Is BM really this far ahead of Surv now?

[Crowfeather]

Guys help me out for a min here. I was theorycrafting on FD and noticed that there's a whopping 20% difference in dps that my BM spec has over my Surv spec (407k vs 338k). Crowfeather-US Turalyon if you wanna double-check for yourselves. Gear's 585 with everything heroic except Legplates with Haromm's & AoC, reforged with crit as priority. My first question is, since I haven't really done any stat crunching on FD since the new patch hit, is this accurate? Has BM pulled so far ahead that there's a 20% difference, or is there something broken with FD?

[Azortharion]

Theorycrafting and FD don't belong in the same sentence.

Just stop using FemaleDwarf. >.<

[Queen of Hamsters]

Guys help me out for a min here. I was theorycrafting on FD and noticed that there's a whopping 20% difference in dps that my BM spec has over my Surv spec (407k vs 338k). Crowfeather-US Turalyon if you wanna double-check for yourselves. Gear's 585 with everything heroic except Legplates with Haromm's & AoC, reforged with crit as priority. My first question is, since I haven't really done any stat crunching on FD since the new patch hit, is this accurate? Has BM pulled so far ahead that there's a 20% difference, or is there something broken with FD?

I've done extensive testing in-game on a dummy, and I don't see that difference. Sustained, SV tends to pull ahead for me by at least a few thousand.

[Crowfeather]

Theorycrafting and FD don't belong in the same sentence.

Just stop using FemaleDwarf. >.<

Ok that's what I needed to know! I'm actually keeping up with a BM hunter in my guild of the same iLvl, so I was totally scratching my head over FD and thinking this can't be right. Thanks mate

[Jeremypwnz]

Make sure you have a shot priority or things get screwy, assuming you want to use FD.

@other guy: Training dummies aren't really for testing dps, but rather testing your rotation. Training Dummy dps is anecdotal and the lack of buffs/debuffs and cloak cleave (if you are at Dummies that are close to eachother) skew results.

Edit: Err. learn your rotation.

[royen]

Ok that's what I needed to know! I'm actually keeping up with a BM hunter in my guild of the same iLvl, so I was totally scratching my head over FD and thinking this can't be right. Thanks mate

It might be that you are just a better player than him + if you get rly lucky with LnL, that helps to.

[Azortharion]

There's no "testing" to be done on a dummy, it's not what they're for.

[Artemishowl]

There's no "testing" to be done on a dummy, it's not what they're for.

Especially in terms of MM hunters. Oh lawdy anyone who "tests" on a target dummy >.<

[royen]

We all know the one and only real test is LFR

[Xeromus]

Especially in terms of MM hunters. Oh lawdy anyone who "tests" on a target dummy >.<

Man seeing bows and totems makes me wish hunters could put down sentry turrets >_> or have a totem pet you can give a weapon to and let it pewpew

[Deleted]

To answer your question, yes and no. The be all and end all of it is this. BM is ahead in almost every scenario as far as parsing is concerned. On single target they are relatively close, but you will see BM pull ahead by a noticeable amount, yet it isn't that far ahead in reality. On cleave fights, protectors for example, there is a sizeable gap. The top 40 parses of both specs probably have around a 100k dps difference on a fight such as this one. That being said, you wouldn't necessarily be 'gimping' your raid by playing SV at all. It excels at target switching, it is good at doing the 'bitch' jobs too, such as engineer on heroic garrosh, which BM can't do as well.

[Kissthebaby]

The answer is no. Need proof, well you can just wowheroes me kid.









Edit: for seriously though BM is better, surv is still great though and real men play that spec.

Edit 2: not "men" but man, I think I am all alone now

[aHeeRo]

I've done extensive testing in-game on a dummy, and I don't see that difference. Sustained, SV tends to pull ahead for me by at least a few thousand.

These are my favorite replies.

[Effinhunter]

Theorycrafting and FD don't belong in the same sentence.

Just stop using FemaleDwarf. >.<

The real answer should have been that he's obviously not using FD correctly and probably has glaring issues in his rotation on the SV profile he was working with. The BiS SV and BM profiles are nowhere near this far off. Depending on stat reforges, you should be within 15k single target between specs.

Shot shifting is a known issue with FD and makes fine-grained analysis difficult. Simcraft, otoh, has had a history this exp of complete mechanic falsities, such as the TotH issue that persisted for a long time, confusing the crafters. Is the stampede model in current simc correct for haste values? I'm not sure, but I feel both sims should be regularly used.

Edit: P.S. One of the reasons that SimC has had issues this expansion is that the results are rarely verified because the module maintainer hasn't been a hunter for a while. FD, otoh, has been always maintained by a hunter. In WoD, I hope that we do have an active hunter developer for our module because there will be some fundamental changes, and I'd rather we have as accurate a model as possible in SimC, to complement the FD results.

[Azortharion]

The problem is that if you're doing anything serious "theorycrafting-wise", there is no compelling reason to use FD for any kind of comparison over SimCraft. It's hardly even faster when you get to know SimCraft which is fairly easy as they have a wiki with all the stuff you can change and add.

We do not have a /hunter/ dev on our SimCraft profile ATM, but we have an amazing warrior one working ours as well as those for warriors and balanec druids and probably a lot more(Collision) and he has already asked for input on the 6.0 version of it - the responses were fairly sparse, admittedly, but that's not our fault.

I deliberately did not elaborate my answer further than that there is no reason at all to use FD over SimCraft., and I don't know if it was sarcastic or not (my response probably warranted some snarky sarcasm if anything), but he did seem to appreciate the answer. Is saying what I did bashing FD a bit too hard? Maybe, it's a great tool, but with SimCraft being as good as it is especially right now I don't see any reason to use it ever. Ask Mr. Robot has a built-in function to make comparing your optimized gearsets in SimCraft easy as fuck all.

TL;DR FD is useless at this point in 5.4 and SimCraft will only suck if we don't let Collision know what's wrong with it.

[Effinhunter]

No one is arguing that FD should be used by "theorycrafters" to dive deeply into mechanics, reforges, etc. It is, however, a pretty good gauge of macro theorycrafting. It's bane is the model is still based off a deterministic simulation that doesn't accurately reflect the underlying RNG. But just because Simcraft has the potential to more accurately model the real hunter mechanics, doesn't mean the mechanics are accurately implemented.

Again, the hunter Simcraft module has a history this expansion of having poorly implemented mechanics. Imo, Femaledwarf has utility in that it can help you quickly evaluate if Simcraft is correct in the macro sense, especially when you discover boundary conditions in haste/mastery/crit values that should be further analyzed in Simcraft. For most people, FD will be fine for hunters just wanting a sense of the stat weights (like favoring haste over mastery, which is reflected in both simulators).

As someone who maintains multiple active open source projects with half a dozen regular committers each, I can tell you from experience that just because something is open sourced, does not mean that people are actively taking deep dives into the code. Millions of casual users could use an open source project and not one of them will actually find a serious flaw in it (e.g. recent root exploits in OpenSSL that have persisted for nearly a decade). That task tends to fall on the developers, who need to be actively using the module (which is why I'm saying that FD will continue to remain a good complementary tool, if for no other reason than to validate macro-level simulation results found in Simcraft.)

I brought up the Simcraft haste/stampede results you had mentioned in other threads as an example, because it is something I've wanted to test for a while, but I simply haven't been lucky enough to get the gear (e.g. Protector's weapon, HWF TED, etc.) to try a gradient of haste. Since it has been used to justify arguments that haste should just be stacked regardless of perceived mastery weight, and the result doesn't seem to immediately agree with FD, I think it's something worth looking into, but we're extending H Garrosh for the next few weeks, so I'm unlikely to get my hands on gear to try to validate Simcraft results anytime soon.

Anyway, nothing wrong with saying FD isn't a theorycrafter's ultimate tool. I don't agree with saying the tool is worthless or useless though. Zeherah puts a lot of work into the tool, and I appreciate what it is and what it is not.

[Azortharion]

You don't need gear to test SimCraft stuff, you can just import gearsets from Ask Mr. Robot that you set up yourself.

As someone who -really- wants to be the best I can be in terms of gear and whatnot, I've tried all angles on the Haste/Mastery kinda thing, and while there are some breakpoints, they almost never mean that you should stop going for Haste, and I cannot get Mastery to > Haste globally in any kind of gear combination.

Shot-shifting completely fucks over FemaleDwarf so I do not think you can reasonably start doubting SimCraft results if FD does not agree. If you drop 1 haste below 15002 in FemaleDwarf you lose like 3k DPS because of shot shifting, which also makes this the breakpoint that FD loves and the BM BiS set is based around even though it's not actually a correct thing/practice.

I'm curious exactly what you'd look into and how because then I'll just do it or something.

[Effinhunter]

I can't imagine a threshold where mastery will overtake haste at all this expansion. That's not what I was meaning. I am just interested in whether or not the extra attack curve/formula for Stampede is anywhere close to what Simcraft is saying it is.

What I mean is that I'd like to test out various haste levels from say 6000 to 18000 in increments of 250/500 and see how closely the Simcraft engine mimics the haste increases versus number of attacks. There's nothing major about the difference between haste/mastery reforges between FD and Simcraft until you get to the 14743 breakpoint that you datamined in Simcraft. Even after the ilvl upgrades, I can't get to that point yet, so it's one of those tests I just haven't been able to do (and I don't understand how AskMrRobot would help me there). I'd just like to see if these reforge points are valid or not, and if not, how close does the Simcraft stampede attacks curve come to reality in regards to haste.

[Alvarie]

Especially in terms of MM hunters. Oh lawdy anyone who "tests" on a target dummy >.<

You mean the crit bonus on targets above 80% health will give me more DPS?

I haven't used FD in a few years, it's a solid website it just doesn't calculate common sense. Numbers are numbers but most of the time being realistic is more DPS than breaking a haste breakpoint on Mars.

BM beats Surv just because it constantly sims better, as Surv relies heavily on RNG from 4 set and LnL. If you aren't the one to care about being the best no one said you can't play Surv over BM in a normal raid environment.

[Azortharion]

SV doesn't rely on RNG to do good damage, it just has an extra component of RNG that can make or break parsing damage, but even with average RNG you're sorta well off. Not compared to BM, of course. It takes outrageous RNG to beat a BM hunter with much lower trink uptime. It's also remarkably easier to play while also dealing with raid mechanics because the room for error is fairly big.

BM beats Surv because BM does more damage in 90% of cases single-target, and 100% of cases whenever you can Beast Cleave.