Feral: An in-depth look at energy pooling, AMR and SimC

[Zoopercat]

Hey everyone. Swol, Ask Mr. Robot's main theorycrafter, did a deep dive into a few things that get brought up about AMR and SimC. One of the main things that was brought up recently, was a discussion on energy pooling.

Swol wrote a blog post about it - it's long, but it's easy to follow along. He gives simC config files so anyone can verify his work, or build upon it.

The TLDR, based on his tests (in Simc), pooling for rip is not crucial to DPS. He also shows that AMR and SimC match very well in both simulators, so you can use whichever one you prefer - both are accurate.

Please read the blog post before responding.

- - - Updated - - -

Update: swol updated part of the article.

By request, I will clarify the main difference between the two APLs, and why I am calling the SimC APL “Pooling” and the AMR APL “No Pooling”.

The SimC APL specifically pools energy before using any finisher. The only time a low energy Rip could be used is if Rip is not ticking or Rake will fall off in less than 1.5 seconds. The only time a low energy Savage Roar can be used is if Rake will fall off in less than 1.5 seconds. Ferocious Bite is held as an energy dump and used only when in danger of capping energy or when you have Berserk or will soon use Tiger’s Fury.

The Swol/AMR APL places no energy restriction on using Rip or Savage Roar. The only energy restriction placed on Ferocious Bite is that you must have enough for a max damage Ferocious Bite (50 Energy or 25 with Berserk). Ferocious Bite will not be used unless there is more than half a Rip left on the target. So, there are a few times when some passive, incidental pooling will occur if you have more than 12 seconds left on Savage Roar and less than 8 seconds left on Rip, but more than 4.8 seconds left on Rip. It will wait until Savage Roar drops below 12 seconds or Rip drops below 4.8 seconds to act.

If you want to convince yourself that no energy pooling is occurring in the Swol/AMR APL, add this line after Ferocious Bite:

actions.finisher+=/rip,if=(energy>=50|buff.berserk.up&energy>=25)
That line will ensure that you never sit on your combo points for any reason. It doesn’t change the damage of the simulation because the case in which you would actually wait is so rare that it falls beneath even 0.1% margin of error.

Some confusion was being caused because I had added a line to my APL that was pooling energy once in a while when the conditions for Rip and Savage Roar were not met. I have removed that line from the APL and updated all of the tests in this blog post using the updated APL in the config file below that definitely does not pool energy.

[Heywes]

That was long but very interesting.

[Deleted]

I dont want to be the lazy guy but can some1 explain to me what the pooling discussion was about? Pool energy while having rip on the target? What would be thw point of doing it? As long as we maintain buffs and rip/rake it seems to be meaningless if we spend rhe energy we get or wait for 100 energy before spending .Pooling because of what and when?

Greetings

[Deleted]

Because Ashamane's Bite. You want to use as many generators as possible when your rip uptime is still long to get most out of that trait.

[Nfinitii]

I dont want to be the lazy guy but can some1 explain to me what the pooling discussion was about? Pool energy while having rip on the target? What would be thw point of doing it? As long as we maintain buffs and rip/rake it seems to be meaningless if we spend rhe energy we get or wait for 100 energy before spending .Pooling because of what and when?

Greetings

The point was to pool nrg before u cast rip so u can use as many combo point generators as possible after so ashamane's bite passive could copy rip at higher lengths to maximize dps. The article talks about pooling and not pooling and comparing the numbers. Tldr the difference r so small it doesn't matter.

[Deleted]

Simcraft APL Swol shared is amazing. It sims soul of the forest higher than incarnation with default simc APL.

[Thorim]

I've found SimC and AMR aren't very similar. SimC repeatedly shows different sets of gear/especially trinkets to be a better choice than AMR does..and when using the SimC suggestions in raids, I have always done better than with the AMR suggestions. One example is that out of the gloves/bracers/chest/waist legendaries...it suggests using the chest and waist lol I've done multiple raids trying variations of these and the bracers/gloves always outperform the chest which gives *nothing* besides stats?! Why on earth would the chest *ever* be better than the gloves or bracers?

I also don't really see how pooling can't be important? Having say, 20% uptime on AB vs 45% uptime is going to be a huge difference.

[Zoopercat]

@Thorim - when you are looking at a ranked gear list on AMR, did you use custom weights, generated by our simulator (the ones that take an hour to run)? If not, what you are doing is comparing a default list (like a list from a guide, that isn't custom to you), with SimC results, which are custom. So of course the custom ones would be better

Lastly, a few runs to test items doesn't work. What if you got a lucky crit streak? What if the boss died faster (your DPS is higher on shorter fights). A simulator can run 1000s of iterations to smooth out those variables, which is why people use them to test out gear ideas. 1 attempt on a boss is not a statistically significant set of trials

To run a custom setup on AMR, follow this guide (The top half): http://i.imgur.com/r2gpuBW.jpg

Also, did you read the full article posted? It talks specifically about pooling and AB uptime.

[Thorim]

@Thorim - when you are looking at a ranked gear list on AMR, did you use custom weights, generated by our simulator (the ones that take an hour to run)? If not, what you are doing is comparing a default list (like a list from a guide, that isn't custom to you), with SimC results, which are custom. So of course the custom ones would be better

Lastly, a few runs to test items doesn't work. What if you got a lucky crit streak? What if the boss died faster (your DPS is higher on shorter fights). A simulator can run 1000s of iterations to smooth out those variables, which is why people use them to test out gear ideas. 1 attempt on a boss is not a statistically significant set of trials

To run a custom setup on AMR, follow this guide (The top half): http://i.imgur.com/r2gpuBW.jpg

Also, did you read the full article posted? It talks specifically about pooling and AB uptime.

I sim my character in SimC and also use AMR to get custom stat weights tailored for my character, I'm not sure why you would assume (and spend most of your reply hammering it home in kind of a dickish way) that I'd just use the default weights when comparing gear-sets. so how would your simulator tell me that the legendary chest was better than the bracers/gloves?

Multiple raids, hundreds of mythic+, hours on the dummies and dozens of sims/showing real-world comparisons between SimC suggestions and AMR suggestions repeatedly give me better gear-sets and closer matching SIMDPS in SimC's camp, sorry. It would be interesting to kow if you raid/do high level mythic+ as a feral druid? Because all of the top logging ferals/feral-specific theorycrafters (like on fluiddruid) kind of disagree with you in terms of gearing/play-style.

[Zoopercat]

Hey, I didn't mean to sound rude, I'm really sorry about that

I had run a custom gearing strat before asking and didn't see the same suggestions you were, so that's what lead me to ask that. When I use the strat I did for you, I see the waist and feet being recommended, not the chest.

Waits + Feet sim: 670,964 DPS
Waist + gloves: 655,366 DPS
Waist + chest (which I don't see recommended for you with custom gear): 626,271 DPS

So the simulator agrees, the chest isn't good A custom stat weight set wouldn't rank it higher. Can you link me the custom one you ran (the one that takes about an hour to run?). And also tell me what settings you chose? For example, I ran this one for Ursoc, if you ran it for AoE or have other different settings, maybe that's what you saw?

As for the legendary question: when you run custom stat weights on AMR, we also sim every single item with a proc. Then that information gets bundled in the stat weight report, and then those items are ranked directly from simulations. Which is why the custom rankings are actually really accurate (assuming no bugs, of course). There is about 1% margin of error from the suggested DPS of a Legendary to what the sim might show - a tradeoff of 1% accuracy for generating a list in less than 1 second.

As for the feral theorycrafters you mention: we addressed that in a recent blog post. If you haven't seen it yet, it's worth checking out http://blog.askmrrobot.com/feral-amr-simc-pooling/

[Thorim]

Hey, I didn't mean to sound rude, I'm really sorry about that

I had run a custom gearing strat before asking and didn't see the same suggestions you were, so that's what lead me to ask that. When I use the strat I did for you, I see the waist and feet being recommended, not the chest.

Waits + Feet sim: 670,964 DPS
Waist + gloves: 655,366 DPS
Waist + chest (which I don't see recommended for you with custom gear): 626,271 DPS

So the simulator agrees, the chest isn't good A custom stat weight set wouldn't rank it higher. Can you link me the custom one you ran (the one that takes about an hour to run?). And also tell me what settings you chose? For example, I ran this one for Ursoc, if you ran it for AoE or have other different settings, maybe that's what you saw?

As for the legendary question: when you run custom stat weights on AMR, we also sim every single item with a proc. Then that information gets bundled in the stat weight report, and then those items are ranked directly from simulations. Which is why the custom rankings are actually really accurate (assuming no bugs, of course). There is about 1% margin of error from the suggested DPS of a Legendary to what the sim might show - a tradeoff of 1% accuracy for generating a list in less than 1 second.

As for the feral theorycrafters you mention: we addressed that in a recent blog post. If you haven't seen it yet, it's worth checking out http://blog.askmrrobot.com/feral-amr-simc-pooling/

What ilevel range did you select? I'll do another run but I did a 1hr one only last week. I've had a lot of odd things happen recently with AMR, like suggesting off-tier pieces rather than the 4pc, different trinkets than what I know are better etc. For NH the 4pc and bracers clearly come out on top.

I still don't understand how letting rip tick down with no energy (because you didn't pool before hand) is better/even with pooling to get a full duration rip up then the maximum chance of procing AB though.

[Zoopercat]

I set it from 880-920 (you want it a little below and a bit more above than your current gear, or anything you are targeting in a BiS). I used ursoc, potion, flask, augment rune. I used the default ursoc fight length (which is a bit short, if you run it again, you could set it to the length of your average fight in NH these days).

Also, change the 'bloodlust mode' to 'rotation' - that uses it at the start of the fight. When it is set to boss, the boss script can use it at the end, if that's what the strategy calls for.

You can see what custom weights look like from Krosus, however, that's still in beta. It is looking good for most specs, but I am not done fully vetting that.


As for an explanation for the rip thing beyond what the article says, I'll ask Swol to stop by and add more commentary on that. (I'm not a feral expert, just a simulation expert).

[Thorim]

I set it from 880-920 (you want it a little below and a bit more above than your current gear, or anything you are targeting in a BiS). I used ursoc, potion, flask, augment rune. I used the default ursoc fight length (which is a bit short, if you run it again, you could set it to the length of your average fight in NH these days).

Also, change the 'bloodlust mode' to 'rotation' - that uses it at the start of the fight. When it is set to boss, the boss script can use it at the end, if that's what the strategy calls for.

You can see what custom weights look like from Krosus, however, that's still in beta. It is looking good for most specs, but I am not done fully vetting that.


As for an explanation for the rip thing beyond what the article says, I'll ask Swol to stop by and add more commentary on that. (I'm not a feral expert, just a simulation expert).

I did another one and got different weights, and it still suggests the chest, probably because of the differing options.

However, I used your weights and BIB.......and it suggests I use my eye of guarm and Chaos talisman....*over an 895 convergence of fates!*, that is just........crazy lol COF means berserk has a 1.5-2 min CD, so an averageish fight of 6 minutes is a full extra berserk, see this is one of the reasons i don't trust these things. It also still suggests the chest by the way - even using your sim lol Oh and it suggests the soldier enchant on the neck not satyr...which...is just wrong.

[Zoopercat]

Can you send me the BiB string you are using? That will let me test this out specifically against your bags. Appreciate it.

[Swol]

@Thorim

The article I wrote specifically addresses whether pooling energy before using Rip significantly increases Ashamane's Rip uptime. The findings showed that you don't get a significant increase in the uptime. Depending on what legendaries you have, the uptime difference between pooling before Rip and not is in the 1-4% range.

Regarding the ranking of cooldown reduction effects such as convergence of fates: the duration of the fight you simulate will have a large impact on the ranking, especially for a long cooldown ability like berserk. If you use the 210 second Ursoc script to make the stat weights, CoF could easily rank very low, since you won't end up getting more Berserk uses even with the CDR. Increasing the fight length to 300 or 360 seconds will see that you get more use out of it. The scripts include a default fight length variance of 12.5%. Some of those fight durations will get another Berserk use, some won't. The result you see is then an average.

If you run a gearing strategy with two specific legendary items already equipped, you are essentially telling the simulator that you are going to use those legendary items. When you are looking at a BiS list, you should just lock in the two legendary items you had on when you made the gearing strategy.

[Thorim]

@Thorim

The article I wrote specifically addresses whether pooling energy before using Rip significantly increases Ashamane's Rip uptime. The findings showed that you don't get a significant increase in the uptime. Depending on what legendaries you have, the uptime difference between pooling before Rip and not is in the 1-4% range.

Regarding the ranking of cooldown reduction effects such as convergence of fates: the duration of the fight you simulate will have a large impact on the ranking, especially for a long cooldown ability like berserk. If you use the 210 second Ursoc script to make the stat weights, CoF could easily rank very low, since you won't end up getting more Berserk uses even with the CDR. Increasing the fight length to 300 or 360 seconds will see that you get more use out of it. The scripts include a default fight length variance of 12.5%. Some of those fight durations will get another Berserk use, some won't. The result you see is then an average.

If you run a gearing strategy with two specific legendary items already equipped, you are essentially telling the simulator that you are going to use those legendary items. When you are looking at a BiS list, you should just lock in the two legendary items you had on when you made the gearing strategy.

Thanks for the reply.

Any idea why AMR would suggest soldier over satyr?

[Swol]

By default we recommend Satyr, but Soldier isn't massively far behind. I'd have to see the gearing strategy you created to make a guess as to why it came up as soldier. Maybe your gear is making Mastery a bit stronger because of some combination of special items. It's entirely possible that the values of those two enchants could creep into the margin of error range.

[Zoopercat]

@Throrim, In your case, the chest piece is better. Her's why:

1. If you don't equip the legendary chest, your next best item is the 885 Chestguard of Insidius Desire. This is a loss of 1264 agility. 605999 DPS
2. If you don't equip the legendary gloves, your next best item is the 910 Temporally Displaced Gloves. This is a loss of 576 agility. 602,539

If you look at those 2 reports, you can see under "other" that the gloves got you more combo points from primal fury. But with the chest piece, you can see that Rake and Rip (for example) are hitting a lot harder because of the extra agility.

This is where best in bags and simulators comes in handy - the combinations really matter. Static legendary rankings are just a base guide that can't take your character into account.

Here's another report that does the combinations for you, to see how they stack up.

[Zanzha]

Realistically though the gloves are still the better choice in almost all situations.

This PR drivel on the other hand is just a farce, players?maybe not even that, with minimal experience with the class essentially trying to tell us that we've had it wrong for years. Pooling with feral is nothing new in legion and has been a core part of playing the class for since it's conception, if for anything because of the part it plays in maintaining good uptimes. To even dare suggest that it is of no consequence is borderline idiocy.

I mean, there's a reason you guys got kicked off the discord for spreading misinformation, changing the platform / audience doesn't make it any less incorrect.

You clearly have no interest in actually helping or interacting with the community or so to speak in any way and only crop up to shill your broken product when someone badmouths it.

[Thorim]

Realistically though the gloves are still the better choice in almost all situations.

This PR drivel on the other hand is just a farce, players?maybe not even that, with minimal experience with the class essentially trying to tell us that we've had it wrong for years. Pooling with feral is nothing new in legion and has been a core part of playing the class for since it's conception, if for anything because of the part it plays in maintaining good uptimes. To even dare suggest that it is of no consequence is borderline idiocy.

I mean, there's a reason you guys got kicked off the discord for spreading misinformation, changing the platform / audience doesn't make it any less incorrect.

You clearly have no interest in actually helping or interacting with the community or so to speak in any way and only crop up to shill your broken product when someone badmouths it.

I know a lot of 'common sense' things sometimes turn out to not be true, but I still cannot wrap my head around how pooling energy and then getting the maximum number of CP generating abilities off in quick succession on a high duration rip would just not be absolutely better than this blogs suggestion that pooling doesn't matter.....that's the one thing I was/am having the hardest thing believing. Getting multiple 20, 18, 15 etc second ashamane's bite procs is *surely* better than being energy starved and maybe having it proc at 12, 10 etc seconds.......

That's the big reason I was arguing against OP's blog post. Are they just basically BSing Zanzha?