Mostly for Anvil and the healers, and I didn't want to put this as an in-game mail.
I ran Sanz through some Patchwerk-style simulations and tweaked the alchemy buffs. 1000 iterations, and sorted by dtps (I'm a tank after all!)
Here's the results:
Str flask:17129 dps 30657 dtps 18749 hps 132.1k TMI +6.5%
+ armor prepot
Str flask:17014 dps 30707 dtps 18671 hps 132.0k TMI +5.8%
+ str prepot
Str flask:16924 dps 31038 dtps 18549 hps 133.0k TMI +5.2%
Stm flask:16255 dps 31349 dtps 17804 hps 126.3k TMI +1.1%
+ armor prepot
Stm flask:16148 dps 31467 dtps 17743 hps 126.0k TMI +0.4%
+ str prepot
Baseline: 16078 dps 31683 dtps 17612 hps 134.7k TMI -
Stm flask:16053 dps 31716 dtps 17601 hps 126.9k TMI -
I suspect the baseline isn't at the bottom simply because of margin of error.
Question for you healers out there: the stamina flask gives +30K HP. Would you prefer an extra 30K shock buffer or 1K less damage/sec?
I'd favor best DPS in most fights, even for tanks. Shorter phases and shorter fights generally mean less chance of mistakes, the primary source of wipes. Based on those SimC numbers I'm not seeing a huge change either way in DPS, dtps, or TMI. IMO this makes the small boost to DPS worthwhile.
When there's significant chance of tank deaths I'd generally favor Stamina especially if it pushes you from 2 shot to 3 shot territory.
In farm content we've got ridiculous levels of overhealing, partly because of the absorbs interaction and because we're geared to the point of having room for small mistakes.
In progression content we need to be able to absorb mistakes made by tanks, healers and DPS. Most healing comes in small chunks every 1-2 seconds. We may need to send once of those chunks to the raid instead of the tank, we need the tank to stay up until we can return attention and send another GCD your way.
Imperator is kind of a special case, we've got a long fight with very low raid damage and then a final phase asking for maximum healing output and minimal mistakes. I suspect even here the difference between mitigation and DPS for tanks would be small compared to effective play overall as a raid and extra DPS might just be what it takes to get through the final phase.
The other night we discussed flasks being about 4% DPS boost and pre-pot + pot during fight being about another 4%, has anyone dug in deeper on those numbers?
It's tricky in the simulator because potions can be triggered for DPS events (bloodlust) and survival events (low health). I'll do some edits on the table so the increase from baseline is more obvious.
Based on a few other simulations:
Strength flask alone: +5.2%
Pre-pot armor and pot: +1.3%
No pre-pot, just pot: +0.6%
Some rounding error there, the impact of pre-pot and pot look to be equal. It hopefully has a bigger impact on DPS than tanks. Interested in me running the numbers with mutagen as the baseline?
In farm content we've got ridiculous levels of overhealing, partly because of the absorbs interaction and because we're geared to the point of having room for small mistakes.Should we go with two healers, moar dots?
I'd rather increase the number of competent DPS. Dropping 33% of our healing (3 to 2) leaves little margin for error, inflating DPS from 5 to 9 doesn't feel so bad (and lets us drop sub-par performers when pugging, which is really another conversation).
When I'm speaking of ridiculous amounts of overhealing, it is about the amount that would allow dropping 3 to 2 in a ~10 man if everyone was playing very well. Partially that's the stupid situation Disc is in, I'm blowing up meters as Disc while Rada, Hobugz, and Stonestorm are putting up more raw healing but raid is already at 100%, resulting in 30% - 50% overhealing for them. Problem is when we need 100% from all 3 healers during heavy damage phases (Imperator P4, dropped balls at Margaret, lotsa stacks at Butcher, fire bathing at Twins, etc)
Pre-potting increases your primary stat by 20% during a KEY phase of the fight. In short, it's at least 2% of your overall dps on a 5 minute fight, and can be a lot more if you are matching it up with dps cds and bloodlust. Remember that all of your stats interrelate in a multiplicative way, meaning that having 1500 extra Attack Power (given by any pot) will increase the value of your Haste, Crit, Multistrike etc which are all based on your AP.
Reviewing the logs from our last Kargath Heroic. We could improve on use of pots.
That is included in the simulation setup. For DPS, pots are used at the max_dps stage (heroism/et al). The tank simulation is a bit different. It can be used at max_dps or max_survival. I'm not sure how the simulation decides when to use it; it may be the first event. Look at the table above -- unless the simulations are set up wrong and/or usually getting triggered during max_survival, it provides 0.6% additional DPS. I don't think anyone is questioning whether we should prepot more. The question is "how beneficial is it?"
"Someday soon" I'll run the same sims with a DPS class to see whether there's a bigger difference with the pewpew classes. I think there will be.
Oooh... same day turnaround on someday! :)
I ran a sim using Mutagen. More on the sim below.
Baseline (pot + prepot + flask): 25918 dps (+7.8%)
Flask + pot (no prepot): 25688 dps (+6.8%)
Flask only: 25318 dps (5.4%)
Newb: 24030 dps
The DPS benefit of the potion is better for DPS (at least Mutagen) than tank (at least Sanzaru). Fortunately, that's what I expected.
Now as for the sim: it is a Patchwerk-style fight (tank and spank) with minimal other mechanics to worry about. IMO it's the easiest to sim, though other options (e.g. heavy movement) can be done too. The fight is set up to be 450 seconds +/- 20%. Just for fun, I did a sim of Muta on a 1.5 minute fight to maximize the pre+pot impact (29.9K dps). As you can imagine, a 50% uptime on a potion has a big impact.
But on a 450 second fight, that 25 seconds has a lot less impact (25/450->math).
Now, I'm not saying "don't prepot" I'm only answering the question from guild chat of "how much impact does prepots and flasks have?"
Sanzaru, I think the difference is largely impacted by stacking cooldowns, which depends on the talents you use for a fight. I simmed my setup with seraphim/holy avenger. With pre-potting and potting, it sims out at 21217 dps, including a 42k dps burst on the pull. Taking out the pots, that burst drops to 34k and the overall dps is down to 20531, for a 700 dps loss.
Of course, it's going to take some more practice for me to hit any of those numbers.
Here is a log of a paladin tank playing heroic butcher the way I dream about doing it.
https://www.warcraftlogs.com/reports/XQxZtRa4zGrw721f#fight=4&type=casts...