I’ve received a lot of questions and comments about secondary stats lately, and I thought it might be worth clearing some stuff up. Here’s a sample of the kinds of questions our weights are raising:
- Why is mastery the worst hunter stat for all specs? [it's not]
- I thought mastery was supposed to be the best stat for all classes? [It's not]
- How can mastery be bad for BM? [It's not] It boosts our biggest damage!
- How can mastery be bad for SV? [It's not] since it boosts all of our damage!
A lot of these questions aren’t so much about what weights are so much as why they are the way they are. BM hunters in particular seem to be struggling with the concept that boosting their pet damage isn’t as good as other stats.
I think there are two main things contributing to the confusion, so let’s start there.
Qualitative Measuring
You have to keep in mind that when we talk about stat weights and reforging, we’re talking in qualitative terms. You have to understand that there’s a difference between “better and worse” and “good and bad.”
Let’s say I give you $100 just for being a dues paying hunter. That’s good, right? You’re happy. But what if I gave you $120 instead? That’s even better! But that doesn’t mean that me giving you $100 is bad — it’s just not as good.
Same thing with our secondary stats.
Just because crit is better than mastery doesn’t mean that mastery is bad. Just because mastery is the worst stat for your spec doesn’t mean it’s a bad stat. This is a very important distinction! Especially now when the two bottom stats for a spec are often very close to each other in value.
Part of this good or bad thinking may be left over from Wrath era haste — haste wasn’t just the worst hunter stat, it was worst by leaps and bounds. We got used to thinking of haste as barely better that nothing at all (and, I should note, this was sloppy thinking, but very prevalent).
So again, just because a stat is worst than another stat doesn’t mean it’s a bad stat. That’s like saying agility is a bad stat because it’s worse than ranged dps. Agility is a great stat! So is mastery, and haste, and crit. They all increase our dps — just some of them are better. If the stats aren’t all perfectly balanced, one of them has to be on the bottom.
Pet scaling
Another contributing issue, particularly on the BM side, is pet scaling. Again I suspect this is Wrath-era thinking leaking into our new Cataclysm class design. Our pets now scale with all of our dps stats — and you cannot ignore the benefit of pet dps even in MM and SV specs. They contribute a lot of dps, even though BM pets contribute more.
This means that when we evaluate stats, we have to evaluate their effect on pet dps as well as hunter dps. I think a lot of hunters are used to assuming that all stats but AP affect only us, and not the pet.
Some Examples
I’m not going to go through every single explanation of why one stat is better than another for every spec — stat weights are calculated either based on exhaustive testing under very controlled circumstances or logical use of spreadsheets like Zeherah’s DPS Analyzer or, ideally, both. There’s a lot of tiny things that go into stat weights. Just about every single thing you do to optimize your hunter, in fact.
BM mastery is a frequently-asked one though, and many BM hunters email me to point out that they get a ton of dps from their pets, so increasing their pet damage should clearly be huge. And you know what, it is! Just not as huge as other stats.
BM mastery vs crit
Let’s take BM mastery vs crit, since that’s an easy one to explain and the most commonly-asked one I get. Your typical BM hunter gets maybe around 45% of their dps from their pet — that’s including Kill Command, by the way. There is of course a decent amount of variability to that, but 45% is a good median.
Now 1 mastery (not mastery rating, but 1 mastery) increases our pet’s dps by 1.7%, and that also boosts Kill Command damage. Since pets + KC contribute around 45% of our dps, 1 mastery is increasing our overall dps by 0.765%
Now 1% crit chance increases our crit chance by 1%, but also increases our pet’s crit chance by 1% and in addition gives us more dps from crit-based procs like Sic ‘Em and Invigoration.
For most of our shots 1% crit (well, from 0 to 1%) is a 1% dps gain, since we use the 200% physical damage multiplier. This is complicated a bit by our meta gem boosting crit damage, and the fact that we still have a couple shots like Serpent Sting using the 150% spell damage crit multiplier — but in general nowadays 1% crit is a bit more than a 1% damage increase — compared to the 0.765% of 1 mastery.
So 1% crit is a much larger dps gain than 1 mastery, and since they both use the same combat rating conversion 1 crit rating is equivalently better than 1 mastery rating.
Now let’s look at mastery for survival hunters.
SV mastery vs crit
For SV one point of mastery increases all their magical damage by 1%. That is the vast majority of their damage, certainly, but not all of it. That doesn’t affect auto shots, kill shots, or pet dps.
On the other hand 1% crit for SV increases the crit chance of all SV magical shots and auto shot and kill shot and – this is important — increases their pet’s crit chance. So 1 mastery is giving us less than a 1% dps gain, since it’s only 1% to our magical damage, white crit is giving us over 1% dps gain since it’s affecting everything, including our pet.
Since it takes the same rating to get 1 mastery or 1% crit, this makes crit a clear winner for SV, even before considering the benefit of talents like Sic ‘Em and Toxicology the further increase the benefit of crit.
Mastery is still good, and can still be used to balance dps
But just because crit is better than mastery, doesn’t make mastery bad. Mastery is still increasing your dps, and doing it by a substantial amount. Yes, if we have the choice we’ll reforge as much of our mastery as we can into crit. Heck, depending on your spec you might even wish all of your gear had only crit, hit and haste on it (certainly not for MM though).
But even if you somehow managed to have zero mastery on your gear, you’d still have 8 mastery, since that’s the baseline that you start with. Thus Blizzard can still tweak the balance of your class by tweaking mastery numbers (as well as tweaking specialization numbers).
Just a pointer:
Crit, Hit and MM’s Mastery are still “chance %”, RNG-based stats, and thus have diminishing returns. They may increase your dps on average, but if you are already very high on these stats (something we’ll see only in a year and a half from now), getting higher won’t get you the same dps gain as before.
In other words, a very lucky person won’t get much richer by being 1% luckier.
Registered just to comment on the above reply – I see it all the time, and it’s just the wrong way of thinking :)
“In other words, a very lucky person won’t get much richer by being 1% luckier.”
1% is 1% – it doesnt matter if it goes from 1% to 2% or 98% to 99%. It’s still 1%.
“getting higher won’t get you the same dps gain as before.”
This is true if you end up over the cap. But unless you are over the cap (don’t think it’s possible to reach the cap in any other stat than hit right now) that 1% will always be as valuable as the next. Going from 17% crit to 18% crit is in no way better than going from 98% crit to 99% crit.
Just wanted to add that since I’ve seen these kind of comments everywhere. “I don’t need any more crit since I’m already at 61%”. So? Getting 62% is still the same upgrade (again, unless you end up over the cap) as every other %-increase.
And it’s not really RNG, since the percentage will decide “exactly” how many more times you will crit or hit for those added stats.
thanks for this blog – it’s always fun to read :)
A person with 99% crit chance will do 199% of the damage he would do without crit, right? increasing that to 100% crit chance would increase his overall damage by about 0,5%.
A person with 0% crit chance will increase his overall damage with 1% if he gets 1% increased crit chance.
Kinda — it’s actually more with meta gem and all the proc stuff. But yes, going from 99% to 100% crit is the exact same dps increase as going from 0% to 1%. However, since your overall damage is increasing, it’s a smaller percentage of your damage.
To the above posts….you are both right and both wrong. More crit is always better this is true. And yes it is also true that once capped there is a diminishing return if not a null return on any more of a stay. However…where crit gets tricky is that it is a “chance” to proc a critical strike. It is NOT a gaurantee!
So if given an infinite sample size and you have 50% crit, yes half of the shots fired will be crits. However, an encounter or fight is not infinite. You can fire 20 shots and crit on only 5. Then fire 20 more and crit on 15. Neither sample was the 50% you have but the average between is.
Mastery is clearly a great stat as it provides bonus to extorting magical including the crates themselves. I think frost is dead on accurate when he states crit is better, butastery is really only a notch below. Not a slouch in it’s own right
Back in Classic, we used to use fractions to explain the diminishing return on stats like crit and how much each percentage is worth as you get more.
Going from a one in five chance of critting (1/5, 20%) to a one in four chance to crit (1/4, 25%) only takes 5% crit. Going from one in four to one in three (1/3, 33%) requires 8% crit.
It’s not a perfect explanation, but it gives a very simple view of how you need more of a stat to get a noticeable increase in proc chance as your rating increases.
Which is bullshit of course. :-)
You need more crit to go from 1/4 tot 1/3 than from 1/5 tot 1/4 simply because the distance is greater: 8% vs 5%. However the 1% crit still gives you the same gain. Just think about 20/100 to 25/100 vs 25/100 to 30/100 instead of to 33/100.
The only three diminishing return factors in crit returns are
1) There is a hard cap at which you 100% crit. Because not all shots get this at the same level due to talents and glyphs, you see soft-caps.
2) We have got secondairy talents like focus regen for pets. If – hypothetical – our pets are focus capped, more crit won’t help for that secondairy part. Sedondairy stats are usually not lineairly. For the damage part however 1% = 1% up to the (soft) cap.
3) RNG. On average this has zero effect, so it is not a real diminishing return, but often felt as if it is so by players.
How come? If you fire 20 shots with 50% crit, you can still have 20 crits or 0 crits. On average you will have 10. But the way random works, on most fights you will have not exactely 10 crits but a different amount. The amount of crits is distributed using standard statistics.
Now, vary this 50%. At low crit values the chance an individual fight has zero crits is greater than at high%. So the higher the crit, the smaller the chance that you’ll have a fight with (close to) zero benefit at all from crit.
This is how it sometimes feels as if a low crits small amounts of gain help more than at higher.
However again, I say feel, because this is individual and on average for every of those fighst with low crit, there will be equal other fights where you have more crits than the average %.
Now, there is in the article from Frost however one thing missing. Of course, if Frost had added it, it would probably be even more complicated, so I can understand he left it out. But that thing is: stat weights. If you reforge 1% crit, you get more than 1% hit/haste/mastery back. So even if 1% crit is better than 1% e.g. mastery, regorging crit to e.g. mastery may still be a again.
Note, I’m not saying it is, it is just an example. My point is, in addition to the effect of 1%, you also need to look at how many points are needed for that 1%.
@Armin Actually it doesn’t matter for stat weights that 1% crit or 1 mastery is more rating than 1% haste. Our stat weights are based on 1 rating, so that’s already taken into account.
If 1 haste rating is worse than 1 crit rating, we don’t care what the combat ratings conversions are, reforging haste into crit is always a gain.
Yeah, I should’ve clarified that it isn’t a real diminishing return in the sense that any one percentage point is worth more or less than another. It was all based on perception and how noticeable the gains were compared to how much you already have.
actually I wasn’t wrong in anything I said and nothing you wrote, Quori, pointed to any error. I think you are wrong when you say “more crit is always better this is true” though, cause crit will do nothing for you when over the cap. Unless you are over the cap it’s always worth the same, percent for percent. There is no diminishing returns on crit other than a cap.
Stating that you can’t fully “trust” percentages because it’s no guarantee doesn’t make any sense to me personally. You’re just as likely to crit more than your given percentage as less, but as long as the average stays the same it’s not an issue. I can, and will, expect to crit ABOUT 100 arrows out of a 1000 if I have 10% crit. I might only crit 90, or I might crit 110 – but none of this changes the fact that 1% crit will always be as good as the next percent (if below cap).
I hate auto correct. I truly do
Quori…what is Butastery :)
lol, BM Mastery
Thanks for this Frostheim, I came to the conclusion a while ago that crit > mastery for SV hunters, what I was wondering is where haste, agility and hit fit in there, in your opinion.
I was thinking hit (until cap) > agility > crit > mastery > haste.
That about right?
First off, I’d like to thank Frost for the very informative posts over the past few weeks about Cataclysm. I’d raise a pint of ale but we Blood Elves prefer a glass of wine. I’ve been playing around some BM stats since Cataclysm since I always strive to get the BM spec to be accepted by my guild as a viable DPS spec and would like to share my insights on what I have so far (no hard numbers though).
Looking at the secondary stats for BM, hit rating is no longer part of the equation since all hunters need to be hit capped and agility is always best since it provides dps and crit, I have tested between haste, crit, and mastery to see which one would benefit me more during 5-man dungeons in Cataclysm. I always believed that balancing stats is much more important than boosting a specific stat (like what happened to armor pen back in late wotlk) and I think it was implemented pretty well for Cataclysm. So basically, for BM hunters, I think its reaching a soft cap for haste then crit then leftovers for mastery. Its basically timing the channeling of our Cobra Shots to be in synch with our KC to make for a smoother “rotation”. With current gears its hard to actually determine the soft cap for haste but if we base it on our wotlk experience, it sits around at 20%.
Haste I think is important since it contributes to a number of factors which are vital to BM dps:
1. improves focus recovery. It is like the hunter’s version of spirit.
2. adds to a lot of autoshots, which improves with crit for pet focus recovery, which in turn improves hunter focus recovery with pet basic attack crits.
3. pets now scale with hunter haste, more haste = more pet auto attacks.
4. improves CS cast time, which when in AotF becomes very important in movement fights.
So right now my secondary stat priorities are: haste (until I cap) > crit > mastery for BM. My ideal numbers would look like +20% pet damage with mastery, 20% haste and all crit. At this moment am standing at 12.7% haste, 20% mastery (+3.7 pts), and 15% crit and am doing well in 5-mans averaging around 7.5k dps on mobs and 8.5k during movement boss fights and 9.5k up on standard burn fights with ilvl333 gears.
I highly suggest that you post both this and the previous post in the guides section. It will most certainly help hunters manage their reforging regardless of their spec.
Ok, beat me with a stick if i am wrong but..
If i currently have 50% crit chance, i’m already doing 150% damage on average.. so if i add 1% crit chance i’ll end up doing 151% damage on average, which is an increase of 1/150 = 0.66% instead of 1/100 = 1%.
If i get 1% DAMAGE INCREASE instead i’ll end up doing 1% more damage both on normal hits AND on critical hits, thus ending up doing 150% + 1,5% with an increase of 1,5/150 = 1%.
Of course this kind of issue is already taken into account in spreadsheet simulations, and on the other hand i understand that this subject would be ginored while preparing a “flat and simple” explanation about stat weights… but it brings me to think that probably stat weights are bound to reach some kind of balance with one another so that any paticular stat will become a bit less desirable when you already have loads of it.
If this was true, the fact that we have at least 8 mastery to begin with may be part of the motive behind Mastery being not particularly shiny as a secondary stat.
Two things, first and foremost, I think what Frost is saying is that you want to take items with the MOST overall stat increase and give consideration to the weighting. So, if an item has say +10 agility, +10 Hit +10 Haste and +10 Mastery, vs an item with +10 agility, + 0 Hit + 5 Haste and +100 Mastery, You’d want the latter because even though Mastery is our worst stat, there is a lot of it and enough of it to override the benefits of the other “missing” stats and, some of it can be reforged into the missing stat.
Secondly, a question. Wasn’t the meta gem issue supposed to be “fixed”, or the change of mind implemented by now? It doesn’t seem to have been? Should I go out and get more blue gems than red?
Thanks!
Blizzard announced they were going to revert the change, however it was going to have to come in a minor content patch, rather than the regular hotfixes we’ve been getting. We are still awaiting that patch, and no target date has been given. From the list of bugs I’ve seen thus far, I’m guessing they’re working on a pretty decent sized patch and are testing it internally as we write.
This is just slightly off topic, but it’s something interesting I’ve noticed regarding SV at the moment. I’m not sure if this is intentional or not, but currently Cobra Shot is increasing the duration on Wyvern Sting. From my limited testing Wyvern Sting seems not to scale as well as Serpent Sting, although on my character it’s currently ticking at 914 every 2 seconds for 457 a second, vs Serpent Sting which is ticking at 1280 every 3 seconds at 426 a second. I’m unsure if talents such as Improved Serpent Sting and Toxicology are effecting Wyvern sting, however it would be very interesting if they did. I’ve been using Wyvern sting a lot in heroics as a very effective method of cc, so I’ve noticed the sting up quite often. I’ve also noticed that if your pet is on Defensive, and you hit a target with Wyvern Sting, the pet will rush to attack that target, so keep your fingers on that passive button!
While we’re arguing about percentages and weighting, just one incidental thought…
If I do 1000 non-crit DPS, adding 1% crit is going to add 1% of that 1000 regardless of how much crit I already have. However, note that this is only after all factors other than crit are considered. Why does that matter? Simple, because that means that as you increase your other stats, crit becomes more valuable….and conversely, as you increase crit, your other stats become more valuable.
Thus, if you have a ton of crit and no power behind it, adding AP is way more valuable than adding more crit. Conversely, if you have no crit, but a ton of power, then you’ll get a lot of bang for your crit rating buck compared to what you would get for the AP. Replace either crit or AP with any other DPS stat, and the argument remains the same. The more you stack one attribute, the less valuable it becomes compared to everything else. In theory, there should be some idealized balance between stats that you should achieve over the long-term, but realistically you will probably never strike that balance.