to ER makes that reasonable.

A fighter who doesn't have 18 strength early on should still put on plate asap in general, but a fighter will probably have 18 strength most of the time. And raising it to 18-20 later on isn't a bad idea if you don't already have it. But you probably don't want to raise it to 25-30, by that point dex or int would have been better (if you're casting, int is probably better sooner, but now we're getting into a "how much do you cast" sort of question).

Here's a link to said curves should anyone be interested:
viewtopic.php?f=5&t=14075&p=195665#p195708

Str >= ER is in the approximate area where I personally feel like more strength isn't worth the investment, plus it's easy to remember. I often tell people to use this value "or a little more on heavier armours" but as you can see, there's not a specific optimal number.

[CanOfWorms]

This isn't my experience. I've seen skills I have no XP in get small boosts (less than a skill level), and then once trained larger boosts - the most I've seen at max piety and fully bound is around 4 or 5 skill levels. It seems to vary non-linearly depending on how much skill you have in it. For example I've seen skill level 1 boosted to about 3.5 and skill level 10 boosted to about 14. To hit a boosted 27 seems to take 24 skill or so, at least with a high aptitude.

I don't know the exact mechanics and I'm happy not to. All I know is he gives large skill boosts and training the skills further is always worthwhile.

I guess that's what I get for reading off 0.14 Ashenzari's data :v I tracked down the most recent skill formula, which gives a steady increase to ~3/~4.5/~5.5 at skill level 5, and then starts dipping down around skill level 9.

[Pollen_Golem]

Got some more questions on ID tips:

11) Weapons that are glowing/runed/enchanted but are in grey font are unbranded. You must have witnessed a monster wielding that weapon because any existing brands are identified on sight when the weapon is wielded by player or monster. Therefore, grey and potentially cursed weapons should be discarded without a second thought. Except rare bases.

12) You should generally not bother with blue-font weapons that belong to a school in which you have no investment and which is not among the better weapon schools for your species (in terms of aptitudes and other factors) except maybe rare bases. There are too many for you to ID or de-curse, anyway.

13) 1 rank of "resist mutation" is good enough to justify quaff-identifying potions. (-66% chance to mutate at each attempt)
14) An amulet of resist mutation is good enough to justify quaff-identifying potions. (-90% chance to mutate at each attempt)
15) Of course, that should be adjusted for which potions you have and haven't identified - unless you've already identified potion of mutation, in which case it's ok to quaff the rest (barring obvious cases where only decay/confusion/poison are left unidentified)
16) Scroll-identifying potions usually works best by targeting the smallest stacks.

17) Baiting monsters to identify stuff for you is nonsense.

[Wahaha]

13), 14) and 15) No, because it is a waste of good potions.
16) It's too hard to tell anything from potion stack size, except for the biggest stack probably being curing. Don't bother. When identifying scrolls I do target specific stack sizes depending on what I want to id.
17) Doesn't work anymore and was useless when it did.

[Pollen_Golem]

a waste of good potions? I usually quaff-ID before entering a tough fight from a safe place, so might/haste/brilliance do not go to waste.

I know you can't guess a potion from stack size besides curing/HW, but if you're seeking to ID mutation, is it very-rare or medium-rare, IE are you likelier to find it in stacks of 2 or single potions?

[celem]

You think in absolute terms far too much, most of these answers involve if's, when's and but's.

Grey weapons (and grey items in general) are useless to your character due to species, forget them entirely. This is your one absolute.

Blue weapons are identified as branded and the brand given if a monster appears holding them, these are rarely (never?) cursed if branded but may be if 'enchanted'. Blue items on the ground are total mystery and may be cursed/distort. You sometimes see green, that means youve seen a monster unwield it and its therefore not cursed.

I will totally abandon everything for the right weapon at the right time, regardless of species aptitudes. An elec dagger at pretty much anytime before D4-5 for instance, that will go almost to the end and you can imediately start reading your EW.

Uh, potion quaffing. I dont normally quaff-id until i've got a few basics known, namely mutation. Its not super rare but as said the stack size is not predictable, its not the biggest stacks (probably), but thats about it. I read until I find id, then ID potions starting with singles and smallest stacks until I find mutation, at which point I quaff-id anything thats in a stack of 2 or more, likewise for scroll reading. The only thing that gets an ?id after that is singleton scroll/potion and artifacts incase of contams/curses and so on.

[jejorda2]

16) Scroll-identifying potions usually works best by targeting the smallest stacks.

Identifying large stacks means you have more potions available to help you out of tough spots.
Identifying small stacks means you know what the rare potions are sooner. But you still might not know what the common ones are, which you actually need to use.

[partial]

I use ID scrolls almost exclusively on potions