Gear and math

It’s been a nice relaxing couple of weeks in my WoW world. In my guild, we all took a break from what was becoming a very dull Nighthold raid circuit, and I seized the opportunity to work on a couple of alts — my balance druid and my destruction warlock. I find I enjoy playing them both, but the lock possibly a tad bit more than the druid. I still find the boomkin tedious for its long casts, but it gets better with better gear stacked for haste.

Both alts are hovering very close to ilvl 900 or a bit under, and the one thing that amazes me is how much better they are simply by virtue of having better gear. Trust me, in the last two weeks I have not suddenly become vastly more proficient on either one, but the difference in damage for both is pretty astounding. The only change has been upgraded gear. In some ways this is fun, because gear is relatively easy to get, even without subjecting yourself to LFR or mythic dungeons. But in other ways is seems kind of cheesy and not quite right. I guess it is an inevitable result of Blizz stepping away from the “bring the player not the class” philosophy — class/spec mechanics and gear seem to count for more and more these days. Nobody likes to blame gear for poor performance (well, okay, maybe some people like to), but that excuse is actually becoming more and more reasonable as Legion goes on.

I was thinking about this as I started last night to prepare my main hunter for resumption of raiding Tuesday when Tomb of Sargeras opens. Patch 7.2.5 brought some changes to BM hunters, and in spite of giving us a baseline 2-charge Dire Beast/Dire Frenzy, it is looking like overall we are in a worse place damage-wise than we were for Nighthold. Seems like Blizz just could not stand to have BM hunters close to the top, had to take away more than they gave. There will still be some class tweaks coming along in hotfixes, but honestly I am not holding my breath that any of them will include buffs for BM hunters.

At least two sites I read regularly have openly stated that MM is clearly — and by quite a ways — top of the hunter heap. From the IcyVeins BM hunter guide:

Now that 7.2.5 has released, we can say with reasonable confidence and assuming no major changes, that Marksmanship will be the optimal raiding spec going into Tomb of Sargeras, mostly due to the potency of its new set bonuses.

Beast Mastery remains a solid choice, though rather than being very competitive and sometimes even better at single-target than Marksmanship in ideal situations, it is now fair to say that its potential output is less than Marksmanship in nearly all situations.

And even the redoubtable Bendak, in his most recent BM post about Patch 7.2.5, is brutally realistic about BM, stating it will likely fall out not only in the middle of the damage pack, but likely in the lower middle at that.

Whatever. I am a hunter in WoW, that is who I am. And since Blizz has seen fit to destroy the essence of my vision of “hunterness” in MM and SV specs, I really have no choice but to continue playing BM. Numbers have never meant that much to me anyway, so what seems to be a sudden plunge from lower-top to lower-middle position is not a calamity. Some class/spec has to be in that position, it is the nature of rankings. Still, I will be interested to see what the actual numbers spread is when the ToS results start to become available. If the spread between top and bottom is large, then Blizz will have once again failed in its never-ending attempt to “balance” the class/spec mess they themselves caused.

My alt gear-centric push over the last couple of weeks also served to reinforce to me the utter insanity of Legion’s gear complexity. On my alts the calculus was relatively easy, since I never intend to actually raid with them: higher ilevel = good, secondary stats pretty much be damned. But when I started to weigh gear and talent combos on my main in preparation for ToS, I found myself once again despairing over the sheer mathematical enormity of the task.

It has gotten so bad that AskMrRobot is now implementing a SETI-like mass computer sharing approach to solving the gear problem for players. Mind you, modern computers already have pretty massive computing power. Certainly enough that even a middle-level server could perform general arithmetical comparisons, even for thousands of users at a time. But Blizz’s insane interdependencies of gear stats, talents, different types of raid bosses, RNG-dependent proc rates, and specialized legendary and set bonuses have gone exponentially past arithmetic calculations. To properly assess the relative value of gear, only massive computer simulations approach accuracy. One or two simulations at a time are handled (though slowly) on a home desktop computer, but if you are trying to do it for large numbers of players, you need vast computational resources, and the cheapest way to get them is to set up a distributed grid of community computers. (I applaud AMR’s ingenuity here, but honestly I would like to see a bit more detail on their app’s security setup before I open my computer to it.)

The point is, you need the power of modern computers to decide if a piece of WoW gear is actually an upgrade for you, or to decide which legendary works best with which set of talents. 

But Reforging was “too much math” for us.


See you after the release of ToS.

9 thoughts on “Gear and math

  1. We got my wife’s DK from 803 into the mid 850 range in about an hour. That was a huge difference. We, all 900+ geared, went into some heroics Tank and 4 DPS. Some things that barely hurt us one shot her. The difference in health pools is a big factor.

    The Stat weights and values have become so complex I don’t even think many theory crafters can decipher any more. I try looking at guides and seeing their comments Haste is wort 0.87, Mastery 0.74, Crit etc. and then the changes once your Haste is above X%, then values change. It is all so complicated that I just look at, does it have higher iLvL, are most stats green for increases, do I lose a ton of Haste. I have disenchant so much gear where I would lose over 1000 Haste but gain 1200 Versatility. I wish they would just stop trying to come up with new systems for stats, and just refine one to where it makes sense to the everyday player. Most of us get 1 decent piece of gear. We don’t have the luxury of stockpiling multiple sets tuned to specific fights. Even though the seem to feel we want to.

    1. I agree that even the theory crafters seem all over the place. I did a small experiment a week or so ago using my main with current gear and talent set, simple Patchwerk scenario, 8 minute fight. I plugged the data into my computer using SimC, into Beotorch, and into Raidbots. I happen to have several sets of gloves at around 900 ilevel in my inventory, and when I imported the results into Pawn, one sim said my equipped gloves were best, and the other two each identified a different set of inventory gloves as an upgrade. Three different sims came up with 3 different answers using the same data input. So yeah, it has gotten so complex no one really knows, not even the computers….

      1. Do fights even last that long? I see notices in guild chat. Guild has defeated Heroic X in 3 minutes 45 seconds

      2. LOL, well I know some of them seem to last that long! I think the tough progression fights often come close to that. I know several of our raiders can use their 3-minute CDs at least twice, and they don’t pop them at the pull. Six minutes probably would have been a better setting for what I was looking at.

  2. For a bit of fun on Sunday night our guild did an alt-run through EN Heroic. As we’d progressed through it we were wary of the mechanics, split people into groups like we used to……

    Total waste of time. Even on alts it was like running NH LFR. Hit hero at the start, most bosses were dead before it ended, and 90% of mechanics could be ignored. I have never seen a power escalation like this before, especially within the same tier. We certainly couldn’t have cleared ToT or HM like that.

    I went off AMR during WoD as it was giving me crazy gear suggestions which were plain wrong: I just don’t trust it now. I’ve been learning SimC this xpac, and found out about Raidbots through Reddit. Well worth checking out. I’ve found that SimC accurately predicts achievable damage, and that by following the stat weights it gives for gear with my main stat on it (never trinkets, necks or rings) my dps improves.

    I can’t wait for ToS to open and progression to start again: it’s why I play and love this game.

    1. I usually only use AMR for my alts, where I just want a quick and dirty recommendation for some gear, relative value of upgrades, fast assessment of enchants and gems. But you are right, even for alts I never pay any attention to their recommendations for rings, necks, or trinkets. I do like Raidbots — I use it or Beotorch for relatively quick sims when I don’t have time to do my own SimC runs.

      I, too, am really looking forward to ToS. Our first raid night resumption is actually tonight (Tuesday), so I hope things go smoothly on the Blizz tech side. Unfortunately, last night I had a serious file corruption error and had to reset my UI and run a scan/repair of game files, which caused most of my complex addons to completely reset. So I am still in the process of redoing my action bars, keybinds, and bag/bank inventories. Hoping Weakauras did not also reset (haven’t been brave enough to check yet) — if so, I could be screwed for efficiency tonight in raid. (Maybe I will be forced to blame my gear for poor performance. 😉)

    1. Yup, with the few thousand difference in damage numbers, seems to relate mainly to players looking for the edge to get or maintain a spot on a mythic-level raiding guild.

Comments are closed.