So, we’re comparing bloodvine vest, 13int, 27sp, 2% hit to nemesis robes, 16int, 32sp, 1%crit, 23sp bonus

Just that one item in a vacuum looks like a clear downgrade going to bloodvine.

Looking at shadow bolt damage: 510 base, 0.8571 sp coeff

Right now he has 9% hit on gear, for a 16 - 9 = 7% chance to miss, 10.9% chance to crit + 5% from talents.

15% damage from sacrifice, 10% damage from curse of shadows, and lets say he can get full benefit from imp shadow bolt, 20%

(510 + 578 * 0.8571) * 1.15 * 1.1 *1.2 = 1526.2

The average shadow bolt will be:

(Base * 2 * crit chance + Base * (1 - crit chance)) * (1 - miss chance)

(1526.2 * 2 * 0.159 + 1526.2 * (1 - 0.159)) * (1 - 0.07)

1680.423

If we swap out the chest item we get:

(510 + 550 * 0.8571) * 1.15 * 1.1 *1.2 = 1489.773

The average shadow bolt will be:

(Base * 2 * crit chance + Base * (1 - crit chance)) * (1 - miss chance)

(1489.773 * 2 * 0.149 + 1526.2 * (1 - 0.149)) * (1 - 0.05)

1626.162

This isn’t the full calculation that our model of the game does, but this is the meat of what you would calculate for a warlock and clearly illustrates what we deem the correct way to factor hit into a mathematical model. If you miss 5% of the time, you do 5% less damage. In statistical terms, it is not a binary event that has value unto itself. 2% hit is essentially 2% damage, and then you have to calculate if the stats you lose (in this case crit and spell power) are worth more or less than 2% of your damage.

In this case, missing 2% more often with higher stats will be a DPS increase.

– I noticed yojji is optimizing with the setting to use corruption, just thought I’d mention it in case that is not what is happening in-game. Usually you can’t use corruption in raids as a warlock. Doesn’t affect this discussion really, but one of the nice things about our optimizer is that we take into account a lot of different variables like fight length, optional spells, etc.