Something you may want to consider is:
If CH is hitting 2,5 targets for the earthliving proc, that is probably a raid wide damage or partially raid wide.
In that case, HST overheal should be less than 75%.
On the other hand if the raid is getting some spike damage, the time between the dmg and the raid is getting topped shall not allow more than 2 ticks of earthliving procs, which is between 6 and 8 sec. So probably most of the procs will be overheal (50% overheal for a single proc)
Even more, in that case, only the first CH cast procs will be able to tick more than one time
If the raid is taking more time than that to get topped, probably something wrong is happening.
With the HST you will get the same ticks numbers but in 5 persons at the same time, a 4 procs for a single CH cast is very rare.
To me sounds better the HST glyph than Earthliving one, but is all about your play style
I had the same idea, when Riptide came out and posted the upstanding image into this board, but this makes just sense in some situations. I most situations where you can't preview raiddamage and incoming heal, you shouldn't build up a static healing-target-order just bump up your ch-amount. In many cases you would have to heal targets, that have full hp instead of people that might need heal, just to keep up your order - that would be absurd (that's what I was talking about in the posting before you ;D)
But still, if you are able to survey your ticking riptide-hots it's a good idea to turn them into huge CHs, when this target gets damage.
Something you may want to consider is:
If CH is hitting 2,5 targets for the earthliving proc, that is probably a raid wide damage or partially raid wide.
In that case, HST overheal should be less than 75%.
CH is a smart heal, and hits those that need heals. The ELW proc will proc on those that did take damage. CH may have healed them to full.
HST is party-only. If half the raid takes damage, and the other half is still at full, HST will do 50% overheal, on average.
Assuming CH & ES for 2 of our glyphs, the debate seems to be between EL & HS for the 3rd. As Skyhoof showed, HS glyph is 83 hps, and at a reasonable amount of haste, EL glyph is 172 hps. Both spells seem to situationally overheal between 50% and 75%.
I've been using glyph of HW as an experiment recently. This is a recent enough venture that I only have data from one boss kill. It was a 4HM "And they will all go down together" achievement kill, and I was healing in the back of the room where the dps spreads out preventing CH jumps. For that fight, the glyph was 191 hps, at 65% overheal, making it a competitive alternative.
There are a few spots where the RT bonus to CH can be effectively used. First, consider that if CH can be cast effectively (hit multiple targets), then cast it instead of RT. There are a few current encounters where RT can be used to build up CH. In general, RT is useful when you cannot cast Chain Heal.
Malygos Vortex: Cast one RT before Vortex, one at the beginning of the Vortex, one near the end. When everyone lands, you can cast three empowered Chain Heals.
Lotheb: Cast two Riptides before the healing window. Then during the healing window, you can cast two empowered Chain Heals,
Gluth: Cast three RT before the Decimate, and three empowered Chain Heals after.
Any: cast RT when moving, cast empowered CH after.
The RT glyph makes the HoT last 6 seconds longer. How often will that allow an empowered Chan Heal, when the 15s Hot wouldn't?
I've been using glyph of HW as an experiment recently. This is a recent enough venture that I only have data from one boss kill. It was a 4HM "And they will all go down together" achievement kill, and I was healing in the back of the room where the dps spreads out preventing CH jumps. For that fight, the glyph was 191 hps, at 65% overheal, making it a competitive alternative.
I was a huge fan of GoHW when I was doing heroics and 10's. There were frequent times when HW was the right spell to cast and I also needed healing. In 25's, if HW is the right spell, typically it is even better for another class to heal it. Exceptions like 4HM prove the rule.
Hots don't tick when player is at full healt. Healing stream tick when players are full healt. Healing stream do 75% ovehealing. You could say then that players are at full healt 75% of time approximate. Combat log parsers can't calculate overhealing from hots precisely.
Originally Posted by Daidalos
I think you can infer overhealing done from the duration of the hot and the number of ticks you should see in that given duration on that player (of course if its 100% overheal for the entire duration of the hot you won't know how much it would have ticked for but you could use averages). As far as I know there is no direct way to calculate overhealing with hots. Otherise the only overhealing you see if from a tick of a hit that takes a person to 100% health and over any subsequent ticks will not show up in the log.
OK, I failed to realize that the combat log wasn't recording the ticks of HoTs when a player was at full health.
It's impossible to know how many times Riptide was cast since it lasts 15 seconds and ticks every 3 seconds but can be consumed early by a chain heal. So perhaps Riptide was cast 1,000 times. That would give us about 18,000 chances to proc Earth Living and we would have gotten 3,600 EL procs without the glyph (or 14,400 EL HoT ticks). The log shows earthliving hots hit 4,975 times. Leaving aside the issue of EL procing on a target that already had the EL buff, this would indicate that 65% of the time earth living procced on a target already at full health. In addition, 10% of the time it did proc, it over healed giving us about a total overhealing of 75% (as Pitbuller pointed out).
OK, so let's run the number again.
If HS provides 230 HPS and overheals 64% of the time, it provides effective healing of 83 HPS.
If Earthliving provides 265 HPS and over heals 10% of the time and doesn't proc 65% of the time, it provides effective healing 66 HPS.
If I cast 100 Healing waves in 250 seconds (2.5 second cast), the glyph would cause Earthliving to proc on 5 of those heals.
(5 x 3176 X .25)/250 = 16 HPS
In this case Healing Stream glyph at 83 HPS would be better.
If I cast 125 Healing waves in 250 seconds (2.0 second cast), the glyph would cause Earthliving to proc on 6 of those heals
(6 x 3176 x .25)/250 = 19 HPS
Healing Stream glyph at 83 HPS would still be better
But what if I cast Chain Heal instead of Healing Wave? Earth living has a chance to proc on each jump of Chain Heal. Based on the log parsing, Chain Heal hit an average of 2.9 targets.
If I cast 100 Chain Heals in 250 seconds (a 2.5 second cast) and we assume Chain Heal hits an average of 2.9 targets, the glyph would cause Earth living to proc on 14 targets
(14 x 3176 x .25)/250 = 44 HPS
Healing Stream glyph is still better at 83 HPS
If I cast 125 Chain Heals in 250 seconds (a 2.0 second cast) and we assume Chain Heal hits an average of 2.9 targets, the glyph would cause Earth living to proc on 18 targets.
(18 x 3176 x .25)/250 = 57 HPS
Conclusion: Healing Stream glyph provides more HPS than Earthliving Glyph (I will make some note of this in the original post or simply remove it)
I haven't had a chance to look at the math posted on the Riptide Glyph but I agree with Vuldunobetra's conclusions: You should always strive to cast a Chain Heal just as the RT on a target is about to expire. I also agree that this is not possible most of the time. As a result, RT does far less healing than I would like to see and our CH isn't boosted as much as I would like to see. I would suggest that Blizzard simply change the mechanics so that CH doesn't consume RT but you still get the boost to your Chain Heal (much the way you do with LHW and ES).
In a perfect world the RT glyph would be very powerful. In reality,you're better off using one of the other glyphs.
One guild on the PTR puts up their stats. They do have 3 shaman healers for most of these tests. I looked into a dozen or so logs for HST vs ELW numbers. This was from 3/16 & 3/17. I don't know what was active on the ptr at that time, nor do I know which talents and glyphs the players were using. Raid History
In general, fights where CH was used mostly favored Earthliving. The fight where LHW was used (Patchwerk) favored HST. No surprise there. I'm slightly surprised at just how much healing both were doing overall.
Could you show me how you got the 75% overhealing? I'm still seeing 10%. Stassart's mod appears to track only healing done in combat, much like Recount. Obviously, effective healing will vary from fight to fight and shaman to shaman but I have also found Earth Living and Earth Shield to be our most efficient heals.
The combatlog does not show Earthliving ticks when the person is at full health. So the overhealing number for earthliving is only going to be based on the ones that actually tick. As I can determine when earthliving weapon is applied, when it is refreshed, and when it is renewed I can determine how much it actually ticks and give a real overhealing number and give data per tick. Adding that to my TODO list.
Just to clarify, shaman_hep does not just show combat healing. For a lot of the calculations it only uses in combat heal data. I should split out combat healing in the healing report like I do in the mana usage report. Adding that to my TODO list.
I do currently run with the earthliving glyph, but I am quite interested in coming up with numbers for evaluating the actual benefit of the different glyphs so that an informed decision can be made.
Just to clarify, shaman_hep does not just show combat healing. For a lot of the calculations it only uses in combat heal data. I should split out combat healing in the healing report like I do in the mana usage report. Adding that to my TODO list.
Shaman_hep 1.44 now displays both total healing and in combat healing in the healing report. In version 1.44 all calculations are done using in combat numbers unless the report shows both in combat and total, in which case the calculations are done on both.
I just uploaded Shaman_hep 1.44 to Curse and should be available for download within the next 24 hours.
I only mention this in this thread because people are using the output of shaman_hep for this thread and I want to provide the best data possible.
I considered this an item that should be fixed asap as it should be clear what is in combat and total data.
On live with 2710 spellpower it's ticking for 281 (without the glyph)
On ptr with 2710 spellpower it's ticking for 287 (without the glyph)
So the math should look like this:
250 + 25 (no talents)
250 + 31 (live 5/5 Restorative Totems)
250 + 36 (PTR 3/3 Restorative Totems)
So over a 5 minute fight, 3/3 Restorative totems adds a total of 8,250 healing, if there is no overhealing, and every party member is always within range.
It's impossible to know how many times Riptide was cast since it lasts 15 seconds and ticks every 3 seconds but can be consumed early by a chain heal. So perhaps Riptide was cast 1,000 times.
It's even worse than that, shaman_hep actually has separate figures for both Riptide and Riptide_hot. There were 3116 riptide hits (with crits, so definitely only the initial hit), and 4161 riptide HOT ticks, so there were a couple of thousand more chances for Earthliving to proc and do useful healing.
I'm kind of glad the Glyph of Healing Stream is looking good in comparison. I like the totem and I look forward to dropping it more often.
PTR Build 9722 just went live but 3/3 Restorative Totems is still bugged. Instead of boosting the total healing done by your Healing Stream Totem, the talent is only boosting the base healing (25 healing every 2 seconds). So instead of buffing this totem, it is actually worse on the PTR than the live server.
Live with 5/5 Restorative Totems and 1949 healing: 211-212 per tick
PTR with 0/0 Restorative Totems and 1943 healing: 185-186
PTR with 3/3 Restorative Totems and 1943 healing: 197
Expected value on PTR with 3/3 Restorative Totems and 1843 healing: 268-269
PTR Build 9722 just went live but 3/3 Restorative Totems is still bugged. Instead of boosting the total healing done by your Healing Stream Totem, the talent is only boosting the base healing (25 healing every 2 seconds). So instead of buffing this totem, it is actually worse on the PTR than the live server.
Live with 5/5 Restorative Totems and 1949 healing: 211-212 per tick
PTR with 0/0 Restorative Totems and 1943 healing: 185-186
PTR with 3/3 Restorative Totems and 1943 healing: 197
Expected value on PTR with 3/3 Restorative Totems and 1843 healing: 268-269
I am assuming this is a bug until devs say otherwise. I'll leave my spreadsheet as is for now. If its not a bug it makes me wonder where I will spend those points.
Instead of boosting the total healing done by your Healing Stream Totem, the talent is only boosting the base healing (25 healing every 2 seconds). So instead of buffing this totem, it is actually worse on the PTR than the live server.
I logged on the PTR and submitted an in-game bug report quoting the above text.
Posted on the PTR forums about healing stream not getting the 45% increase to amount healed and that it is being added to the base heal of the spell instead. Dresorull said they are looking into it.
There are a few spots where the RT bonus to CH can be effectively used. First, consider that if CH can be cast effectively (hit multiple targets), then cast it instead of RT. There are a few current encounters where RT can be used to build up CH. In general, RT is useful when you cannot cast Chain Heal.
Malygos Vortex: Cast one RT before Vortex, one at the beginning of the Vortex, one near the end. When everyone lands, you can cast three empowered Chain Heals.
Lotheb: Cast two Riptides before the healing window. Then during the healing window, you can cast two empowered Chain Heals,
Gluth: Cast three RT before the Decimate, and three empowered Chain Heals after.
Any: cast RT when moving, cast empowered CH after.
The RT glyph makes the HoT last 6 seconds longer. How often will that allow an empowered Chan Heal, when the 15s Hot wouldn't?
What do you mean by x empowered Chain Heals? Is that 2 different casts or just the bounces of CH?
If you mean CH bounces then I can confirm it doesn't work that way. I just tested this after reading your post above and RT got consumed only when the CH was cast on the primary target having RT-HoT (exactly as the tool tip says). 2nd target who did have RT-HoT on him did not get the benefit and the RT-HoT was still there.
What do you mean by x empowered Chain Heals? Is that 2 different casts or just the bounces of CH?
If you mean CH bounces then I can confirm it doesn't work that way. I just tested this after reading your post above and RT got consumed only when the CH was cast on the primary target having RT-HoT (exactly as the tool tip says). 2nd target who did have RT-HoT on him did not get the benefit and the RT-HoT was still there.
The bounces of CH do not consume RT but all bounces are increased if the initial hit is on someone with RT.
OK, I failed to realize that the combat log wasn't recording the ticks of HoTs when a player was at full health.
OK, so let's run the number again.
If HS provides 230 HPS and overheals 64% of the time, it provides effective healing of 83 HPS.
If Earthliving provides 265 HPS and over heals 10% of the time and doesn't proc 65% of the time, it provides effective healing 66 HPS.
<snip>
Conclusion: Healing Stream glyph provides more HPS than Earthliving Glyph (I will make some note of this in the original post or simply remove it)
I haven't had a chance to look at the math posted on the Riptide Glyph but I agree with Vuldunobetra's conclusions: You should always strive to cast a Chain Heal just as the RT on a target is about to expire. I also agree that this is not possible most of the time. As a result, RT does far less healing than I would like to see and our CH isn't boosted as much as I would like to see. I would suggest that Blizzard simply change the mechanics so that CH doesn't consume RT but you still get the boost to your Chain Heal (much the way you do with LHW and ES).
In a perfect world the RT glyph would be very powerful. In reality,you're better off using one of the other glyphs.
I was thinking a little more about this data, and, after first admitting that I may be reading the data incorrectly, I think that the test conditions overstate the value of the ELW glyph, and in practice that glyph is actually even worse than the data suggests.
* Healing Stream Totem's overheals are always reported in the combat log, so a combat log parse will always show an accurate amount of overheal for HST. Therefore, it is fair to take parses for HST at face value.
* ELW, like most hots, will not display in combat log parses if it is 100% overheal. Therefore, ELW overheal tends to be grossly underreported.
* Given that nearly all resto shaman will have 2/2 Blessing of the Eternals, heals on targets below 35% health will proc ELW 100% of the time, making the glyph a complete non-factor in these situations (I'm going to refer to this phenomenon as "ELW execute".) ELW executes further reduce the relative power of the ELW glyph. As no combat log parser I have seen is capable of tracking player health percentages, it's impossible for us to know how many ELW procs were due to ELW executes.
In short, I think that the data seems to understate the difference between the ELW and HST glyphs, and that HST is actually far, far superior. However, I can't think of a good way to account for ELW's invisible overheals or how to properly factor out ELW executes. Does anyone have any ideas how we can accurately measure ELW procs?
With regard to the riptide discussion, I think that the empowered CH mechanic as it stands now is interesting and gives us another dynamic/strategic decision to make. However, I don't think that the RT/CH interaction needs the kind of buff that Skyhoof is suggesting. If RT stayed up for the duration regardless, on the majority of fights I would maintain RT on rogue or DPS warrior and use them as my spam CH target. We already know from years of data and calculations that CH is superior HPS and HPM with even one jump; by essentially giving CH a passive 25% boost in any fight where the melee takes damage, we become CH spammers again. This is even more pronounced if we start maintaining two riptides with 100% uptime (presumably one on ranged and one on melee, in which case we have two main targets for the majority of our chain heals.)
As things stand right now, empowered CHs are difficult to achieve in practice, with the exception of fights like Loatheb, in which we have a lot of time to set it up. The mechanic rewards thinking two or three steps ahead; if riptide always lasted its full duration, the mechanic would reward repetition and inflexibility.
Having said that, installing a HoT status mod for Grid has really helped me keep track of my Riptides and make quicker, more accurate decisions regarding empowered CHs. HoT status mods aren't just for resto druids anymore, and if you (a generic you) are looking for ways to better utilize the CH/RT interaction, try downloading some form of HoT timer.
Last edited by Philondra : 03/26/09 at 12:24 AM.
Reason: Fixed a slight logical fallacy
One guild on the PTR puts up their stats. They do have 3 shaman healers for most of these tests. I looked into a dozen or so logs for HST vs ELW numbers. This was from 3/16 & 3/17. I don't know what was active on the ptr at that time, nor do I know which talents and glyphs the players were using. Raid History
The 3 of us use the following variations for our dual-specs:
I parse combat logs for "Gained Earthliving" and "Refreshed Earthliving" and sum the events up. I multiply that by 4 ticks, then by the average tick amount. That's your total potential ELW healing. Divide your recorded Effective Healing by the Potential Healing. 1 - result = actual ELW overhealing. It takes a little bit of digging, and usually takes me 2 or 3 tries until I get the WWS script syntax correct, but it is possible to find the number.
03/20/09, 12:04 PM
Now I can see a lot more scenarios where I can use this combo given my guild's raid compo.
Similar Threads
