RPS Stats and You

December 9, 2012

Hello Guardians!

I’ve been wanting to do this sort of research for a while, but haven’t really had the time until this weekend. I was looking at how the various Rage generation stats interact, and determining if any one stat was better than the others. I was inspired by something Theck did for Protadin DPS stats some time ago. Thankfully calculating Rage generation is much easier, but that didn’t stop me from making it complicated :)

I had originally thought I would use a similar approach – have a set of assumptions and use a spreadsheet to produce a pretty graph to explain the results. However I realized that was a terribly huge amount of work and went looking for ways to make it easier on myself. Then I realized it would be massively easier if I could just alter the base stats and buffs via options on the spreadsheet itself. In short, I taught myself a bit about how to code things into Excel to produce what I was looking for. Lets look at the results using me as an example:

It turns out there were a couple bugs in the original version of the spreadsheet. While I guess that isn’t really a surprise, It’s unfortunate that I kinda violated my own mantra of “Make damn sure everything is right.” In any case it’s fixed now, and now the real results are shown above.

So it turns out Haste does not actually generate more RPS than Crit. However what is still true is that the RPS increase required to power the additional T&C procs is about half of the total overall RPS gained by choosing Haste over Crit. That being said since T&C is quite worthless for some encounters (Lei Shi, Stone Guardians) and not as powerful as FR – especially not with the 4pc bonus – whether or not it’s advantageous to choose one strategy over the other is debatable.

What has definitely been confirmed is choosing Haste for fights which high physical EH requirements (Heroic Sha of Fear comes to mind) is a good idea. T&C plays very nicely with Savage Defense on fights with high damage melee swings if you have the extra Rage to power it. You can play around with this yourself and move the Haste value up and down.

So what does this mean for you?

Well for heroic 25m encounters with high melee damage it primarily means you’re going to want Haste instead of Crit. There are far more people to make up the tank DPS difference between the two, and you gain increased mitigation by going Haste – which is the big difference for tanks in 25m. For 10m it’s a bit more of a tough sell. Tank DPS is significantly more important in 10m than 25m, so it’s possible that the DPS you lose by going Haste over Crit could make or break a kill. Furthermore bosses don’t hit as hard in 10m as they do in 25m, thus meaning you can definitely afford to sacrifice that little bit of mitigation to do more DPS.

Once we get into the next tier of content I expect Haste stacking to become significantly more common in both raid sizes.

If you want to play with the spreadsheet, you can find it in the “Maths Corner” menu above. Some quick notes on how to use it:

  • Put in your unbuffed stats/ratings in the green section.
  • On the options page you can set the options that will affect the output.
  • It is assumed you aren’t a baddie and always have MotW up.
  • Bear Form, LotP, and Crit Suppression are automatically calculated.

If you find anything you think is an error, feel free to post a comment below, or contact me on twitter or some other means.

The next thing for me to look at is DPS. But I gotta figure it out first.

11 Comments

  • Moonspelled says:

    How come your results differ so much from the values in Tangedyn’s Level 90 MoP Bear spreadsheet? For RPS his spreadsheet has haste at approximately half the value of crit. I haven’t had time to look at it more closely yet, so I was wondering if you had.

    • Arielle says:

      Tangedyn never updated his SS with the buff of auto attack Rage to 10.85 since it happened after he quit.

  • Abc says:

    You can calculate attack speed with 10% haste buff this way: =2,5/(((HASTE_RATING/(425/1,5))/100+1)*1,1)
    Haste seems to have too high value for you compared to my own calculations, it’s better than crit but not as much as in your spreadsheet. My calculations might be wrong too, didn’t really check yours too deeply.

    • Arielle says:

      I just looked at it again and found at least 1 bug with the haste table that wasn’t related to Haste. I’ll look at the haste calc again when I get home from work tonight and upload a new version.

    • Arielle says:

      There we go, fixed it so it’s actually right this time. Thanks for pointing that out, appreciate it.

  • Aqusinna says:

    Haste stacking bears… hm

  • Moonspelled says:

    I’ve had a closer look at the SS and have some questions:

    You have Mangle at a fixed rate of one every 4 secs? Doesn’t the amount of Mangle CD resets by Lacerate, Thrash and FF depend on your hit/exp?

    Also, isn’t one every 4 secs a bit optimistic? I did some quick napkin math at hit/exp hardcaps and came to one every 4,571428571 secs. Note that this is different from the value in Tangedyn’s SS, so I might be wrong. I also didn’t factor in haste because I think guardian GCD isn’t affected by it, but I’m not entirely sure.

    Also (a non-crit) Mangle gives 5 rage on use, if I’m not mistaken. Not on hit.

  • Arielle says:

    Technically yes, the number of Mangles per second does depend on your Hit and Exp. The problem is it’s not possible to actually calculate this within the confines of a spreadsheet (at least not in any manner that I know of) and still be correct. A sim would have to be built for it. Unfortunately that’s not something I know how to do.

    Mangle only generates Rage if it hits. During Berserk the extra targets hit only generate Rage if they crit.

    • Helistar says:

      With a 1.5s GCD (i.e. ignoring the fact that FFF has a 1sec GCD) the average mangle interval is:

      1/((1-m)*0.25*(1/3)+((1-(1-m)*0.25)*(1-m)*0.25)*(1/4.5)+(1-(1-m)*0.25-(1-(1-m)*0.25)*(1-m)*0.25)*(1/6))

      where m is the miss chance (0 if you’re capped, otherwise hit+expertise miss chance):
      best case (capped) = 4.5714
      worst case (m=0.075+0.15) = 4.8161

      Both cases are a slight overestimate since FFF has a 1-sec GCD, so there’s a chance to gain 0.5s here and there. At the same time it assumes that you always hit mangle when you can, which is not trivial if you have a bit of lag (the mangle “flash” suffers from the same problem as the extra combo point from crit which cats have).
      (BTW the difference between the two cases is approximately 5%, a lot less than what I expected, actually…. at the same time the chance of reset is only 25%, so the impact of a miss is roughly 1/4 of the miss chance, which at worst is 22.5%, so it looks reasonable).

    • Arielle says:

      I thought about doing it that way originally, but then got myself confused and grossly overcomplicated how I was thinking about it.

      I’ll go ahead an add it to the SS. I’m expecting the relative value of Haste to increase a little bit as compared to everything else as a result.

    • Arielle says:

      Updated SS with the Mangle formula from Helistar. Results exactly as predicted. Crit and Haste now much, much closer together in terms of RPS.

Leave a Reply

Your email address will not be published. Required fields are marked *