Something is off in BiB

So, I reported a similar issue before with my Shaman where the stat weighs seem to be leaning way too heavily into versatility. This time it’s with my Mage:


Crit = 2.90
Haste = 3.47
Intellect = 8.35
Mastery = 3.71

For DPS specs, this shouldn’t happen unless one is on the edge of DR for everything else.

RaidBots with the same set up:

Intellect = 6.59
Mastery = 3.80
Critical Strike = 3.33
Haste = 2.29
Versatility = 2.06

There’s no such thing as “right” & “wrong” in WoW, unless Blizzard alone says so;
AMR & RaidBots offer different advice and as with all advice it can be taken or left.

While I accept that this thread is on the more assertive end of the query scale, it isn’t as harshly worded as some I’ve seen.

Does RaidBots deal ‘exclusively’ with a single spec. scenario? AMR can (well; does) factor in all spec’s that you have set up for the character data you import. If your game setup only has one, then fine… but AMR doesn’t look at the game from a “you must” stand point and - to be fair - any site that offers such a “black & white” version of appraisal is probably best avoided… that tends to fall into the category of TELLING you how to play, ratrher than suggesting.

All that said, I’m sure @Swol & @yellowfive may be able to go into a better worded explanation as to why there are differences… so I’ll leave that side of things to them.

Try to remember, you came to AMR for advice; advice was offered - what you do with it is up to you.

There may be different “strategies”, but there are also gear pieces/setups that are objectively better than others.The entire point of the BiB is to provide users with a quick glance at the best setup they currently have in their bag. If the system is over-estimating the value of a given stat (in this case versatility), then the results will be wrong.

AMR’s multi-spec feature works on a priority system based on how the specs are listed from top to bottom. In any case, the particular instances I’m referring to are all characters with only a single spec.

I am aware that our code likes versatility more than simc for frost mages. I’ve gone through the code at great length… I’m not sure what I would change, really. I’ve compared my calculations to spell breakdowns from logs and tried making tweaks for different aspects of the kit… Nothing looks far off from what people are doing in-game.

To me, versatility makes sense for a frost mage in decent gear. It’s not hard to get to a point where crit and haste have significant diminishing value. So that just leaves mastery, which is good, but my calcs don’t like it as much as the simulator.

The only way I could see my calcs changing, is if people aren’t playing how I expect. And that would generally mean using less ice lance. I see in some of these logs that players aren’t always using two ice lance after a flurry. I’m not sure why that would be the case. My code assume you always use two ice lance after a flurry.

I’m not sure on the semantics of fine-tuned rotations and how that plays into things. It just seems to me that versatility is over-valued by a notable margin, at least in my case. With the snapshot above, you can see that my crit is only at 16%, which most guides say out-weighs versatility until ~33%.

I get that these are general recommendations and that a regeneration of stat values is required for each new gear option, etc, but I don’t think I’m at a point where versatility is actually better than crit, haste or mastery and it doesn’t seem like the BiB should be prioritizing that. Obviously, I could be wrong. Also, I’m having a similar issue with my Enhancement Shaman (there’s a post further down on that).

First, I wouldn’t use the stat weights generated for the pawn output as much of a measure of what general stat strength is. We use a very simple method to make those weights, and it will be prone to bouncing all over the place.

I’ve tried so many different ways to get the calculations to favor more crit… I just can’t make it happen unless I hard-code in a multiplier or shatter less spells. But you should be trying to shatter as many spells as possible, so it doesn’t make sense to change the code to play sub-optimally.

Overflowing Energy is going to make getting more crit less valuable, and the value of overflowing energy goes down the more crit you have. So right there you have this weird relationship where you get more crit from that talent by having less crit on gear. Just taking that talent into account, I’d put that “soft cap” on crit down to at least 27%.

The value, point for point, of crit has a very complicated relationship.

I’m still messing around with enhancement shaman code, so we’ll see where that falls once I’m done.

I mainly use stat weights to see if something is an upgrade at a glance or just to see where my stat distribution is at with my current gear. I use BiB as a way to quickly ensure that I’ve got the best of what’s in my bag equipped. I don’t raid, so I generally only sim as a means of comparisons to get a general idea of where I should be.

I only made this (and the Shaman) post because I was concerned about the stat weight determinations vs what I’ve read and simmed. I’m trying to optimize where I spend my gold on gems/enchants. I do appreciate you responding and explaining the logic behind the results I got, so thank you for that. :slight_smile: