Roll the bones

[Deleted]

There's actually another issue here that hasn't bee addressed. Consider:
"Fishing for Shark Infested Waters, no matter how many buffs you had: 220,219"
Ok, fair enough. But what also has to be considered is how many times you can reroll before you encounter a diminishing return.

To take an extreme example just to make my point: what if you had to reroll 10 times? That's 50 (or 60) wasted cp's, plus the loss of the other buffs. Is SiW really so strong that it's preferable to 3 other buffs (we won't even count GM) and the loss of 10 eviscerates? Probably not.

So where i'm going with this is that there should probably be some kind of modifier in the above statement, so that it reads something like, "Fishing for Shark Infested Waters--up to X rolls-- no matter how many buffs you had: 220,219

The amount of rolls it takes is irrelevant. You can't ever know how many rolls it's going to take this time, so it isn't a meaningful decision you can make. But all of that is already been factored in by the sim. Sometimes you can get a really shitty streak and not get SIW the whole fight, but still in the grand scheme of things fishing for SIW is still the best option. If you stop after say, 10 rolls as you suggest, then then damage is already done and you just settled for suboptimal buffs anyway.

Now, a real qualifier would be fight length. I can't see much sense fishing for SIW if the boss is going to die in lets say 20 seconds for example. Would be nice to get sims on where these kinds of breakpoints would be. I could possibly even see this being a staggered thing; for the first half or 2/3 of the fight you fish for SIW or 3+ buffs and then later on SIW or 2+ buffs and in the very end settle for whatever you can get.

[Tellof]

So where i'm going with this is that there should probably be some kind of modifier in the above statement, so that it reads something like, "Fishing for Shark Infested Waters--up to X rolls-- no matter how many buffs you had: 220,219

The problem I see with this is that you have no idea what buff set X+1 will be. What if it's Grand Melee? You would have to suffer 40+ seconds of one of the worst solo Buffs until you use it up, all the while having lower throughput. I would guess the answer to this would be that after X re-rolls you continue to re-roll but settle for a set of 2 good ones... something like that. The reason the heavy re-rolling works out in the end is because when you do hit that money shot your DPS goes through the roof and you more than make up for your slump in the re-roll phase. If it didn't happen that way then the sims wouldn't show what they do.

Great point though, and it's something we should explore the potential of.

[Zoopercat]

I actually had the reroll count on that full blog post: http://blog.askmrrobot.com/2016/07/o...the-bones-rtb/ For a 450 second average fight length.

If you never reroll, you apply Roll the Bones 11 times in a fight. If you reroll for shark or any-3, you use it 30 times. So on average, you are rerolling twice (for a total of 3 rolls) to get either any shark combo, or any 3 buff combo.

But as people already pointed out, you don't know how many it will take when you start rolling. It could take 10 on one try, 1 on another. So it isn't helpful in the decision making process. But the simulator does this over thousands and thousands of iterations, giving you the average, and accounting for the DPS loss by not casting something else.

[Tellof]

I actually had the reroll count on that full blog post: http://blog.askmrrobot.com/2016/07/o...the-bones-rtb/ For a 450 second average fight length.

If you never reroll, you apply Roll the Bones 11 times in a fight. If you reroll for shark or any-3, you use it 30 times. So on average, you are rerolling twice (for a total of 3 rolls) to get either any shark combo, or any 3 buff combo.

But as people already pointed out, you don't know how many it will take when you start rolling. It could take 10 on one try, 1 on another. So it isn't helpful in the decision making process. But the simulator does this over thousands and thousands of iterations, giving you the average, and accounting for the DPS loss by not casting something else.

General question about your simulator: Are you aggregating simulation results, or at least some key performance indicators, across everyone's sims? I think it would be interesting to see this data visualized since it's more representative of a broad range of gear and talent choices as opposed to doing min-maxed simulations like Simulationcraft. Thanks!

[Blayke]

I actually had the reroll count on that full blog post: http://blog.askmrrobot.com/2016/07/o...the-bones-rtb/ For a 450 second average fight length.

If you never reroll, you apply Roll the Bones 11 times in a fight. If you reroll for shark or any-3, you use it 30 times. So on average, you are rerolling twice (for a total of 3 rolls) to get either any shark combo, or any 3 buff combo.

But as people already pointed out, you don't know how many it will take when you start rolling. It could take 10 on one try, 1 on another. So it isn't helpful in the decision making process. But the simulator does this over thousands and thousands of iterations, giving you the average, and accounting for the DPS loss by not casting something else.

I understand that you don't know how many times you will have to reroll. What doesn't seem to be considered in all of this is that there is an opportunity cost to using 5 or 6 cp's on an ability that does no direct damage over one that does. So in my example--an admittedly random number, btw--I'm simply wondering if getting a SiW buff on the 10th roll is still a dps increase over 9 eviscerates, not too mention the loss of other buffs that are also quite useful, even if not necessarily as powerful as SiW.

Again, I understand we have no idea how many rerolls it takes to get what we want, but we DO know that every time we do, we have forfeited an eviscerate. So the question becomes, how many eviscerates can we sacrifice before it simply no longer worth it. Obviously that number must exist. It seems to me not unreasonable to suppose that there is some number between 1 reroll and some other, higher number, where the most prudent course of action is to spend the cp's on an eviscerate before resuming fishing for the "ideal" proc.

Not trying to be argumentative here, just questioning the notion that fishing for "ideal" procs is always a dps increase, no matter how many times we have to reroll to get them. Common sense would suggest that there is a number, beyond which it actually becomes a dps loss, and that perhaps that number can be figured out with modelling.

I also understand that this is largely academic since, in reality, SiW seems to come up with shocking regularity. But it would still be nice to know.

Also, just as an added note, bear in mind this is the pre-patch, Blizzard's version of limbo. Does anyone really doubt that come level 110 this will all be irrelevant?

[RobertoCarlos]

To take an extreme example just to make my point: what if you had to reroll 10 times? That's 50 (or 60) wasted cp's, plus the loss of the other buffs. Is SiW really so strong that it's preferable to 3 other buffs (we won't even count GM) and the loss of 10 eviscerates? Probably not.

I'm not a math expert, but I've played WoW for a long time, and getting over 80% crit by fishing for shark infested waters (excuse the pun) is probably going to be better than 10 run through's. Think about it, 45secs of nearly 90% crit maybe even higher in t19 on a boss fight where there is a burn phase or you get increased damage for example in an encounter. Bloodlust, potion. Its going to be crazy

[Zoopercat]

The AMR simulator is updated with the RtB changes for today's hotfix. Go to http://beta.askmrrobot.com to do your sims.

And with that, I tested out the various combos again and found that rerolling for ANY 2 buff combo is best. But having just shark infested waters is so close, you might as well keep it, though that's up to you. Here's the update blog post about it.



edit: updated image with Slice and Dice DPS from the buff.

- - - Updated - - -
@Tellof Interesting question. We have some batch simulations that will be looking at a bunch of different gear sets and analzying things on a very large scale. So stay tuned for that

[Varaben]

The AMR simulator is updated with the RtB changes for today's hotfix. Go to http://beta.askmrrobot.com to do your sims.

And with that, I tested out the various combos again and found that rerolling for ANY 2 buff combo is best. But having just shark infested waters is so close, you might as well keep it, though that's up to you. Here's the update blog post about it.



edit: updated image with Slice and Dice DPS from the buff.

- - - Updated - - -
@Tellof Interesting question. We have some batch simulations that will be looking at a bunch of different gear sets and analzying things on a very large scale. So stay tuned for that

Wow, so SND is actually pretty close now. That's <4% worse and you never have to reroll. That would be pretty boring, but it's one less thing to worry about I guess. I'm curious how that will change with higher gear.

[kindath]

I think those tests were done before the buffs to run through and between the eyes were added - but it doesn't make as much of a difference as you might think. Having 2+ buffs up lets you run through more often and/or makes run through hit harder.

[Zabatakis]

Wow, so SND is actually pretty close now. That's <4% worse and you never have to reroll. That would be pretty boring, but it's one less thing to worry about I guess. I'm curious how that will change with higher gear.

Well it is just a hair better than never rerolling bones, and in that case you would use it the same way (just keep it up, who cares what buff). Which would be ok if you didn't need a talent for it, or at least not the lvl 100 which to me should be better than that.

Well, I have no doubt they will tune it or figure something out.


Edit: clarity.

[Touch My Bankai]

Wow, so SND is actually pretty close now. That's <4% worse and you never have to reroll. That would be pretty boring, but it's one less thing to worry about I guess. I'm curious how that will change with higher gear.

Honestly, I can live with RtB, but I am NOT a fan of it... if they can get SnD to be somewhat comparable I'll be using that. Call it boring or whatever, but I prefer the steady-ness of it over the randomness of RtB

[Tindr]

Randomness of this skill makes me mad... Playing whole day in dungeons, HFC(hc and mythic) and I have 6 buffs maybe three times...

[Mamut]

Because you are not supposed to have 6 buffs all time, when you have two or crit one you are already good.

[Lunareste]

I'm probably gonna stick with RtB and only reroll when I proc Grand Melee by itself. I dislike the Rerolling aspect when I have to constantly do it =/

[shaunika123]

I'm probably gonna stick with RtB and only reroll when I proc Grand Melee by itself. I dislike the Rerolling aspect when I have to constantly do it =/

grand melee is the 2nd best roll now

you basically reroll till you get 2 buffs, which is like 40% chance afaik so it's pretty easy to do

[WskyDK]

grand melee is the 2nd best roll now

[citation needed]

I don't see it being better than any other buff
Edit: attack speed is still all but useless.

[shaunika123]

[citation needed]

I don't see it being better than any other buff
Edit: attack speed is still all but useless.

literally look like 6 posts above you? lol

[WskyDK]

literally look like 6 posts above you? lol

The chart?
That shows that GM is within the margin of error for solo buffs. It also shows the disparity between SiW and all the others.

If anything, that chart should encourage people to reroll more. With 4/5 solo buffs being basically the same dps, re-rolling can only benefit you at this point; you either do the same damage, or get more.

[shaunika123]

The chart?
That shows that GM is within the margin of error for solo buffs. It also shows the disparity between SiW and all the others.

If anything, that chart should encourage people to reroll more. With 4/5 solo buffs being basically the same dps, re-rolling can only benefit you at this point; you either do the same damage, or get more.

i didnt say you shouldnt reroll, but you reroll any 1 buff equally
grand melee isnt the worst to have

also since all buffs are about equal you still have to wage wasting a run through for it now

[WskyDK]

Hmmm, I'd still probably re-roll at least twice trying for a two buff before I worried about run-through