Mathematical Damage Mitigation Analysis
#1
Hi there!

I found this particular analysis to be something i'm sure a few folks around here would be fond of. Apparently a math major has derived a few equations that help determine whether that +10 stam is actually better than 100 more AC in terms of physical damage "survivability".

I thought i'd share it with all of you before it gets lost amongst the daily drivel that spews out of those forums.

Linky: http://forums.worldofwarcraft.com/thread.a...ior&T=43274&P=1

Sorry if this has been found already, but also, I'm not a math major so I can't personally verifiy if his/her calculations are indeed accurate.

After all, look where it was found :P
Reply
#2
Not as useful as he thinks. Since we dont know the exact formula of how AC applies to mitigation is all just speculation as to actual effect.


Reply
#3
Ghostiger,Feb 4 2005, 02:04 PM Wrote:Not as useful as he thinks. Since we dont know the exact formula of how AC applies to mitigation is all just speculation as to actual effect.
[right][snapback]67193[/snapback][/right]
I would agree, but with one addemdum. He makes many assumptions as to how equations are exactly determined. This would be fine if he was trying to just get ballpark or rough numbers as to what would generally benefit him more. Instead, however, he works out things in very precise detail, going for numeric relationships and "end-all, be-all" data. Approximations can only hold up to a certain extent for calculating things like this, and are much better suited for estimations and general discussion of trends.

That said, the way that he went about modelling things was worth the read and gives some insight as to some general trends.. just ignore the numbers he throws around in later posts. If you're looking to figure out a better model than what he gives, the base assumptions and the way things are put together is worth the read, too. :)

More than anything, I think his writing makes me think that it might be a good idea for someone to write up (or link to) a basic guide on mathematical modeling to give people some basic ideas to start with if they wanted to work toward figuring out how WoW works.
-TheDragoon
Reply
#4
Ghostiger,Feb 4 2005, 02:04 PM Wrote:Not as useful as he thinks. Since we dont know the exact formula of how AC applies to mitigation is all just speculation as to actual effect.
[right][snapback]67193[/snapback][/right]

Not true.

Because they show DR for equal levels in the UI, it's pretty easy to converge on a formula. The formula they ended up with was:
Damage-Reduction = armor / ( armor + 85*level + 400 )

The poster didn't know this, and assumed a slightly different formula that is close:
Damage-Reduction = armor / ( armor + 100*level )

What we don't know is the factor of mlvl vs. clvl, which I would speculate is just clvl/mlvl with an if/then for mobs more than 3 levels above clvl that which are clearly treated very differently.

This, of course, assumes the UI is correct. If any of you were around in DiabloII times, you know there were several things that displayed one way in the UI, but actual game effects were different, most prominently the early bow bugs with prefixes and suffixes and a bug with Amazon's strafe skill.
Conc / Concillian -- Vintage player of many games. Deadly leader of the All Pally Team (or was it Death leader?)
Terenas WoW player... while we waited for Diablo III.
And it came... and it went... and I played Hearthstone longer than Diablo III.
Reply
#5
So we do know the formula for how AC applies to mitigation (or an emperically derived approxomation).? That does change everything.


But given what that guy had posted I was correct. His conclusion was based on a signifigant guess.
Reply
#6
Ghostiger,Feb 5 2005, 02:47 AM Wrote:So we do know the formula for how AC applies to mitigation (or an emperically derived approxomation).? That does change everything.
But given what that guy had posted I was correct. His conclusion was based on a signifigant guess.
[right][snapback]67239[/snapback][/right]

The derived formula (assuming it is indeed accurate) was posted by someone else in post #5 of the thread. However, the original post was not so much about the exact formula, but an attempt to clear up a common misconception. It reminds me a bit of the resistance vs. life concept in Diablo. You need to have sufficient amounts of *both*, because they have a multiplicative relationship. But while going from 50 life to 150 is much more significant than going from 650 to 750, going from 5% resistance to 15% is much less significant than going from 65% to 75%. For some reason, a lot of people never catch on to that.

The obvious complications are that resistance (or AC, in the WoW) is useless against certain types of attacks, and on the hand using higher HPfe instead of damage reduction means that you require more healing. These are issues that can't simply be plugged into the formula.
Reply
#7
Man, I can only understand rough 3/4 of what is said... I need higher level math courses ASAP.

As near as I can tell from what is said, the man is right for physical damage. The diminishing returns on AC are countered by the increasing returns in the percentile reduction.

The arguement stamina is much more useful against spell damage is true, but all this makes me consider is:

1) Acquiring gear with both AC and Stamina.

2) Acquiring a set of Stamina/Resist gear for caster heavy areas, and AC/Stamina gear for normal life (this will be done via richness and enchanter friends). If anyone has read the long post on Blizzard's warrior philosophy, then they know that spells assume negilgable resistance on most everyone.

Personally, it's also worth noting the damage reduction that comes from a shield that comes with the extra boost in AC, and usually some Stamina. But I guess shield was assume in use anyway.

Oh, and I wish that man was posting here. His insight, even if not perfect, is wasted on those forums.
Men fear death, as children fear to go in the dark; and as that natural fear in children, is increased with tales, so is the other.

"Of Death" Sir Francis Bacon
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)