Haste... for windwalkers?

Snapshot ID: 1c99b40c04a14b3cbd4f54313496260b

Far be it from me to question the wisdom of the robot. However, I’m really curious about the best in bags suggestion for my windwalker. It’s telling me to stack a lot of haste, to the point that haste is now my second highest secondary stat at 24%. This runs counter to every guide I’ve read; Wowhead, Icy Veins, and Method all put haste dead last for windwalkers. In tiers past, I’ve literally run 0% haste.

And, of course, Pandaren are not hasty by nature. The mantra of my people is SLOW DOWN.

Any thoughts?


I can take a look – what I’m seeing in that snapshot (and trying out BiS with a few setups) is that the optimizer generally sees all four secondary stats as good, but haste is edging out mastery a little as the 3rd best stat.

I can mess around with it and see if there is a tweak to the action priority being used that would push mastery above haste – I doubt it will impact performance much one way or the other because all four stats have good value for windwalkers, so I have no problem pushing it towards a more popular build.

If you say it’s OK, I trust you. It’s just unexpected, that’s all. However, I will also point out that the stat weights that come out of the optimizer also follow the other guides I have read. Here’s what I get:

Versatility=5.03, CritRating=4.97, MasteryRating=3.92, HasteRating=3.38

My reading of this is that versatility and crit are 49% and 47% more valuable that haste. If that’s the case, stacking haste seems like not the best idea.

That being said, my logs are looking really good, so I guess I shouldn’t argue. I even have a 100th percentile fight on mythic Volcoross for my ilvl bracket. I’m pretty far behind on gear because I didn’t star this tier until after the New Year.

Thanks for everything!


100th percentile… can’t beat that!

Yeah stat weights are always a bit tricky… For example, it could be the case for a spec that a high crit/haste/vers build is nearly as good as a high crit/mastery/vers build, but stat weights by their nature tend to only recommend one of those solutions (they are a 1st order linear regression, and thus don’t have the ability to describe as complicated a solution space).

I’m still looking at it – but got sidetracked trying to figure out what’s causing this server issue all of a sudden.