Restoration Itemization - Page 57

Drane · 09/16/09, 2:34 AM

Originally Posted by Arythorn

So, just saw my first 4 pc T9 in the wild (granted it was a mix of 2 pc 232 and 2 pc 245 T9 but 4 pc nonetheless). I'd love to see some live data of parses to see what kind of totals we are seeing on effective heal in TotC for this bonus. So, all you early adopters out there, let's get the information flowing!

Just got my 4pc T9 (All 4 triumph) literally an hour ago, here's me just healing a lifetapping warlock to make sure no ticks were overheal: World of Logs - Real Time Raid Analysis

My crit at that moment was 20.15%, so the proc rate was pretty close. I know its an extremely small sample size; I'll upload my 10/25 man parses throughout the week (or maybe just do some random heroic) if anyone wants to see a larger sample.

An interesting note, while doing a heroic previously (forgot to turn on combatlogging for that >.>) rejuv NEVER crit if the HoT was pure overheal. This is why I did the testing making sure it would never be overheal.

Also, Nature's Grace does NOT proc from rejuv critting.

Shelendil · 09/16/09, 4:38 AM

Originally Posted by Drane

Also, Nature's Grace does NOT proc from rejuv critting.

It appears Living Seed does not proc from it either. Is that correct?

ganuard · 09/16/09, 5:42 AM

Originally Posted by Drane

Just got my 4pc T9 (All 4 triumph) literally an hour ago, here's me just healing a lifetapping warlock to make sure no ticks were overheal: World of Logs - Real Time Raid Analysis

My crit at that moment was 20.15%, so the proc rate was pretty close. I know its an extremely small sample size; I'll upload my 10/25 man parses throughout the week (or maybe just do some random heroic) if anyone wants to see a larger sample.

An interesting note, while doing a heroic previously (forgot to turn on combatlogging for that >.>) rejuv NEVER crit if the HoT was pure overheal. This is why I did the testing making sure it would never be overheal.

Also, Nature's Grace does NOT proc from rejuv critting.

Sweet, someone has it. You should get some larger samples for us to tear apart. I should have my fourth piece of T9.5 tomorrow, so I should be able to add to the discussion as well.

I don't know about you guys, but seeing Rejuv tick for 3.4k makes me.. uh.. well you know. When I get it, along with a few other small changes I'll be making in the near future, I'll be at 360 haste rating and will have just shy of 31% crit in a raid. Can't wait to see the results. I'll also be finishing up my Val'anyr next week, and something about critting rejuvs and Val'anyr just seems too good.

Arythorn · 09/16/09, 9:56 AM

Originally Posted by Drane

Just got my 4pc T9 (All 4 triumph) literally an hour ago, here's me just healing a lifetapping warlock to make sure no ticks were overheal: World of Logs - Real Time Raid Analysis

My crit at that moment was 20.15%, so the proc rate was pretty close. I know its an extremely small sample size; I'll upload my 10/25 man parses throughout the week (or maybe just do some random heroic) if anyone wants to see a larger sample.

An interesting note, while doing a heroic previously (forgot to turn on combatlogging for that >.>) rejuv NEVER crit if the HoT was pure overheal. This is why I did the testing making sure it would never be overheal.

Also, Nature's Grace does NOT proc from rejuv critting.

All good stuff -- definitely looking forward to seeing the in-raid parses. From the very limited data, even at a low unbuffed crit rating of 20% (actually only 18.5% crit realized in the parse), the 4 pc bonus uplifted your total Rejuv output by ~8.33%. I used =((3473.2-2328)*12)/165061) as the formula to arrive at that percentage. That's a nice uplift -- if we start looking at raid buffed crit and a strategy of stacking crit beyond 360 haste, no reason we couldn't get to 30% crit raid buffed. So add another 50% to that and perhaps it's a 12% RJ uplift. If I take it one step further and apply that 12% uplift to Rejuv to a parse from last night where Rejuv made up 42% of my healing (and that coincidentally / conveniently has a good cross-section of of Koralon, ToC normal, ToC heroic, and Ulduar 1 Light Yogg World of Logs - Real Time Raid Analysis), we get a number of ~5.0% effective healing added by 4 pc T9 versus the ~11% effective healing that pre-nerf 4 pc T8 generated on this same parse.

Now that is a heavy amount of extrapolation based on a very limited sample size with no overheal so i think we can call that ~5.0% effective healing granted by 4 pc T9 officially SWAG (sophisticated wild-ass guess) material but it is interesting nonetheless. This would put it fairly close to a post-nerf T8 which should be checking in around ~5.5%. At that level of parity, I could see moving forward with a mix of 232/245 T9 4 pc instead of my prior premise of needing to wait until 258. Very interesting indeed. If you can't tell, I'm itching to put a reasonable value to 4 pc T9 so we can better compare it to T8 and I can make an informed decision about using 232,245,258 T9. Don't beat me up too badly on over-extrapolation.

Drane · 09/16/09, 10:15 AM

Originally Posted by ganuard

Sweet, someone has it. You should get some larger samples for us to tear apart. I should have my fourth piece of T9.5 tomorrow, so I should be able to add to the discussion as well.

I don't know about you guys, but seeing Rejuv tick for 3.4k makes me.. uh.. well you know. When I get it, along with a few other small changes I'll be making in the near future, I'll be at 360 haste rating and will have just shy of 31% crit in a raid. Can't wait to see the results. I'll also be finishing up my Val'anyr next week, and something about critting rejuvs and Val'anyr just seems too good.

Yea, I was pretty excited about getting Val'anyr and 4pc in back to back weeks

Also, living seed does NOT proc off it as well. It seems that the 4pc just acts as a 1.5x modifier for the tick; nothing that procs from crits procs from the rejuv tick critting.

I have class till ~2 CST, then I'll definitely go run some 5mans and get a good sample with regular overheal, tank healing, etc.

(I have Heroic Jaraxxus/Faction Champs/Twin Valk attempts tonight, so I'll definitely upload WoL after those)

Arythorn · 09/16/09, 11:15 AM

Originally Posted by Drane

while doing a heroic previously (forgot to turn on combatlogging for that >.>)

Also, all you T9 4-piecers, go get Loggerhead -- Curse Download and set it for EVERYTHING. No more missed parses I tell you!!

Drane · 09/16/09, 11:20 AM

Originally Posted by Arythorn

Also, all you T9 4-piecers, go get Loggerhead -- Curse Download and set it for EVERYTHING. No more missed parses I tell you!!

Yea, I have loggerhead since I do the uploading for my 10/25 man runs, I just don't have any Northrend heroics enabled so I don't have to go into the log and delete those when I upload the night's raids >.>

Arythorn · 09/16/09, 11:26 AM

Yeah, I don't have mine set to include Heroics either -- so the same thing would've bit me in the arse as well. I'm mostly just adding some levity . . . but change it now!!

Drane · 09/16/09, 4:23 PM

Heroic Nexus running with 30.88% crit (22.56% unbuffed plus MotW, Elemental Oath and Moonkin Aura): World of Logs - Real Time Raid Analysis

Gonna try to run VoA25 as well, before tonight's raid.

Arythorn · 09/16/09, 5:07 PM

On the above parse (World of Logs - Real Time Raid Analysis) as it stands now, at 29.7% actual crit rate, the 4 pc bonus is about ~8.75% uplift to Rejuvs (versus the ~12% I predicted at ~30% crit) and 4.4% uplift to total healing on that heroic versus the ~5.0% I predicted.

It's worth noting that on the prior parse with no overheal, your average crits were valued at a clean 1.50 your average tick as expected (avg crit 3473, avg tick 2328 = 3473/2328 = 1.49. On this parse, your average crits are valued at about 1.30 your average tick which was not expected (average crit is 871, avg tick 667 = 871/667 = 1.3)

This perplexed me a bit but I added in another parse that you have and my synapses begin to fire a bit (been a long day). On this parse World of Logs - Real Time Raid Analysis, we get even another flavor. Crit rate = 31.7%, Avg crit = 1676, avg tick = 961 which values the average crit at 1.75 the value of average hit. On this parse, the crits were almost a ~25% uplift to Rejuv and a ~13% effective heal uplift.

This wide variation really comes down how many crits get absorbed into overheal and how many normal ticks get absorbed into overheal. It'll really take a view (or several views) of Heroic ToC to get a real feel for how crits and ticks are fitting into that damage profile. Anyway interesting to see how overheal starts to mess with the valuation as in the first parse, this was a completely absent factor.

ganuard · 09/16/09, 5:15 PM

Originally Posted by Arythorn

On the parse as it stands now, at 29.7% actual crit rate, it's about ~8.75% uplift to Rejuvs (versus the ~12% I predicted at ~30% crit) and 4.4% uplift to total healing on that heroic versus the ~5.0% I predicted.

However, something is afoot at the Circle K. On the prior parse, your average crits were valued at about 1.50 your average tick as expected (avg crit 3473, avg tick 2328 = 3473/2328 = 1.49. On this parse, your average crits are valued at about 1.30 your average tick which is NOT expected (average crit is 871, avg tick 667 = 871/667 = 1.3). Something doesn't jive here. Also the size of the ticks / crits are substantially down from what would be expected. I'm a bit perplexed to be honest.

Sadly we wont get anything useful from a heroics parse simply because there is no damn damage going around. That said, if you use the log browser and set the query to show heals from the spell rejuv, you'll see that it was healing for the correct amount, but it appears the average EFFECTIVE heal, both from regular ticks and crits, were pretty close to what was listed on the healing by spell tab. I suspect World of Logs is causing the confusion here.

A quick bit of the log:
[15:04:37.447] Draenna Rejuvenation Draenna +0 (O: 2741)
[15:04:38.071] Draenna Rejuvenation Drandar +206 (O: 2534)
[15:04:38.583] Draenna Rejuvenation Korgal +*354* (O: 3756)
[15:04:40.971] Draenna Rejuvenation Drandar +0 (O: 2740)
[15:04:44.017] Draenna Rejuvenation Drandar +*2457* (O: 1653)
[15:04:46.940] Draenna Rejuvenation Drandar +*3879* (O: 232)
[15:04:52.403] Draenna Rejuvenation Drandar +2741
[15:04:54.494] Draenna Rejuvenation Harazh +*3193* (O: 918)
[15:04:55.417] Draenna Rejuvenation Drandar +2741
[15:04:55.652] Draenna Rejuvenation Xiah +0 (O: 2740)
[15:04:57.416] Draenna Rejuvenation Harazh +0 (O: 2741)
[15:04:57.925] Draenna Rejuvenation Draenna +0 (O: 2741)
[15:04:58.461] Draenna Rejuvenation Drandar +*4111*
[15:04:58.609] Draenna Rejuvenation Xiah +0 (O: 2741)
[15:05:00.454] Draenna Rejuvenation Harazh +0 (O: 2740)
[15:05:00.898] Draenna Rejuvenation Draenna +0 (O: 2741)
[15:05:01.419] Draenna Rejuvenation Drandar +2740
[15:05:01.695] Draenna Rejuvenation Xiah +0 (O: 2740)
[15:05:03.432] Draenna Rejuvenation Harazh +0 (O: 2740)
[15:05:03.806] Draenna Rejuvenation Draenna +0 (O: 2740)
[15:05:04.509] Draenna Rejuvenation Drandar +2740

Arythorn · 09/16/09, 5:32 PM

World of Logs did have me confused for a bit as I hadn't really used it for this exact purpose before. But, after viewing a second parse with 1.75 value of average crit to average tick, my brain kicked in.I figured that out and edited my original post.

Drane · 09/16/09, 5:33 PM

Yea, I was looking over the average ticks/crits and wondering why they were so low, then realized overheal.
(In my defense, I did tell Drandar to take as much damage as possible, he was pulling 4-5 groups at a time >.>)

We apparently won't have VoA before our raid starts in 2 hours, but I can keep a live log running during the raid if you want to check our wipes/kills as we go.

Arythorn · 09/16/09, 5:36 PM

I'll be stopping to eat dinner with the wife and kids here in a moment but yes, definitely post the results from raids tonight. Val'anyr will likely skew the flavor of things a bit as well but it's definitely interesting stuff to see.

EDIT: I fully expect average crit to gravitate toward 1.5 x average tick even after allowing for the variation of which ticks / crits got absorbed into overheal. On small data collections, we see that skew down to 1.3 and up to 1.75 but law of averages and a larger sample size will smooth that variation out. What it will really come down to is how much is RJ generally contributing to our effective heal output given the damage profile of ToC.

The more I think about it, I should be able to build a table pretty quickly that correlates crit rate and % RJ plays as part of your overall effective heal and generates a workable average for 4 pc T9 given those two criteria. For instance, if we feel RJ will be around 40% of our effective heal output and we use 30% crit as a workable crit number, in a large enough sample size, we should expect 4 pc T9 to value out at somewhere right around of total ~5.25% effective heal in ToC. If I instead used a 27.5% crit rate but stated that RJ would make up 55% of my effective heal, I would get a number closer to 6.65% overall effective heal from the T9 bonus. Near as I can tell crit rate and % that RJ plays of overall effective healing are the two big things that will slide the valuation of this bonus up and down. Overheal, etc. will even itself out over a large enough sample.

EDIT 2: Actually have a workable model now where I've done the above. Base on some initial things I've done, if I state that RJ makes up 40% of my effective healing, I would expect each 1% of crit to buy me an additional .20% more effective healing. So 5% crit to get 1% more effective healing in that scenario. If I up the average percentage rejuv makes up of my effective healing to 50%, each point 1% of crit buys me and additional .25% of effective healing. So only 4% crit required for 1% additional throughput.

Starting to look at it this way will also let us evaluate crit on gear as well as how valuable 1/3 to 3/3 Natural Perfection might be in any talent spec moving forward. Again, we'll see how the data looks but given a large enough sample size, I suppose the gravity of calculation will eventually all model toward a predictable value given crit % and RJ % of effective overall healing.

Mind you, I'm a IT professional not a mathematician but seems to make logical sense to me as I put some critical thought toward solving the value.

Drane · 09/16/09, 5:39 PM

I'm HOPING we blow through heroic Jaraxxus/Faction Champs tonight to get more Twins learning, since that will be the best place for 4pc testing, even with our 2nd resto druid healing.

Live log tonight will be updated here: World of Logs - Real Time Raid Analysis

Even after one attempt on Jaraxxus I'm beginning to agree with Playered/Paininabox- I kept finding myself trying to snipe a 2-3k deficit on someone, only to find that rejuv to be completely overheal. (On the Jaraxxus kill, for example, our holy priest was on the ball with CoH sniping, and managed to beat my effective healing)

I obviously haven't done enough heroic bosses to make a complete observation, however.