Hello there fellow AskMrRobot Users,
i simmed for BiS in Phase 5 and i had a huge discussion with a pally friend about the sim not caring too much about expertise on his RET Paladin. Until now i blindly followed the suggestions from AskMrRobot but i feel like i need insights now to the simming on enhancement shaman, because how can this be my BiS List for the full content !
Most important issues :
Dagger from KJ ?
Not expertise capped.
No Relentless Earthstorm Meta
I took a look at it and refined the calculations surrounding windfury and mis-matched weapon speeds. I hacked our retail simulator to get some data points for windfury proc rates to produce some better values that we should be using for different weapon speeds and haste levels. I had an adjustment in there that would push more of the wf procs to fast weapons, but I didn’t have enough of a penalty to the total windfury proc chance when you mismatch weapons. This update should prefer the slower main hand (and matched speed) weapons even more aggressively than before.
As far as expertise… whether or not it is worth capping that is handled by the calculations. We do everything based on average values. It is often not the best use of stats, on average, to cap out expertise - but sometimes it is. It is too difficult to eyeball it. The optimizer is better at making that choice than a human with all their emotions involved. On my dev server it is liking expertise more than before, but getting less hit rating… so overall it’s still happy with where it was at, more or less. Itemization affects that to a large extent.
The relentless earthstorm will probably be picked with this update.
Hello Swol thank you for the fast response and thank you for the update !
I believe we also need the possibility for simming double wf weapon or wf + flametongue ( for faster offhand weapons) ← or an indicator which one would be better.
The cherry on the top would be a field that you can tick for manual syncing weapon to not consume to many flurry proccs. ( If it’s already implemented im just to blind to see)
Im happy with how far the simming for TBC is but considering its a paid service i would love a bit more out of it, like you already got it for other classes.
I already have it sync’ing the auto attacks because that’s generally what people want to gear for. Crit still ends up being one of the best stats, regardless - the main effect is that same-speed weapons become more valuable.
I could probably make a flametongue option - it looked like there were only some really specific cases where you would ever want to do that, and even then it was going to be a really close thing as to whether or not it would be a damage increase, so I decided to keep it simple.
@Swol
thank you again for your engagement ! Im still not sure how the stats are calculated.
This is an example from the ElitistJerks Forum for Enhancement Shamans. I have a gut feeling that you overvalue haste and/or undervalue strength.
ElitistJerks Archive Enhancement Shaman
When simming it prefers the Crit Gems over the Strength Gems.
Crit EP 1,97 T4 content Str 2 (2.2 with Kings) EP.
Also it values the +10 Haste (1.28EP - 1.92 EP from T4-Sunwell) Enchant on Gloves higher than +15Str (2 - 2.2 static without and with kings)
Of course i can be wrong here and the old stat weights are off… but then your sim proved all that we thought to know wrong from tbc classic enhancement. If this is true i need a new BiS List 
Other examples that i don’t understand →
Why is abacus ranked so high as a trinket (overvalue of haste?)
Why is the relentless not prefered (overalue of haste?)
Thank you for your time. If you happen to be able to look further into it i would be so happy!
Best Regards,
Daniel
PS: Not mentioning expertise here again, but i feel like this should also be ranked higher 
I cant imagine a world where T4 Gloves are better than Kara Expertise Gloves
PS2: 10:38 - The T4 gloves are only ranked higher because of the T4 2Pc Bonus, that actually makes sense, taking the t4 helmet off changes the priority.
+++
A different Bug or “Feature” that i believe i found is that the BiS function tells me that
2* Netherbane is better for me than Dragonstrike + Rod of the Sun King
Since there has been quite a bit of feedback about enhancement in particular, I have already gone over all the calculations again. I have read all of the old elitist jerks threads and taken into account all of what they say in our calculations. I did some actual simulations to help create a model for the strange behavior around windfury.
I’m not sure why “they” don’t value haste as much. The only way I could get haste to be valued less would be if I hard-coded something into the model to just make the dps go down if you attack faster… which would be extremely artificial and make no sense.
Can you post the snapshot for this case? (use the help link right above the gear table). The optimizer not picking the highest scoring set is a different issue. The dual-wield specs are the most difficult case to deal with, but with specific examples we can try to improve it.
Hey,
4b9f95c5728d4c6d87d68f06770c9321 is the ID from the Setup.
Alright i see what i did wrong. Blessing of Kings wasn’t ticked. That explains the weird socketing.
Still something that came up here :
Band of the Ranger-General is 0,46% better but doesn’t replace Ring of a Thousand Marks when Re Optimizing. Thalassian Wildercloak is 0,13% better than Rezor-Scale Battlecloak, but goes for Razer-Scale.
518c313e47204e81be5a03255f3ef11d ← this setup.
Thanks again for your help. I feel much better now about the simming. Just have to double check the items as of now.
Oh, and why is Abacus of Violent Odds rated so good. Better than Tsnumai Talisman. This just got to be an error 
I’m seeing Abacus ranked very slightly below Tsunami in the snapshot provided. Abacus has an over-budget haste proc, so it ranks very well for pretty much all melee specs.
I’ll look into why the optimizer is getting stuck and not finding that better weapon combination. The ring is a smaller issue, but we’ll see what we can do there. We have to accept some small margin of error for the sake of performance/speed. Usually the error is less than 0.5%, and almost always less than 1%.
Rod of the Sun King got flagged as dropping in Magister’s Terrace instead of Tempest Keep, so it’s getting filtered out. I’ll fix that in the next update. Since the boss is the same name in two different zones the parser got a bit confused.
