Originally Posted by

**ShmooDude**
Not necessarily, for example a tank dying has significantly higher odds of being a wipe so that's going to raise his survivability up compared to a DPS dying the same amount. Probably not enough to interfere with what you want to do with it though.

Given the way you're doing your calculations, definitely. Or at least increase it to say 6 since even by your own code 3 dps deaths is only a 30% chance to wipe.

Also (if I'm reading your program right) you're using 100-survi to calculate the chance of death, but it should be (100-survi)*2 because you never have 0% survivability rating for a death (it can be anything from 0-99; 2 makes it assume those deaths average out to 50%). That's probably why your numbers aren't working.

For example some log I pulled up of 17 wipes. There's a resto shaman with an 89% survivability rating. In your current program that's an 11% chance of failure. Over 17 wipes that's an average of 1.87 deaths. However the shamans actual deaths (with ignore after = 3) were 4, with a 64.8, 27.8, 93.3 and a 27.2.

Same logs, another resto shaman with 84.6% average. By your program that's 15.4% death chance or 2.62 deaths. Actual deaths were 5, with a 94.5, 52.9, 38.3, 34.8 and 17.5.