Explain AMR vs RaidBots


This is a pretty… epistemological question, but I’m going to ask it anyway. I have quite a few guildies who say there’s no real difference between AMR and RaidBots. What should I say to that?

My basic argument goes like this: If you have even a modest amount of gear, the number of combinations of gear * gems * enchants * azerite traits gets out of hand VERY quickly - the dreaded combinatorial EXPLOSION!. I think it’s completely unrealistic to expect RaidBots to be able to simulate all of those combinations ON DEMAND and deliver the results in anywhere near real time. I could be wrong about that, so please correct me if I am.

The major strength of AMR is that you guys have ALREADY run all those simulations and stored the results. Thus running BiB is reduced to just querying the results. I guess I shouldn’t say “just” querying the results, since that’s definitely non-trivial. But my point is that you don’t have to run through a bazillion simulations to get the answer

At the end of the day, if you’re trying to decide between two items in a single slot, I think RaidBots could handle that pretty well. But if you want to globally optimize over every combination of gear * gems * enchants * azerite traits… AMR seems like the only reasonable way to go.

What else should I tell the Doubting Thomases of my guild?



It sounds like you have a good grasp on one of the major differences:

Once you are manually picking a small subset of setups to check via simulation - there is no way to know if any of the setups are “good” or not. You can only know which of those simulates highest. Lots of times that is acceptable - but sometimes… it’s not!

On top of that, I would argue that comparing small numbers of data points with a simulator is not a good way to use simulation. I talked about it some in this post:

I would then argue that the solution that our optimizer gives is, at worst, just as good as anything you would get manually comparing specifically simulated sets. It is really hard to explain why set A simulating 1% better than set B doesn’t mean that set A is definitely better. That result more likely means: use either one, as they both fall within the bounds of the probable structural error in our simulation model.