|
On March 12 2014 05:07 Therapist. wrote:
The challenge of most enemies in crawl is really proportional to how quickly you can kill them. For example, if you're a melee based character and you're worshipping the Shining One, then Ancient Liches will pretty much no longer be a problem for you as soon as you have your Holy Wrath weapon because you'll blow them away in a few hits before they get the chance to do anything dangerous. Or if you otherwise have just happened to find a powerful weapon, they'll also go down quick.
Pets also will help with anything you're unsure of.
This is taking for granted that you successfully lure it around a corner so you don't have to close distance. A hasted ancient lich can still annihilate you in a few rounds if you're unlucky, TSO or no.
|
I have getting back to crawl recently and I find the new "encumbrance rating" in 0.13 a bit confusing. Is the 3 times rating < strength is a good enough rule of thumb ?
|
Blazinghand
United States25560 Posts
Encumbrance rating is a number representing how much armour harms your spellcasting, dodging, and melee accuracy. Units are "arbitrary units of encumbrance" (aue). There isn't a hard cutoff like there used to be.
Old system: has FDA encumbrance at 3, so you need that *3 = 3x3 = 9 str to wear FDA without massive penalties
New System: has FDA encumbrance at 11, so you want about 11 str to wear it, but there isn't a huge hard cutoff.
I am REASONABLY sure that basically you want your str at ER.
|
I am REASONABLY sure that basically you want your str at ER.
Anyone giving you numbers for str based on the armour you are wearing is making something up.
For any armour with ER > 0, going from str 1 -> str 2 has a larger effect than going from str 20 -> str 21. For any armour with ER > 0, going from str 20 -> str 21 still reduces the penalties (same for 71 -> 72, for that matter). It is now a smooth function with no breakpoints (well, insofar as armour penalties are concerned ... there is still the obvious breakpoint at str=0 and some for melee/ranged damage).
|
Blazinghand
United States25560 Posts
Oh, so they changed the shape of the line entirely?
so this is an accurate depiction of the changes?
![[image loading]](http://i.imgur.com/2LXZa6L.png)
|
wow that's some scientific shit right there bro
|
Blazinghand
United States25560 Posts
look it's not michaelangelo but it gets the point across right
|
On March 13 2014 05:49 Blazinghand wrote:Oh, so they changed the shape of the line entirely? so this is an accurate depiction of the changes?
mostly yes
|
On March 13 2014 05:38 crate wrote:Anyone giving you numbers for str based on the armour you are wearing is making something up. For any armour with ER > 0, going from str 1 -> str 2 has a larger effect than going from str 20 -> str 21. For any armour with ER > 0, going from str 20 -> str 21 still reduces the penalties (same for 71 -> 72, for that matter). It is now a smooth function with no breakpoints (well, insofar as armour penalties are concerned ... there is still the obvious breakpoint at str=0 and some for melee/ranged damage).
It's smooth, not linear (see above) so there is still a relationship where the benefit of increasing str isn't going to significantly impact the effect of ER. The point where that occurs depends a lot on the ER of the armor in relation to your Strength.
You actually contradict yourself really since you acknowledge that at some point additional str has a diminishing effect, but make that first statement. The STR = ER recommendation is now just a short hand for what STR you should want for a certain armor since in general that's when the benefit of additional str has a diminishing return on ER based rolls.
------------------
Also on ancient liches they're somewhat vulnerable to ranged ammunition from velocity branded weapons, especially steel, since they have relatively mediocre defenses and hp and their only respectable defense is AC not evasion. That is until they summon a meatshield of demons.
|
You actually contradict yourself really since you acknowledge that at some point additional str has a diminishing effect, but make that first statement.
No, I'm pointing out that anyone telling you that 15 str is better for your armour than 16 or 14 is making it up (well, ok, it's better than 14 since it helps more, but I know you understand what I mean here). Giving an actual number is making something up if you are recommending str for an armour that has encumbrance.
edit: perhaps I should clarify ... these made-up numbers are not necessarily bad suggestions (getting to 18 str for plate is a perfectly reasonable amount of str). I'm just pointing out that it's not terribly important which number you pick, so don't treat "18 str for plate" as some immutable number that you must achieve before you wear plate armour, or anything.
edit2: You actually contradict yourself really since you acknowledge that at some point additional str has a diminishing effect
It has a diminishing effect in that going from 1->2 is better than going from 2->3 is better than going from 3->4 ...
So yes, if you mean that "at some point" is at 0 str.
|
Ok thanks guys.
I am still confused about how I should know what's best to wear for a "midrange" character.
Leather armour / ring mail / scale mail ? Should I care about leveling strength above the wished carrying capacity? How do you guys decide during your games ? Or should I Just go for the shiniest armor (and whatever during the early game)
|
@crate there's a difference between each point being worth less and str being a diminished return over int or dex when you consider all of the effects of a reduced armor penalty.
@mavignon Usually if you want to cast + wear armor your probably going for one of the mid-EV dragon armors (fire, ice, pearl). I think any reasonable str is enough for those (8+) and then just pump int.
The odd thing about the formula is, if I'm not mistaken, that +1 int will always help spells succeed more than +1 str. The trade-off is if you want to be able to dodge and/or have higher melee accuracy you'd need str up to a certain point (or just all str idk) and the ability to cast comes as a bit of a side effect.
I don't know... for all the talk of break points it seems like I only ever pump one stat with some smaller consideration for the others if any. Either it's str + heavy armor with possibly 10-13 in int for casting, all int with 8+ str, or dex. But then again I'm not a good crawl player so I don't know.
|
The odd thing about the formula is, if I'm not mistaken, that +1 int will always help spells succeed more than +1 str.
Here is an easy counterexample: 5 30 20 stats (str/int/dex), 27 tloc, 27 spellcasting, 0 armour skill, wear plate armour. This should give 85% fail for apportation. Now raise str to 10. 1% fail (at 30 int). Raise int to 35 (at 5 str). 67% fail.
Obviously this is an extreme example but it shows that you are definitely mistaken.
As for what stat to raise ... in general all three stats give reduced returns per stat point as you raise them higher (str as discussed above, dex has EV stepdowns, int has spellpower stepdowns).
Also, it doesn't actually matter much at all which stat you pick. Whenever someone asks in ##crawl "what stat to raise" I just reply with "!rng str int dex" and tell them to do what the RNG says because if the answer isn't obvious it doesn't matter.
A simple and not-awful heuristic for deciding on armour, assuming you are past like d:5 or so: just wear the armour that gives the highest sum of AC+EV. This fails a bit very early on since AC is much better than EV point-for-point on d:1. If spell success is a thing you care about then you just have to make your own decisions.
|
Well ok, I wasn't considering the extremes properly or calculating correct sorry; but for practical stat values is there ever a case when str mitigates more failure than int?
edit:
Is the wiki's http://crawl.chaosforge.org/Spell_success formula correct? Or is the correct formula somewhere? Maybe that's where I'm getting a little lost as the formula seems to ultimately suggest only moderate impact of STR on spell casting success.
|
Version 0.10: This article may not be up to date for the latest stable release of Crawl.
(str has a different effect on armour back then)
That article is not really useful anyway imo ... it is too complicated for you to figure out generalities and wizmode is better for figuring out specifics.
For many plate characters, or most characters in GDA/CPA it is pretty common for str to help spell success more than int.
|
Finally getting back into my old DeCj, who is now XL:25 and sweeping through Vault:5 with a newly aquired posse of simulacrums. I had been beginning to slack off on the whole necromancy front lately, but a few high level zombies or simulacrum's can be invaluable for taking down larger hordes of enemies.
EDIT: Down to Zot:2, thinking I may grab some other runes before getting orb
DOUBLEDIT: XL:27!
|
Rolled a MiFi, and just found "+27 Lear's hauberk" chain mail before I found the temple.
this item.
wat.
|
Problem with that is that it fills up body, head, feet and hands slots, which means it is REALLY HARD to get resists. Still, in the beginning of the game it is INCREDIBLE because there is nothing you need resists for.
|
resists are overrated anyway, lear's is extremely good
|
I'd agree, but diminishingly so towards the later game.
|
|
|
|
|
|
EPT
