[Armstrong]

As I was getting to work on updating the AP coefficients post, I found myself looking for any kind of research about the new version of Blood Boil, but couldn't find anything (let me know if you did though). So after some testing and math, here's what seem to be the way the spell works now:


With no diseases on the target:
Base damage is 180-220 (average 200)
Attack power coefficient is 0.060 (up from 0.04 pre-3.1)

With one or more diseases on the target (2nd or 3rd do not add more damage)
Base damage is 275-315 (a flat increase of 95)
Attack power coefficient is 0.095 (an additional 0.035 * AP)

Another way to describe it would be to say that when Blood Boil hits a unit afflicted with one or more DK diseases, it hits for an additional 95 plus 0.035 * Attack Power.




Testing:

Level 80 Human DK in Frost Presence, specced 0/0/0, attacking the three level 80 practice dummies in Ebon Hold. Blood Boil Rank 4 was used, which has a base damage of 180-220, or 200 average.

To figure out the attack power coefficient, it's a good idea to calculate based on the average damage dealt using two different AP values. If the AP coefficient is the same at both AP values, you can be sure that no new hidden flat value or multiplier was added to the ability.

Using the gear in my bags, combined with Horn of Winter I was able to create a low value of 1066AP and a high of 3073AP. I made sure to keep HoW up for the high value and that no other DKs applied any buffs on me or debuffs on the dummies during the testing.



No diseases:

At 1066AP Blood Boil averaged 264 damage with no diseases, and 396 with a disease present.

Subtracting the ability's 200 base damage we are left with 64 damage from attack power with no diseases up.

1 ÷ 1066 × 64 = 0.060


At 3073AP Blood Boil averaged 383 damage with no diseases, and 587 with a disease present.

Subtracting the ability's 200 base damage we are left with 183 damage from attack power with no diseases up.

1 ÷ 3073 × 183 = 0.060

So the AP coefficient for Blood Boil without diseases on the target is 0.060.



With diseases:

At 1066 AP Blood Boil was hitting for 396 on average with diseases present. That's 132 more damage than with no diseases.

At 3073 AP Blood Boil was hitting for 587 on average with diseases present. That's 204 more damage than with no diseases.

Because we tested using two different AP values, we are able to see that the bonus damage is not a flat value. It's also not a straight multiplier of attack power, as 1066 ÷ 3073 does not equal 132 ÷ 204. More likely it's the combination of a flat value and a multiplier of AP.

To find those values, we will work with the difference between our two tests.

204 - 132 = 72 damage

3073 - 1066 = 2007 AP

So an increase of 2007 AP creates an increase in damage of 72. This allows us to find the AP coefficient for the bonus damage Blood Boil gains from hitting a diseased unit.

1 ÷ 2007 × 72 = 0.036


Now that we know the AP coefficient, we can find the static portion of the bonus

0.036 × 1066 = 38

132 - 38 = 94

OR

0.036 × 3073 = 111

204 - 111 = 93

So the bonus damage on diseased units appear to be a flat 93 or 94 combined with 0.036 × Attack Power. I think it's safe to say that the actual numbers are probably 95 and 0.035 since plugging them back in still yields the same results as the tests I did.



Hopefully this will be helpful to some. Let me know if you see any problem in the testing or if and how your own findings differ.

[Satorri]

I will do some testing of my own, but let me plug in some of my own very rough numbers to see if that ballparks with what I've seen.

In my tanking gear, with only HoW buffing, I have ~3500 AP. I'm draenei, Blood spec'd (no spell increases, but Bloody Strikes = "increases BB dmg by 30%"), and I have no other buffs possible. My BB hits for about 850-900 with diseases up.

So the (VERY rough) math:
0.035*3500 = 122.5 + 95 = 217.5 (round 218) + 200 (base) = 418*1.3 = 543

I may have misapplied, but I'll do some testing and ammend/add to/edit this post in a little bit with real numbers.

================================================== ============================
Setup 1: my blood tank spec ("30% increased BB dmg"), full tank gear + HoW (3543 AP)
BB spam, no diseases, I rounded up lots of lvl 80 ghouls hit a couple DSs to get all death/blood runes, then spammed BB to get lots of hits. I realized if my health is above 75% it will buff the dmg by 10% so I kept my health below for the test data.

With 0 diseases:
144 (non-crit) hits = avg 566, min 512, max 643 (11 crits = avg 1110, min 1031, max 1217)

With 1 disease (FF):
142 (non-crit) hits = avg 818, min 526, max 924 (5 crits = avg 1649, min 1586, max 1746)

With 2 diseases (FF/BP duh):
151 (non-crit) hits = avg 831, min 768, max 885 (7 crits = avg 1630, min 1554, max 1712)

It looks like you're right, it's not scaling per disease (the 1 disease sample looks like a non-disease hit or two got through and brought down the average a little), it's with or without.

I swapped a couple items to get a different AP value, and dropped down to 2444.

With 0 diseases:
280 (non-crit) hits = avg 447, min 415, max 564 (22 crits = avg 904, min 837, max 980)

With 1 disease (FF):
188 (non-crit) hits = avg 674, min 417, max 873 (8 crits = avg 1261, min 852, max 1442)

With 2 diseases (FF/BP duh):
162 (non-crit) hits = avg 741, min 442, max 820 (8 crits = avg 1529, min 1386, max 1631)


Feel free to use the numbers to add to the calculations or I will a little later.

[Armstrong]

Judging from how the average damage is consistently way off from the median, you have some stuff proccing that temporarily increases your damage, probably Bloody Vengeance.

[Satorri]

Bloody Vengeance only applies to Physical dmg, but I think you are right. When I was doing the second set of damage I was fielding some questions from someone on vent so I was slightly distracted and I caught myself instinctually applying HoW (part of how I reduced AP was not using it). I clicked it off, but it probably slipped in to mess up some values. I was careful but there may have been some points where my health popped over 75%, or when my diseases fell off for a hit that I didn't catch.

The first set was more rigorous though, and I can repeat the second set, or more 'a' second set later.

[Nunes]

All I know is that if you're Blood, it crit's for 2k+

[Armstrong]

Source: Satorri The first set was more rigorous though, and I can repeat the second set, or more 'a' second set later.

Well if it will save you some time, I wasn't really asking for more data per se. I was just saying let me know if you've done or seen other tests and have results that vary wildly from what I posted. So far I'm pretty sure the damage formula at the top is correct.

[Satorri]

=) Just pulling out data to check the formula.

So, set 1:
0 diseases = 3534(AP)*0.060+200=412*1.3 = 536 (actual: 566)
1 disease = 3534(AP)*0.095+95+200=631*1.3 = 820 (actual: 818)
2 disease = 3534(AP)*0.095+95+200=631*1.3 = 820 (actual: 831)

Set 2:
0 diseases = 2444(AP)*0.060+200=347*1.3 = 451 (actual: 447)
1 disease = 2444(AP)*0.095+95+200=527*1.3 = 685 (actual: 644)
2 disease = 2444(AP)*0.095+95+200=527*1.3 = 685 (actual: 741)

Set 1 hits the numbers dead on, set 2 was a little flukey, but like I said above, I trust the numbers from set 1 completely, set 2 was a little harder but the numbers seem to support the formulae.

Nice job. =)