36 réponses à ce sujet

[SuicidalBaby]

hes good at math, execution is his downfall

[SuicidalBaby]

mr_afk wrote...

The number to stop at would be at 100 - which is the cap.
The only times it becomes more worthwhile to pump dex (crit chance) is if you have a base damage of 100+ (which is almost impossible as I don't think staves get to 50+dmg),

I posted this staff in the nuke mage thread some time ago

found the best staff @ lvl 25
Magus' Goods, Wounded Coast
Staff of the Magister - 10.8g - rare(orange) random
50 cold damage
+17% fire damage
+10 cold damage
blood mage +1
42 mag/con required

Modifié par SuicidialBaby, 03 juin 2011 - 04:31 .

[Mr_Raider]

With 1.03, willpower requirements are down, spike damage is down from staggers, and you can't one shot kill lieutenants any more. I would surmise that pumping magic up sky high to increase DPS is the way to go, with some points thrown in con for just enough health to last between heal cooldowns.

[mr_afk]

ishmaeltheforsaken wrote...
Aren't you the "worst player ever"? Why should I trust your commentary? 

hahaha damn Arelex, I see it's caught on 

suicidalbaby wrote...
hes good at math, execution is his downfall

Hey! How am I bad at execution?!! Just because my party is clueless and Arelex can get a few seconds faster than me doesn't mean anything! haha

Anyway, that staff looks pretty nice- but even with that staff, you would require 110magic to reach a base damage of 100, which is impossible. But given a little bit of critical damage% (from items) you'd definately be better off with some dex with that staff.. 
I wonder what base damage malcolm's honour has at ~level 25?


---------------------------------------
Anyway, I'll see if I can explain it succinctly for a change:

Each point into magic gives 0.5 base damage
Each point into dex gives 1% crit chance
Each point into cunning gives 1% crit damage


Generally speaking, you can calculate damage via 'expected damage over 100 hits' where

100hitsDmg = Base Dmg*(100-crit chance)+Actual Crit Dmg*(crit chance)

Dividing this by 100 gives you average damage, but it really isn't necessary for what we're about to do.


Anyway, what you should see from that formula is that there are three main components - base damage, actual critical damage, and critical chance.

Base damage = (magic-10)/2+staff base dmg
Crit damage% = 50+(cunning-10)
Crit chance = dexterity-10

Actual critical damage = base damage*(100+crit dmg)%

why cunning is ****
For example compare;
actual critical damage = (50.5 base damage)*150% = 75.75
to
actual critical damage = (50 base damage)*151% = 75.5

For any base damage less than ~75.5 you'll find that increasing base damage by 0.5 does more for actual critical damage than increasing the critical damage percentage by 1%. (75.5 seems to be the breakeven point)

When you take into account the formula for dps and the spell mechanics - where base damage is of greatest importance and critical damage isn't even taken into consideration more than half the time (based on critical chance), pumping cunning is never a good option (unless you are a rogue with devious harm).

E.g. Even my level 21 pure magic/con crit-mage with 67% critical chance, 88%crit dmg and 80.5 base damage still benefits more from pumping magic than cunning:
100hitsdmg=80.5*(100-67)+(80.5*188%)*67=12796.28, or ~128dmg on average

Assuming a level-up
100hitsdmg=82*(100-67)+(82*188%)*67= 13035
100hitsdmg=80.5*(100-67)+(80.5*191%)*67=12958


So basically, a pure magic mage with a high base damage will pack a greater actual critical damage than a cunning build despite however nice the large critical damage% may look.

dex
So, I'm doing a really bad job at been succinct, but anyway - we've covered critical damage but we still haven't considered critical chance - as busting out crits = ~50%+ more damage which is pretty good right?

Firstly, critical chance is a lot more useful than critical damage%
However, unless you either have a very high base damage or a very high crit damage% you're better off pumping base damage.

This draws back to actual critical damage.
Basically, unless your actual critical damage is high enough, it isn't worth increasing the chance of triggering it over simply increasing your non-critical damage.

Using a non-crit mage build (and heroic aura) you would have a crit chance around 12%.
If it had a similar stat distribution to my mage, the pure magic/con level 21 mage would have 89magic (base damage ~80.5) and critical damage% of 53%:
100hitsdmg=80.5*(100-12)+(80.5*153%)*12=8562
Assuming a level up,
100hitsdmg=82*(100-12)+(82*153%)*12=8722100hitsdmg=80.5*(100-15)+(80.5*153%)*15=8690

This means that unless you have your actual crit damage very high there's no point pumping dexterity. In other words, unless you have a specialised crit-build which uses item properties to achieve high crit damage% to increase actual crit damage, there isn't any point pumping dex, ever - well, at least until you hit the 100 cap.
Even my crit-mage build hasn't reached the point where dex is better than magic:
100hitsdmg=82*(100-67)+(82*188%)*67= 13035
100hitsdmg=80.5*(100-70)+(80.5*188%)*70=13009

However, in a few levels (once I reach ~91base damage) it may be beneficial to pump dex instead.


An interesting thing to consider however, is the added benefit from the ability for spells to crit-As spells have a much higher damage (as the base dmg is affected by a multiplier), in terms of spell damage it may be beneficial to pump crit chance earlier. 
However, as most of your damage will probably be coming from staff attacks and the recent patch has neutered the damage potential of spells like haemorrhage, I would say that pumping magic is the way to go.

Hope this helps/was interesting

Modifié par mr_afk, 03 juin 2011 - 05:08 .

[SuicidalBaby]

put that graph you made up

[mesmerizedish]

I like graphs.

[mr_afk]

I've never done a graph but I did a few tables. And they were only optimising for a certain level.
e.g. at level 14:


It doesn't take into account the large amounts of critical damage/chance that you can get by getting the right equipment later game.

The easiest way to picture it is as 'actual critical damage' and how large that 'actual critical damage' is determines whether you should pump dex or not.
For non-crit mages without the boost from items it is almost never going to be better to pump dex than magic until you reach 100 magic.

Modifié par mr_afk, 03 juin 2011 - 05:10 .

[Jack-Nader]

Magic yields dimminishing returns as you go higher. It is best to look at it in % form.

2 magic points = 1 base damage point.
2 Dexterity = 2% critical chance

Going from 49 base to 50 base is a 2% increase in damage.
Going from 99 to 100 base is a 1% increase in damage.

At 50% critical damage which is stock standard for everyone, Adding 2% critical chance is the same thing as adding 1% damage to your character.

Therefore with a critical damage base of 50% you should always increase your base to 100 before you add any points into dexterity.  Note that mages will actually reach the 100 magic cap before they reach 100 base damage, so as a rule of thumb just pump magic untill you slam into the cap and then pump dexterity.

Modifié par Jack-Nader, 03 juin 2011 - 06:38 .

[Sylvius the Mad]

That's a great explanation, mr_afk. It's good to know that pumping Magic is still the way to go in DA2, as it was in DAO.

[andraip]

@suicidal  He is good in math comparing to what? His Dragon Age 2 skills? Then he is VERY good in math

@the_worst_player  You typing to much

For warriors & Mages:

Go for Dextery if your base_damage = 5000/crit_damage, go for cunning when your crit_damage is lower then your crit_chance. Go for STR/MAG if this is not the case. PERIOD.

@mr_afk  There is no difference between base damage or spells, if it turns to be viable to go for crit in base damage it also turns to be viable for spells. In exception the talents who give you 100% crit chance, then the best damage wise for that talent is crit_damage when your base damage is 50, and 25 for people who have devious harm. If your damage is below that it's more viable to go STR/DEX/MAG.

EDIT: forget the sublined part, it's total bullsh*t, brb with correct data. I should stop talking about math when I'm still sleeping.

CORRECT DATA

when having 100% critical chance it's best damage wise to have crit_damage = 2 x base_damage wihtout devious harm and crit_damage = 4 x base_damage with devious harm.

Modifié par andraip, 03 juin 2011 - 10:03 .

[Mr_Raider]

mr_afk wrote...

I've never done a graph but I did a few tables. And they were only optimising for a certain level.
e.g. at level 14:

You should redo that table with a post 1.03 optimised stat distribution. Cunning and dex reqs are no longer symmetrical. At level 14 you can get away with 21 dex, 14 cun and the rest in magic or con for rogue gear.

[MichaelFinnegan]

mr_afk wrote...

Hope this helps/was interesting

Yes, I was looking around for such an analysis. Thanks.

However, as most of your damage will probably be coming from staff attacks and the recent patch has neutered the damage potential of spells like haemorrhage, I would say that pumping magic is the way to go.

Two other reasons to go for the magic attribute: increase in force power and attack value. Neither of these were there for mages in DA:O (a basic attack always connected - but, then again, there was no critical hit either).

One question: What is mage attack speed a function of?

An interesting thing to consider however, is the added benefit from the ability for spells to crit-As spells have a much higher damage (as the base dmg is affected by a multiplier), in terms of spell damage it may be beneficial to pump crit chance earlier.

Since attack value is low at the start and there is the additional requirement to overcome enemy defense, and so, if the "hit" check fails for a basic attack, damage dealt is merely 10% (?) of the base damage; and since spells always "connect": I think what you say has some merit. But I'm not sure how it'd actually work in practice, since there are other variables (like spell cooldowns) to consider.