Hello,
Can you explain to me the logic behind Azerite Traits recommendation?
So far the tool never suggested me to use Explosive Potential as a Demo Warlock even though it is known to be a top tier traits. Mind that I set up my strategy on multi target/mythic+.
Instead, it always suggests me a weird mix of everything usually including a single Baleful Invocation (even though it is also a very good trait to stack) or a single Demonic Meteor (can be stacked as well). It recommends a lot Umbral Blaze that I’ve never seen on a top tier list so far…
It almost never ask me to stack any traits it is almost like its avoiding me to use a trait more than once…
Really I don’t get it. And if I force trait strategy and simulate the result are usually a massive improvement.
You can say that I can set up a specific strategy, and you’ll be right but: Such illogical results makes me doubt the recommendation altogether, and the simulate tool as well since my understanding was that the recommendations were based on the simulation tool…
Maybe the BIB tool is not what I need to use, or I don’t use it properly?
Interesting fact:
If I select Demonic Strength as a Line 1 Talent it does indeed suggest me Explosive Potential * 3…
Can you explain how it is calculated and why a change in a talent can induce such difference?
If you use the help link above the BiB gear table, you can generate a snapshot id and post it here. Generate the id by using the create support post button.
Do that when you have a set of gear set up that you are finding simulates significantly higher than the default BiB suggestion and I can check it out to see if anything is going on.
Here’s default optimization suggested using my talent tree:
8693a47ff11f49b683b8fe2dbc4a9c46 -> Scores at 66k in Simulate tool
Here’s the optimization suggested when I switch from Bilescourge Bombers to Demonic Strength in the talent tree:
8802eb35e3924c6ca932c13294480ac4 -> Scores a bit more in Simulate tool but its not significant.
It is, however, aligned with what can be found in the majority of resources available online for Demonology.
Finally, here’s the suggestion with the first talent tree and a strategy that mimics the one suggested for Demonic Strength talent tree:
21af2b7b76014ed2ad0508b047ac2454 -> Scores over 68k in Simulate tool
That made me wonder: Why is it not suggesting the same azerite traits when I switch talent even though simulate tool prove them to be more efficient?
Those are pretty close to each other. Explosive potential gives a bit of a damage increase. At this point, there are so many complex interactions between azerite traits, corruption, talents, etc… we’ve had to accept a little bit higher margin of error in the optimizer. We make a trade off where we let it stray a tad more from the simulator data, but we can still go through quadrillions of combinations of gear in a couple seconds. Simulators aren’t perfect anyway, so I’d say this is still within an acceptable range.
These two sets of gear are close enough that you’re not going to feel a difference in-game. The only time you’d notice much difference is if there is a particular fight where you can game a particular trait for increased effect.
There is different underlying data for different talent combinations - that is why you’ll see different suggestions when you change talents. One of the strengths of our system is that we adapt the recommendations to your setup instead of only giving advice for the most common setup.
Ok thank you very much for taking the time to explain stuff to me.
Can you share a little more on the architecture of the BiB Tool? As I understand the recommendation varies depending on the talent chosen is it true regarding the major azerite as well? What has the most “weight” in the calculation?
Also, since the value of each talent change depending on the stuff you have how would you proceed to find the best combination of talent/major essence in regard of the stuff I have at my disposal (including bags)? Is it possible to do combination test sim that automatically adapt the BiB for each iteration of the test?
I feel like there is two-way of doing things: Either picking a stuff and define which talents are the most “effective” with it, or picking the most effective talents and find the stuff that is the most effective with them. I’m not sure how to proceed…
We have you pick your major essence and talents. We have pre-simulated data for all of those combos that we then analyze to pick the best gear for you.
If you go to the guides section of the site and check out the talents section - we have some data there for what major essence/talent combos have the highest potential for different specs if you were willing to use/optimize for them. That assumes you have access to all gear in the game, though.
As far as best major essence/talents for your currently available gear - you’d have to do some simulations on your own for that. The problem you will run into is that there are too many combos to actually do the simulations in a reasonable time. You could run BiB for a few different combos you are interested in and compare the simulations for each of those.
At one point we were thinking of making a tool that would let people use the optimizer to essentially find the sets of gear they want to test in the simulator… but it ended up being not that useful since the optimizer is generally very good at predicting the simulated result without actually doing the simulation - and almost instantly.