Search In
Find results that contain...
Find results in...

## Question

Was doing some math and every time I do a backstab I can't get the math to work out unless I replace 150 with 175% for backstab. Something is off on Graze on Backstab as well, maybe Graze is counting twice? EDIT

Upon further testing it seems like Backstsab is doubling the effect of grazes and crits?

Edited by bleedthefreak

## Recommended Posts

• 0

I was helping him run the numbers and I'll try to explain what seems to be happening.

It seems that on a backstab... crits and grazes are doubled.    If you backstab and crit, you get an extra 25% in there.  If you graze you get an extra -50%.

If that makes any sense.  Everything else seemed to work out right, but when backstab was in the picture it was off by those numbers.

##### Share on other sites
• 0

UI issue, graze is calculated separately, I don’t know the formula, @MaxQuest he found out how the graze is calculated.

Edited by dunehunter

##### Share on other sites
• 0

UI issue, graze is a separate multiplier. Same as bounce penalty, all penalty is calculated as multiplier.

See, we thought that as well, but the math doesn't work out that way either. It does work out if you assume the graze is -100% though.

EDIT - To clarify, the presence of Backstab seems to be making the additive Graze effect -100% and the Crit effect +50% instead of the normal 25%. When there is no Backstab in the mix the numbers work out about the way they should be (some minor rounding here and there).

Edited by bleedthefreak

##### Share on other sites
• 0

UI issue, graze is a separate multiplier. Same as bounce penalty, all penalty is calculated as multiplier.

See, we thought that as well, but the math doesn't work out that way either. It does work out if you assume the graze is -100% though.

EDIT - To clarify, the presence of Backstab seems to be making the additive Graze effect -100% and the Crit effect +100% instead of the normal 50%. When there is no Backstab in the mix the numbers work out about the way they should be (some minor rounding here and there).

##### Share on other sites
• 0

EDIT - To clarify, the presence of Backstab seems to be making the additive Graze effect -100% and the Crit effect +100% instead of the normal 50%.

You mean 50% instead of 25% with crits I presume?

Very weird behavior indeed. Could this be explained with Backstab using the (base dmg roll * (0.5 on graze || 1 on hit || 1.25 on crit)) instead of only using the base dmg roll - or something like that? Also not really if I look at the numbers...

Edited by Boeroer

##### Share on other sites
• 0

EDIT - To clarify, the presence of Backstab seems to be making the additive Graze effect -100% and the Crit effect +100% instead of the normal 50%.

You mean 50% instead of 25% with crits I presume?

Very weird behavior indeed. Could this be explained with Backstab using the (base dmg roll * (0.5 on graze || 1 on hit || 1.25 on crit)) instead of only using the base dmg roll - or something like that? Also not really if I look at the numbers...

Correct.

Backstab Crits were working out to be +50% damage (not +25%).

Backstab Grazes were working out to be -100% damage (not -50%).

As for why... who can say?  It's only with backstab a near as we can tell and only on crits and grazes when using backstab.  Regular crits and grazes are fine.  Hits on Backstab are fine.

##### Share on other sites
• 0

It is weird - but honestly not too bad. I like it. Works really well with assassination (and its increased crit damage).

##### Share on other sites
• 0

EDIT - To clarify, the presence of Backstab seems to be making the additive Graze effect -100% and the Crit effect +100% instead of the normal 50%.

You mean 50% instead of 25% with crits I presume?

Oops, yes, fixed! Thanks!

##### Share on other sites
• 0

This is not bugged, here is the explanation:

1. For graze hit: all the "positive" (actually the correct term is supra-unitary) modifiers have a step equal to the surplus percentage damage. Thus over-penetration has a step of 0.3, sneak-attack has a step of 0.5 etc. The only "negative" (actually sub-unitary is more appropriate term) is the graze modifier. The step in this case is 1-1/value=1-1/0.5=-1. Now all the steps are summed up: -1+0.15+1.5+0.5+0.15+0.3=1.6 Since this sum is bigger than 1 we have a damage of 22.7*(1+sum)=22.7*2.6=59.02, which the UI rounded to 59.1 (the correct rounding should have been 59 but probably there are some rounding errors in the intermediary results, in any case the difference is minimal). Now we apply the might modifier so we have final damage 59.02*(1+0.15)=67.873 (rounded by the UI to 67.9)
2. For the crit this is much simpler as we have only positive modifiers. So we sum up all the individual steps and we get: 0.25+0.5+0.15+1.5+0.5+0.15+0.3=3.35. Since this sum is bigger than 1 we have a damage of 24.9*(1+sum)=24.9*4.35=108.315, which the UI interface rounded to 108.4 (the correct rounding should have been 108.3 but maybe this is explained by rounding errors in the intermediary results, in any case the difference is minimal). Now we apply the might modifier so we have final damage: 108.315*(1+0.15)=124.56 (rounded by the UI to 124.6).
Edited by kmbogd
• 2

##### Share on other sites
• 0

Holy crap.

So we have:

- A 50% bonus like sneak attack adds 50% of base damage as final damage ( +0,5 = +0,5)

- A 50% penalty like graze lowers the number by 100% of base damage (1 - 1/0,5) = -1

If this is correct, saying " graze lowers damage by 50%" feels completely wrong, no matter how you put it.

##### Share on other sites
• 0

This is not bugged, here is the explanation:

1. For graze hit: all the "positive" (actually the correct term is supra-unitary) modifiers have a step equal to the surplus percentage damage. Thus over-penetration has a step of 0.3, sneak-attack has a step of 0.5 etc. The only "negative" (actually sub-unitary is more appropriate term) is the graze modifier. The step in this case is 1-1/value=1-1/0.5=-1. Now all the steps are summed up: -1+0.15+1.5+0.5+0.15+0.3=1.6 Since this sum is bigger than 1 we have a damage of 22.7*(1+sum)=22.7*2.6=59.02, which the UI rounded to 59.1 (the correct rounding should have been 59 but probably there are some rounding errors in the intermediary results, in any case the difference is minimal). Now we apply the might modifier so we have final damage 59.02*(1+0.15)=67.873 (rounded by the UI to 67.9)
2. For the crit this is much simpler as we have only positive modifiers. So we sum up all the individual steps and we get: 0.25+0.5+0.15+1.5+0.5+0.15+0.3=3.35. Since this sum is bigger than 1 we have a damage of 24.9*(1+sum)=24.9*4.35=108.315, which the UI interface rounded to 108.4 (the correct rounding should have been 108.3 but maybe this is explained by rounding errors in the intermediary results, in any case the difference is minimal). Now we apply the might modifier so we have final damage: 108.315*(1+0.15)=124.56 (rounded by the UI to 124.6).

Show this explanation as a background on the screen when Josh explains in front, how much effort they put to more readable rules. LOL

Edited by gGeorg

##### Share on other sites
• 0

Well it's really more complicated than this. It's true that in the current formula maluses have a bigger weight than the bonuses no matter the spectrum of the sums of the steps.

However, if you had only a graze, the actual result is just a -50% dmg: as you have only one modifier with step value -1 this constitutes a negative sum, thus we apply the second form of the dmg formula which is rolled_damage*1/(1-(-1))=rolled_damage*1/2=0.5 *rolled dmg.

Now imagine that besides graze you also have under-pentration with another 25% malus. The sum of the steps is  (1-1/0.5)+(1-1/0.75)=-1-0.33=-1.33. Since this is a negative value we have the dmg formula rolled_dmage*1/(1-(-1.33))=rolled_damage*1/2.33=0.43 * rolled dmg. As you see this is much better than a 75% malus (25%+50%).

The point is that there is a diminishing returns aspect to the formula. But it's equally valid the fact that maluses have a bigger weight in the formula than bonuses have.

Edited by kmbogd
• 1