Why multistrike??

[ellieg]

I understood the multistrike damage was based on NON-crit damage, only the initial damage rolls for multi/multi.
Then the multis can crit, so in your example, a 200 crit is a 100 that can multi for 30 or 60 in case of a critical multi.

But maybe that's outdated info.

ok thanks. way less confusing than what i was thinking. so low end is a 100 non crit, no multistrikes. high end is a 200 crit with 2 multistrikes critting for 60 each.

[brookllyn]

ok thanks. way less confusing than what i was thinking. so low end is a 100 non crit, no multistrikes. high end is a 200 crit with 2 multistrikes critting for 60 each.

This is how I understand it but I have literally no source to base it off of.

[Sanguiris]

MS can't strike but the damage is based on the original attack. If the attack normally is 1000 multistrike will be 300 but if it crits and hits for 2000 then MS will be 600.

[ellieg]

MS can't strike but the damage is based on the original attack. If the attack normally is 1000 multistrike will be 300 but if it crits and hits for 2000 then MS will be 600.

ok so multistrikes dont roll for crit. attack happens, it rolls whether or not it crits, then rolls to see if it multistrikes. Wish blizzard would clear this up.

[Jaydenkor]

There is alot of incorrect information in this thread. (Not you Azor)

Multistrikes are 2, X% chances (based on YOUR Multistrike) to deal 30% dmg of the attacks base, non-crit damage.

EG: it does not matter if your attack crits, or just hits. The multistrikes crit independently of the initial hit.

part of the reason multistrike is better then crit for BM is because of Cobra strikes, which devalues crit for BM slightly.

The generalized *main* reason multistrike is so good for us over crit, is because of scaling factors. We ALREADY have 15% crit, base. We have 0% multistrike base.
Think of this another way (Extreme example), lets say you had 99% crit and 0% versatility.
What would be better? 1% more crit, or 1% versatility?

Math:

Rate of change formula: (New - Old) / Old = %change

200% dmg (100% +100% crit) - 199% (old crit value) / 199 =.005025 (0.5025% dmg increase)

1% versatility, up from 0%, is a 1% dmg increase...
So at this valuation, 1 Versatility is worth ~twice as much as 1 crit.

Now remember, obviously, we're only at 15% crit base ..so the impact on things is not so huge as it is at 99% crit... but you get the idea.. and this impacts why Multistrike becomes better then crit.
Later on, once we have values on our gear like 28% crit (total) and 20% multistrike.. and 15% haste.. and 8% versatility, and X mastery..then we can make better, not-so-wildly-swinging-all-the-time stat values.

[Invrlose]

I'm not arguing that MS doesn't scale the best, we are asking the question, "why?".

I made a spreadsheet based on the math here:http://www.icy-veins.com/forums/topi...ical-analysis/

with a rating conversion of 66 per multistrike % and 110 per crit %.

They have the exact same amount of final damage, even at 6000 rating of each. Here it is: https://docs.google.com/spreadsheets...it?usp=sharing

This means that 1 rating of multistrike increases your average damage by the exact same as 1 rating of crit. They are effectively equal in terms of raw stats, like I said, there is nothing inherent about multistrike that makes it scale better compared to crit, otherwise it would be the best stat for any spec that would ever argue between the two stats.

and Ellieg: MS scaling exponentially isn't quite true since you need to consider the other percent % chance that you get only 30% extra dmg. You can copy my spreadsheet and edit it so that MS % is .1, .2 and .3 and see that every 10% MS increases damage by 6% respectively.

-----

I believe the reason why multistrike scales better than crit in reality is due to base stats and the fact that multistrikes can crit. Since you have 15% base crit, multistrike scales with your base crit and will have a built-in higher dmg per point than crit.

Asking Azor "Why" is never a good idea. He will just attempt to "school" you and never actually tell you the mechanical function of something, instead he will just insult you. The post directly above this one explains the actual math the best in this entire thread.

[Wythel]

Nope, Crit and Ms roll individually.

Just tested this: (disclaimer: I'm Sv, so it's 36% damage, not 30%)
Ms-crit from noncrit attack:


noncrit Ms frim crit attack (included non-crit autoshot for comparison as well):



Multistrikes are not percental damage procs that depend on your original damage, but individual attacks on their own, free from the original attack's roll.

Erm...

First pic :

Initial damage = 684
Multi/Crit damage = 493

Formula is 684*.36*2 = 492.48 rounded to 493

2nd pic :

Initial crit damage = 4788
Multi damage = 862
Formula is 4788 / 2 * .36 = 861.84 rounded to 862

Multistrikes are percental damage procs that depend on your original NON CRIT damage to be exact.
But yes, they have their own crit rolls.

[Deleted]

But 1% crit is more damage than 1% multistrike, IIRC someone did the math and their scaling factors are the exact same unless something has changed.

in terms of raw dps, unless you play a specc with either buffed crit or mult dmg, then yes.... 100 crit rating = 100 multi rating in terms of raw damage potential over time.

Tho ofc, crit + multi = king as it gives you the chance for a crit then 2 multistrikes, that themselves then can crit... and its just awesome when it happens to a big hitter like chimera or kill command.

[Deleted]

There is alot of incorrect information in this thread. (Not you Azor)

Multistrikes are 2, X% chances (based on YOUR Multistrike) to deal 30% dmg of the attacks base, non-crit damage.

EG: it does not matter if your attack crits, or just hits. The multistrikes crit independently of the initial hit.

part of the reason multistrike is better then crit for BM is because of Cobra strikes, which devalues crit for BM slightly.

The generalized *main* reason multistrike is so good for us over crit, is because of scaling factors. We ALREADY have 15% crit, base. We have 0% multistrike base.
Think of this another way (Extreme example), lets say you had 99% crit and 0% versatility.
What would be better? 1% more crit, or 1% versatility?

Math:

Rate of change formula: (New - Old) / Old = %change

200% dmg (100% +100% crit) - 199% (old crit value) / 199 =.005025 (0.5025% dmg increase)

1% versatility, up from 0%, is a 1% dmg increase...
So at this valuation, 1 Versatility is worth ~twice as much as 1 crit.

Now remember, obviously, we're only at 15% crit base ..so the impact on things is not so huge as it is at 99% crit... but you get the idea.. and this impacts why Multistrike becomes better then crit.
Later on, once we have values on our gear like 28% crit (total) and 20% multistrike.. and 15% haste.. and 8% versatility, and X mastery..then we can make better, not-so-wildly-swinging-all-the-time stat values.

Very good explanation. Another way, I would put it is: If you have two stats, that have the same scaling (in the way brookllyn explained it on the previous page), then you get most damage, when you balance them out to have the same amount of rating.

What is the cobra strike thing though you are talking about? Do multistrikes grant additional focus? I dont play BM really, so I dont know.

Idk if this how it works, but if it does, this is why multistrike scales better. Lets say you have 10% multistrike. Thats 2 rolls at 10% chance to do 30% dmg. the chance of both rolls being a success would be 10% x 10% = 1%. So thats 1% chance of doing 60% multistrike damage. Now lets say you have 20% multistrike chance. 20% x 20% = 4%. So thats a 4% chance of doing 60% multistrike damage. So this is scaling exponentially.

Now this was already mentioned before, the average gain of multistrikes is linear, however the chance to get a double multistrike depends quadratically on your multistrike chance. Not that this would matter cause the average is important here, but I wanna point out though that quadratic growth is NOT equal to exponential growth. People tend to call anything higher than linear exponential.

One last thing, as far as I know pretty much most stats scale linearly, an exception being our mastery since it increases Dmg and Crit Dmg (so that kinda double dips when you crit) and old wotlk armor penetration (didnt do the math on that one though, so not 100% sure).

[Ettan]

No idea but id guess pet multistrike hits can give frenzy procs & proc invigoration.
That should be more than enough to tip the scale in favor of multistrike.

[Lunaboreal]

Well, your answers make sense, but when I changed my stat strategy to stack more multi strike, my dps went down almost 2k I have no other reason other than my stats. I tested it out for 5 minutes each on my garrison target dummy.

First test was with haste enchants, reflecting my stat prioritization strategy to keep crit/multistrike (25%) just above mastery (24.5) and 15-20% haste. I'm aiming for that kind of rationing to take advantage of hunter stat weights being close, the synergy and the role each stat plays in affecting dps. With this prioritization I did 16.8 single target damage, unbuffed, on a target dummy. Multistrike was at 7.91%

I changed my enchants to higher-tier multistrike and performed the same test. I did 15k dps, a significant drop from 16.8k. This was on a target dummy, a patchwerk test. I did not nor was there any reason to mess up my rotation. Multistrike was at 10.56%, almost 3% more than the first test. What happened to the exponential scaling??

I'm glad you've all figured out how multistrike is so OP, but it's more damaging to me than anything else. I can't risk doing 2k less dps because of strategies the guides recommend. Has anyone done any in-game trials to see the differences between the stats??

[mtvoelker]

Well, your answers make sense, but when I changed my stat strategy to stack more multi strike, my dps went down almost 2k I have no other reason other than my stats. I tested it out for 5 minutes each on my garrison target dummy.

Individual dummy parses are a horrible indication of performance.

[Lunaboreal]

Asking Azor "Why" is never a good idea. He will just attempt to "school" you and never actually tell you the mechanical function of something, instead he will just insult you. The post directly above this one explains the actual math the best in this entire thread.

Now, now, don't be bitter. You're all doing fine in this thread. I'm getting the "why's" answered and no one is insulting me, so let's not start poking at the hive now.

- - - Updated - - -

Individual dummy parses are a horrible indication of performance.

Are we starting with this again? Dummies are fine for comparing 2 things, keeping everything else the same. It's not for parsing, but for trying and comparing how things work.

[brookllyn]

Are we starting with this again? Dummies are fine for comparing 2 things, keeping everything else the same. It's not for parsing, but for trying and comparing how things work.

Right, but five minutes in each set of gear means you can't really get the most confident results. You could've had the best RNG on the first set and the worst RNG on the second set. Have you tried simming both sets of gear on your char to see if the sims support your (early) conclusion? If they match, then maybe MS is a bad idea for now, but if they don't match, then there is something else wonky that you need to figure out before you decide which stats to run.

Edit:
Not to mention, raid buffs can potentially increase the value of some stats(5% crit comes to mind especially when talking about MS).

[TheTrueM4gg0t]

I'm glad you've all figured out how multistrike is so OP, but it's more damaging to me than anything else. I can't risk doing 2k less dps because of strategies the guides recommend. Has anyone done any in-game trials to see the differences between the stats??

sorry, but it's not possible to win or gain 2k dps by just changing some of our secondary stats through enchants/gems etc.
comparing a setup with the very best second stats to one with the very worst probably wont make a 2k difference, as secondaries are relativly balanced these days. (and we don't have that much on gear yet)
what you're seeing is simply the margin of error you get from dummy testing, and not an actual 2k dps worse performance of the setup.
and Multistrike is not "so OP". it's just our best stat by a small amount. (as explained by Jaydenkor above)