Akraen's (in)Complete Frost Stat Analysis

[Akraen 230]

If the issue is specificity then we need to be specific. Even completely made-up ET%s don't change the outcome of this analysis so the goal here is to be resolute in the verification of my conclusion.

 

I think my next step is to take a range of ET%s with error bars for the top 3 stat ratios (9577, 14242, 19347) and compare those to wrap up the conclusion of the OP, then if we do any specific further analysis or get bored waiting for our beta invites, then we could do a fight-by-fight breakdown just to be helpful to everyone who is stuck on progression still (lots of us).

[TLTeo 29]

If the issue is simulating (somewhat) realistic fights, a reasonable idea might be to have 4 different templates for a fight that alter the parameters of the basic model (say, extra IL damage on a cleave fight). Say, something like:

model 1 - Patchwerk style fight

model 2 - Cleave fight

model 3 - High movement fight (this one would likely be the hardest to estimate, but i think it'd be doable looking at logs)

model 4 (not so necessary really) - AoE/Add heavy fight

 

Also, i'll have some spare time in the next couple of weeks, and i'd be happy to help with the math side of things. I've never done any WoW (or gaming in general) theorycrafting, but i have a good math background (physics degree). PM me if you need any help!

Edited April 14, 2014 by TLTeo

[Akraen 230]

If the issue is simulating (somewhat) realistic fights, a reasonable idea might be to have 4 different templates for a fight that alter the parameters of the basic model (say, extra IL damage on a cleave fight). Say, something like:

model 1 - Patchwerk style fight

model 2 - Cleave fight

model 3 - High movement fight (this one would likely be the hardest to estimate, but i think it'd be doable looking at logs)

model 4 (not so necessary really) - AoE/Add heavy fight

 

Also, i'll have some spare time in the next couple of weeks, and i'd be happy to help with the math side of things. I've never done any WoW (or gaming in general) theorycrafting, but i have a good math background (physics degree). PM me if you need any help!

 

There aren't really fights like that though. They are transitions of smaller periods of any combination of modes. So that's why I'm suggesting a range of hypothetical ET%s so that you can see it doesn't really matter what the fight is like, 19347 wins, except maybe on the fringe. There is a minimum ET you'd have for LB/NT, in that no matter what you are refreshing at least once per 12 sec interval.

[TLTeo 29]

I see, that makes sense (and is probably much easier to do).

[Akraen 230]

If you'd like to help out, just take the DPET values from the OP and add in some ET%s, you could do it by hand or in a google doc very easily. I'm just pretty strapped for time.

[Ahnry 0]

Hey Akraen and everybody!

 

I have the following questions:

1) Should i move to 19437 haste > Mastery build, and sacrifice ~ 800-900 intellect? It is worth to sacrifice that amount of Int/SP? Or should i wait to get better gear ?

2) I have the Normal (non-WF) Thok ring, should i use it instead of HC Protectors ring ?

 

http://www.askmrrobot.com/wow/gear/eu/arathor/ahnry

 

Thanks in advance!

[Akraen 230]

Hey Akraen and everybody!

 

I have the following questions:

1) Should i move to 19437 haste > Mastery build, and sacrifice ~ 800-900 intellect? It is worth to sacrifice that amount of Int/SP? Or should i wait to get better gear ?

2) I have the Normal (non-WF) Thok ring, should i use it instead of HC Protectors ring ?

 

http://www.askmrrobot.com/wow/gear/eu/arathor/ahnry

 

Thanks in advance!

I'm not sure it's best to give up 800-900 intellect, because my stat weights were not so precisely tuned to be "accurate" but just to get the setup that I calculated worked for me. Basically I used AMR as a finishing tool, not so much the whole works.

 

Right now in the current state of frost margins and how tiny differences are in stat weights (I think everyone forgets just how ridiculously close these different builds are!), I think it's best to just maximize your item level. Take crit over haste even, if it's a higher item level. I have Thok's and Galakras' rings 574, but I can say pretty confidently that I am better off with TFP's hit/crit 580 ring over Galakras' ring, despite the loss in haste. That's because I'm able to hit any breakpoint I want without Galakras' ring, so it really is just best to take the extra item level.

 

If you can't hit the breakpoint, it's not enough of a gain for you to give up a substantial amount of intellect, so I would continue to work with a lower breakpoint, like 14242, so that you can better optimize your stats and benefit from PBoI's amplification.

 

That said, keep the 19k breakpoint in mind so that you can work towards it as you get more gear. I've done all my testing at 579-581 numbers so that I don't have to re-do it all. You'll all get to this gear level, there's plenty of time in SoO-- so you know moreso where to work towards, rather than forcing yourself to sacrifice it at a lower item level.

 

Hope that long-winded explanation helps, I want people to see my method of thinking through the entire question rather than just giving you a yes/no answer.

[Ahnry 0]

I'm not sure it's best to give up 800-900 intellect, because my stat weights were not so precisely tuned to be "accurate" but just to get the setup that I calculated worked for me. Basically I used AMR as a finishing tool, not so much the whole works.

 

Right now in the current state of frost margins and how tiny differences are in stat weights (I think everyone forgets just how ridiculously close these different builds are!), I think it's best to just maximize your item level. Take crit over haste even, if it's a higher item level. I have Thok's and Galakras' rings 574, but I can say pretty confidently that I am better off with TFP's hit/crit 580 ring over Galakras' ring, despite the loss in haste. That's because I'm able to hit any breakpoint I want without Galakras' ring, so it really is just best to take the extra item level.

 

If you can't hit the breakpoint, it's not enough of a gain for you to give up a substantial amount of intellect, so I would continue to work with a lower breakpoint, like 14242, so that you can better optimize your stats and benefit from PBoI's amplification.

 

That said, keep the 19k breakpoint in mind so that you can work towards it as you get more gear. I've done all my testing at 579-581 numbers so that I don't have to re-do it all. You'll all get to this gear level, there's plenty of time in SoO-- so you know moreso where to work towards, rather than forcing yourself to sacrifice it at a lower item level.

 

Hope that long-winded explanation helps, I want people to see my method of thinking through the entire question rather than just giving you a yes/no answer.

 

 

Yes, it helps, i think the same, I just wanted to be sure about that! Thanks mate!

[Zildijana 3]

I in the middle of changing from normal for heroic gear, but so far I dont have any hc tier items, do you think there is a point between hc and normal I could got drop the tier and 2 set bonus for the stats I wil get from warforged and heroic item?

 

running with this:

http://www.askmrrobot.com/wow/gear/eu/frostmane/zildijana

http://eu.battle.net/wow/en/character/frostmane/Zildijana/advanced

Edited April 20, 2014 by Zildijana

[Akraen 230]

Never drop the 2pc bonus, why has this been traveling around the internet lately? I just had to tell two people on MMO-Champion the same thing.

 

Who is saying this? Is it someone on battle.net? It is mega misinformation. I hope nobody took something I said about when you get above 28k haste seriously-- nobody in the world can even hit that number yet.

[Avengedlol 0]

Hey guys, would like some input if its worth for me to be at 18,969 haste cap with my current ilvl (561) or would 14,242 be the biggest bang for my buck? thanks.

 

http://us.battle.net/wow/en/character/illidan/Burnindesire/advanced 

[Kver 65]

Play whichever you can reach and play comfortably, there is no 'best' build.

[Akraen 230]

At what point do you throw up your hands, kick the trash can, and walk out the door to the pub instead?

 

[durrtygoodz 60]

Just noticed that some ppl you used on that chart are not currently playing frost, for example Feylaa is arcane therefore using arcane stats.

Edited April 22, 2014 by durrtygoodz

[Kver 65]

Just noticed that some ppl you used on that chart are not currently playing frost, for example Feylaa is arcane therefore using arcane stats.

I would assume he kept that in mind and didn't actually pick the wrong specs :P

[durrtygoodz 60]

I would assume he kept that in mind and didn't actually pick the wrong specs

uh read what i put: Feylaa is currently arcane specced and arcane geared. His arcane reforges are on that graph.

Actually looking at it again that russian "frost mage" with the 19k crit is also fire specced.

[Oltier 471]

uh read what i put: Feylaa is currently arcane specced and arcane geared. His arcane reforges are on that graph.

Actually looking at it again that russian "frost mage" with the 19k crit is also fire specced.

Doh, we'll have to fire Akraen :\

[Akraen 230]

Not to get too snarky here, but the fun thing about trend lines is... they ignore outliers.

 

All pedantry has been acknowledged and since it's completely irrelevant to the purpose and "conclusion" (though there isn't one really, it's just a quick look) is unaltered by removing those three mages, I won't go through the effort of updating the graph.

 

The moral of the story here is: don't go through the effort like I am figuring out frost stats, it really doesn't matter that much what you do, and I'm settling firmly on BiS ideal haste bp being 19347, nobody's opinion will sway me from that, only if someone actually proves me wrong.

[Akraen 230]

For a few weeks I'm going to look at the best frost mage parsers in the world and try to find trends in their stat patterns on various fights.

 

My guild's kill times are pathetic, so I don't expect to rank. But I will be logging my ET%s from fight to fight in this setup and compare it to 19.4k haste.

 

 

fyi: this is a good way to play with AMR, just open up excel/google docs, and jot down all the combinations you can make

[durrtygoodz 60]

I'm going to be playing with 14242 (well 14360 to be exact) this week to experiment with single target dps - Id very much like to increase on Thok/IJ/Malkorok.

[Akraen 230]

Thok is really my main target with this, too. Since we still wipe on it a lot (we're still rebuilding our healing core), it is actually a good chance for me to get lots of logs to check.

 

So yeah this should be a good week, please share your parses if you find anything interesting, I'd love to see!

[Frostedmages 3]

I'm going to be playing with 14242 (well 14360 to be exact) this week to experiment with single target dps - Id very much like to increase on Thok/IJ/Malkorok.

 

Try heavy mastery on thok. I was mastery > haste due to being arcane MS and switching to Frost and was beating out our 14.2k haste frost mage on thok pretty much every pull. I think it's due to the welly being able to go ape shit the entire fight? I had also glyphed him so he never got interupted while I chased Thok around.

Edited April 25, 2014 by Frostedmages

[Akraen 230]

Hehe Frost, with my gear, I can't really go below 14300. I mean, I could, but I'd be intentionally wasting stats to do so.

[durrtygoodz 60]

I actually can't tolerate with such slow casts honestly. 14242 feels fluid.

 

Also this pug has been awful so far so doubt I'll share anything. I had 520k on sha at about 35% or so, but then got banished, then straight into a prison, and we eventually wiped. Shame, could've been big. Oh and frost bomb on Galakras was awesome. Although I'm resigned to the fact that I'll never get the top parse until my main raid stacks 2 or 3 packs for big aoe.

[durrtygoodz 60]

Oh wow we just did the knock away strat on IJ. NO RANKS FOR DANIEL.