General (New version) Forum query


#1

Hi,
Is there a way to track any of our previous Forum input that was done before V2 hit, or have all previous topics been lost in the ether…?


#2

We blew everything away, started fresh. We had a sticky in the old forum letting people know to copy anything they wanted to save. Sorry if you didn’t see it.


#3

Ahh… OK - there wasn’t anything specific, so no biggie.

New forum, new start… seems like a good plan every once in a while.


#4

So I tried out the new simulation and this is what I got.
http://www.askmrrobot.com/wow/simulator/report/0c3e019acf3645ef80547cf8aec5ba2a

If I am understanding it correctly it shows a mastery/haste/crit build is best for me.

When I go to the main page the goal stats are still crit/haste/mastery which seemed to be the default build.
BiS and BiB seem to recommend gear based on the crit/haste model as well.

I thought my new gearing strategy would now be based on the new sim of mastery/haste/crit. I named the sim
and saved it, but at this point I think I am doing something wrong or I don’t understand how it is supposed to work.


#5

After naming/saving it, did you then choose it from the picker at the top/middle of the gear optimizer page? Sounds like maybe you just didn’t select it before optimizing… let me know if that’s not the case and we can look further.


#6

Thanks for your help. Another question.
This is my equipped sim using the my new mast/haste stats
http://www.askmrrobot.com/wow/simulator/report/ae4bf83afa0643e18bbd8c22bc669dfa

Then this is what the BiB recommendation sims at
http://www.askmrrobot.com/wow/simulator/report/a3c23565c6724e3d97ea40e969bf889a

The BiB comes in lower than my equipped, what am I missing?


#7

Those are extremely close, 0.7% difference, so from the optimizer’s perspective, they are more or less equivalent. Within 1-2% is our goal with the optimizer.

The optimizer is able to rank a LOT of gear combinations very quickly (up to several quadrillion in just a few seconds). That speed costs you a very small amount of accuracy, but ultimately gives you better gear rankings than trying to guess at which subset of gear combos to simulate more accurately.

(For reference, it takes my fast but not amazing computer about 24 hours to simulate 100,000 combinations. So to simulate 1 billion combinations (which is not unusual even for just a player’s bags), it would take 10,000 days.)

So rather than make the user guess which small number of combinations are “probably” the best, then simulate them all and wait hours… we use an advanced algorithm that can approximate very closely in a couple seconds. Could you maybe squeeze out another half a percent or 1% DPS with some hand tweaking and simulation after the fact? Sure, sometimes. But you’ll never gain a whole lot, and it got you to a near-perfect solution almost instantly.