HPS Alternative - "Clutch Healing"!

90 Pandaren Monk
6735
Background

People on this forum often complain about how HPS doesn't really matter, and that it doesn't really measure the skill of a healer. The argument against it (a good one) is that not all healing is of equal value. For example, a 30k heal that saves the tank 1 second before he would have died is much more valuable than a 30k heal that snipes a HOT just to top someone off.

It's important to remind everyone that the vast majority of the time, healers only have one job: Keeping people alive!

Healing itself is not their job, it is only the means they use to do their job. As long as there are no deaths in a fight, the healers could not have done any better, regardless of how much healing they did.

There are plenty of instances when healing is, in fact, useless, such as:
• Overhealing (Obvious one)
• If a tank ends a fight with 350k health, then the last 349,999 healing they received was useless, since they would have survived without it.
• Typical HOT sniping situations like the one mentioned above
• Etc…

It is clear what we need. We need a healing metric that measures our performance at keeping people alive, not just our performance at healing!!

Clutch Healing

My idea for such a metric is called Clutch Healing. Measuring Clutch Healing differs from normal healing in two major ways:

1) Only useful healing has any value (see above)
2) The value of a heal increases as the time available to deliver it decreases


Let's explain number 2 with an example! Suppose someone with 100 max HP (lol) goes through these 2 scenarios

A)
Takes 55 dmg (45 HP)
1 sec passes
receives a heal for 15 (60 HP)
1 sec passes
takes 55 dmg (5 HP)

B)
Takes 20 dmg (80 HP)
1 sec passes
Takes 20 dmg (60 HP)
1 sec passes
Takes 20 dmg (40 HP)
1 sec passes
Takes 20 dmg (20 HP)
1 sec passes
receives a heal for 15 (35 HP)
1 sec passes
Takes 30 dmg (5 HP)

In both these scenarios, 15 healing was done 1 second before they would have died, leaving them with 5 hp. A normal healing meter wouldn't differentiate between these scenarios, but wouldn't you agree that it requires more skill to heal through scenario A than scenario B? Why is that?

It's because there was less time to react in scenario A. The guy went from 100 to 0 in 2 seconds, as opposed to taking 5 secs in scenario B. Keeping the guy alive in scenario A requires a faster reaction time, maybe a faster, more inefficient heal, maybe using a cool down. A raid fight full of scenario As would require more skilled healing than a fight full of scenario Bs. It's burst damage that kills in both PVP and PVE, and burst damage requires … clutch healing! That's why the value of a heal should scale with the reaction time needed.

I have a basic algorithm that can calculate the reaction time required for each heal from a basic combat log, but I haven't thought much about stuff like the magnitude of the reaction time scaling factor, min and/or max values etc. I also have no experience creating add-ons or anything like that. I just wanted some opinions on the basic concept before elaborating on this any more.

So what do you think?
Reply Quote
100 Undead Priest
10715
There's no way to calculate what you are trying to calculate. If I play smarter and stand in the right spot, I *create* more time for myself to react and thus would appear to be a worse healer than a purely reactive healer.

There is, similarly, no way to measure the value of a fast Dispel.

When I heal someone and keep them at high health, they are much less likely to pop their Healthstone and Personals, but it cannot be measured if they would have otherwise wasted those abilities.

There is unfortunately no measure for a good healer aside from "did the boss die?" which is a metric for all players involved.
Reply Quote
MVP
90 Blood Elf Paladin
HC
12145
The argument against it (a good one) is that not all healing is of equal value. For example, a 30k heal that saves the tank 1 second before he would have died is much more valuable than a 30k heal that snipes a HOT just to top someone off.


But you're comparing two different situations and trying to find a parallel that explains why one is more important than the other.

That's not really an argument though, that's trying to make something otherwise glaringly obvious (saving a tank) stand next to something that has nothing to deal with it (topping someone off at 80%).

When we speak about HPS, we speak about the relatively value of healing in a specific situation, though not always/necessarily comparing it to something else unless needed.

It's because there was less time to react in scenario A.


Sure, but as healers, when we realize that there's less time to react, we're going to find a way to abuse the most efficient heal and use that over and over again (WoTLK.)

03/17/2013 08:51 AMPosted by Geropy
The guy went from 100 to 0 in 2 seconds, as opposed to taking 5 secs in scenario B.


Yeah, but there's less thought in that fast of healing, which in turn makes it very boring to heal.

So what do you think?


Your argument is that a healer that can react faster is in some form a better healer. While to some degree you may be correct, we've already tried that formula and this is what we've found:

Players will always find a way to make the most efficient heal the only heal to use, even if it's an inefficient heal. Why? Because when you are not given enough time to think about the heal you're using, you're going to go to the spell that covers the highest percentage of problems that you know. In this case, our Flash/Surging/etc heal.

You've actively described WoTLK healing, and while it was cool for those very 'clutch' moments as you've described, you can still have those moments (and still probably do frequently) through saving someone's !@# with a nice Life Cocoon. Or Leap Of Faith. Or Natures Swiftness. Or even a Flash heal, which is (Spoilers!) what they're used for!
Edited by Practical on 3/17/2013 9:14 AM PDT
Reply Quote
90 Pandaren Monk
6735
@Multicidez

I agree that this is definitely not a perfect measure, as it doesn’t measure things like the value of dispels, CDs like Leap of Faith, or if healthstones/CDs of others would have been wasted, but I believe it does a much better job than pure HPS without being overly complex.

@Practical

I have raid-healing experience from WOTLK, so I know what you’re talking about. But even in a slower-paced healing setup, there will always be the need for triage, and so there will always be the need for prioritizing those who are at risk of dying sooner.

I’m not saying that I want to return to a model where health bars are bouncing up and down like crazy. Like you said, these clutch moments probably mean more when they are rarer. However, I would argue that the response to these types of moments, regardless of how rare and drastic they are, is very important to a healer.

If two healers have identical healing output for a fight, but one healer focused more on people at medium-high risk of dying, I would want that healer to have a higher “healing score” for lack of a better term. That’s what I’m trying to do with this.
Reply Quote
100 Blood Elf Mage
18255
03/17/2013 09:35 AMPosted by Geropy
I’m not saying that I want to return to a model where health bars are bouncing up and down like crazy.


I don't really understand when people say this. The idea that health bars are relatively stable or something isn't accurate. I know the developers alleged back when cataclysm came out that they had instituted a healing model of "oh don't worry, you can leave that person at half health for a long time and slowly heal them up with your efficient heals", but that never came to pass. It's just that now, unlike in Wrath, many of our heals are slow. It's why we have a major proliferation of zomg cooldowns.
Reply Quote
MVP
90 Blood Elf Paladin
HC
12145
However, I would argue that the response to these types of moments, regardless of how rare and drastic they are, is very important to a healer.


Why?

If two healers have identical healing output for a fight, but one healer focused more on people at medium-high risk of dying, I would want that healer to have a higher “healing score” for lack of a better term.


You mean HPS.

That’s what I’m trying to do with this.


You can't do that though without creating an environment that's ridiculously similar to WoTLK.

The idea that health bars are relatively stable or something isn't accurate. I know the developers alleged back when cataclysm came out that they had instituted a healing model of "oh don't worry, you can leave that person at half health for a long time and slowly heal them up with your efficient heals", but that never came to pass


Sure it did, but the majority of the time it was in normals, and in some Heroic Modes. Outside of that It's really hard to balance that sort of "let them sit there with that HP" in heroic modes because of the nature of HM's.
Reply Quote
90 Pandaren Monk
6735
Why?


Because these are the types of situations where deaths most often occur.

You mean HPS.


I don't, because in the situation I described, the two healers had identical HPS, but one of them did a better job at prioritzing healing on those who were at greater risk of dying soon. Really all my system is doing is adding some emphasis on healing priority instead of purely on healing output.

You can't do that though without creating an environment that's ridiculously similar to WoTLK.


I'm not advocating any change to the healing environment. This has no impact on the fights themselves, just an extra number to put in recount or something similar.

I'm also not saying that only fast reaction time healing should matter. If a healer does nothing the whole fight but save the tank from burst damage a couple times, their "clutch healing" would be lower than someone who did 10 times their healing by topping people off. There's a balance in the middle somewhere.
Reply Quote
100 Undead Priest
10715
@Multicidez

I agree that this is definitely not a perfect measure, as it doesn’t measure things like the value of dispels, CDs like Leap of Faith, or if healthstones/CDs of others would have been wasted, but I believe it does a much better job than pure HPS without being overly complex.


You will *never* get a perfect measure of a healer.

HPS does not measure quality. It can sometimes point to a weakness in healing, but that is the extent of its purpose.

The problem is too many people try to equate HPS with quality, as if it were a way to measure one's !@#$% size. Then we have people like you who try to come up with a revolutionary new way to try and measure one's ^-*!@ size.

Just stop.
Reply Quote
90 Pandaren Monk
6735
@Multicidez

I agree that this is definitely not a perfect measure, as it doesn’t measure things like the value of dispels, CDs like Leap of Faith, or if healthstones/CDs of others would have been wasted, but I believe it does a much better job than pure HPS without being overly complex.


You will *never* get a perfect measure of a healer.

HPS does not measure quality. It can sometimes point to a weakness in healing, but that is the extent of its purpose.

The problem is too many people try to equate HPS with quality, as if it were a way to measure one's !@#$% size. Then we have people like you who try to come up with a revolutionary new way to try and measure one's ^-*!@ size.

Just stop.


I'm not trying to get a perfect measure of healing. I realize that it's impossible. I'm just trying to get a better one. I'm also a believer that there is no point to using any type of healing measure to compare people from different groups, it's really only useful for analyzing what an individual group needs to do to improve.

The fact of the matter is that, whether it's HPS or something else, people are going to try use it for an unintended purpose, like measuring e-peen. That doesn't mean we shouldn't try to find something better for the people that will use it properly.
Reply Quote
100 Night Elf Druid
12555
I think you rightly understand the problem with weighing healing purely based on HPS, but beyond that, you're kind of missing the point. Maybe you're proposing this because you do a lot of runs with people you don't know (LFR, pugging, whatever), but in a group that you run with regularly, you shouldn't feel the need to lord your healing numbers over your fellow healers. It's the teamwork of the healers that keeps everyone up. Each healer has a niche.

I run my resto druid with a holy pally and a disc priest. The disc priest is great for the first line of defense, especially when major damage is about to hit. The holy pally is great for big heals on one to two targets taking lots of consistent damage. I have some tank healing, some raid healing, but mainly I'm sort of the mortar of the team. I can't keep up with the holy pally's big heals, but I can fill in the small heals better than either of my co-healers and I can use hots preemptively to smooth out the healing. I also have a lot of fast heals, so I can help save someone when the holy pally is busy focusing another target or when his best instant heals aren't ready. We all have different strengths and our relative amount healed varies from fight to fight, dependent on what kind of damage goes out. Each of us is important, and while our team can get by just fine when one of us switches specs for a particular fight, having, say, two holy pallies wouldn't work as smoothly, even though they would probably have great HPS and great "clutch healing".

The point is, it's better to measure healing in terms of "how well do I work with my co-healers" and "Does the boss fall, preferably with few casualties".

The only time I discard this mentality is in LFR, where there's so little damage going out that I could simply not heal and the other healers would be able to pick it up without a problem. But I choose to heal, so I'm going to heal, even if it means all of the healers are sitting there, bored, with only about 30k HPS. That's why I'm wondering where you're coming from with this suggestion. The only time I can see healer competition being at all worth it is when utterly bored in LFR.
Reply Quote
100 Undead Priest
10715
I'm also a believer that there is no point to using any type of healing measure to compare people from different groups, it's really only useful for analyzing what an individual group needs to do to improve.


The "death" meter is more accurate for this than any healing metric you might try to create.
Reply Quote
90 Tauren Paladin
0
<thinks>

HPS still gives several indicators of a healer's degree of skill, so long as all the supporting factors are taken into context, such as utility spells used, raid role, and remaining mana at the end of a fight. It is difficult to quantify on a per boss basis, but when you run several weeks worth of raids, you can start to see strings of evidence.

The concept of measuring a healer based upon clutch healing, however, says that any class with instant spells is rewarded over any class that uses HoTs: suddenly heal sniping is rewarded (in a real raid environment, heal sniping usually means that YOU spend mana on the heal while your teammate doesn't cast one, which means they'll have more mana later to heal snipe you in return, though this isn't always the case).

When a raider takes 50% damage, clutch healing sounds like it would reward the player that throws out faster, mana inefficient heals than one that uses a slower, more efficient manner, such as allowing splash healing or ground AoE's like healing rain to top them off when there is no pressing need to heal. Your system would automatically up the value of tank healers as opposed to raiders (when fighting Elegon, for example, and the add explodes, raid healers have time to get them back up, whereas the tank healer has to hit the tank with heavy heals to prevent death from Breath or claw).

Honestly, while the idea of measuring heals with clutch sounds better than raw HPS, doing so would reward healing in a particular playstyle, which is not available to every class, or those that can game the system. HoT "sniping" may be bad, but rolling HoTs on raiders means they take longer to die which means more mana efficient heals can be used, or allows healers to divert focus to raiders that need it more for a precious few seconds.

HPS (and the breakdown of spells used or not used) may be a poor-mans ep33n measurer, but skilled healers are easily able to tell which ones are pulling their weight. And if everyone is skilled, then there isn't really a problem, whatever the HPS meter says.
Reply Quote
I'm still fairly new to healing, so I apologize if some of this is just plain wrong.

But in my experience healing so far, it seems like on top of different healing styles for different classes, there are actually different viable healing styles within each class. Unlike DPS where there is typically one efficient rotation or priority system to produce maximum DPS, healers seem to have a varied toolbox for different situations and not one specific style is necessarily the "best."

For example, I really like keeping health above 90% as much as possible because anticipatory healing and having a safety net works better for me. Some healers are comfortable letting health dip lower, maybe because they are better at reacting quickly in low health/rapid healing scenarios. Again, sorry if I am mistaken here due to lack of experience, but it seems like as long as the boss gets downed (ideally with all or most players, and tank surviving) either option is perfectly viable.

Trying to create any system that measures healing done/HPS/etc. seems like it would only reward one style of healing at the exclusion of another.

I think addons measuring healing output in different ways are very useful at helping healers optimize their healing, and in the case of new healers, help us understand what is going on with our healing in comparison to others, but I just don't think it's so cut-and-dry that one addon measuring one style of healing output can really judge whether or not you're performing to the best of your ability.
Reply Quote
100 Blood Elf Paladin
9980
Try to find something that will show you heal by heal how spikey your heals were, like the healing portion of the world of logs. Evidently if your heals are spikey, then the tank probably was taking a lot of damage, and you were healing through it as well as low period heals to be stable. That would be the best way to measure your own healing in my opinion. The spikier the better probably.
Reply Quote
100 Undead Priest
10715
Try to find something that will show you heal by heal how spikey your heals were, like the healing portion of the world of logs. Evidently if your heals are spikey, then the tank probably was taking a lot of damage, and you were healing through it as well as low period heals to be stable. That would be the best way to measure your own healing in my opinion. The spikier the better probably.


The game of "how low can you go" is fun on trash, but let's not encourage it on boss kills too.
Reply Quote
90 Draenei Paladin
10900
03/17/2013 08:51 AMPosted by Geropy
It is clear what we need. We need a healing metric that measures our performance at keeping people alive, not just our performance at healing!!


We already have this. Its called World of Logs, and it logs every little piece of possible information that fits within a combat log. To include, talents used, buffs cast, when said buffs were cast, dispels cast, absorbs done, healing done, effective healing done, what healing spells were used and when they were used. It even tells you what that healer was doing when so and so died, who took damage and from what.

HPS is more of a guideline, and it's pretty much agreed on that the only thing HPS really tracks is who is a better sniper. Anyone who is really interested in the skill of a healer, will look at their logs.
Reply Quote
63 Dwarf Priest
0
It's an interesting idea. For instance, taking the heals a player received immediately (seconds?) prior to damage that would have killed them, and then applying "points" based on the percentage of healing that was done by each healer could give you a metric on which healer is "saving" people the most, but...

What's happening in the encounter, who is capable of healing while moving, range, cast times, direct heals vs. HoTs vs Absorbs vs damage reduction, letting someone drop low enough that they are in danger of dying in the first place when it was otherwise preventable, people's healing assignments, etc are all things that skew this system and make its value questionable.

That's not to say that HPS or HPS(e) are perfect measures of healer worth, quite the contrary, but people flock to them because they are readily available and easy to compare, which makes them great ammunition to reprimand a player or condemn a class.

Honestly the best measure of both player, and particularly class healing abilities, would be if things like HPS weren't widely known, and the focus was instead on how far raids with these healers were getting relative to others. Of course this could be skewed by things like a classes representation, but a low representation would also be representative of whether a healer's style or kit is fun/easy to play... which would probably come in handy for balancing, too.
Reply Quote
90 Goblin Priest
14695
Are you failing to kill a boss? Look at the deaths and find the root of the problem. If it's a healing issue then analyze your logs for healer cd/spell usage both at the point of death and over the course of the fight. Know your healers' roles and take them into account. Outside of this situation I fail to see the need for your "HPS Alternative."

Stop trying to assign some half-baked numerical method to ranking healers based on a bunch arbitrary parameters you have decided on. Healing is like a vector, magnitude and direction are both important. Use logs to find which of those is the problem and address it.
Reply Quote
90 Pandaren Monk
6735
It seems like over-estimated the usefulness of something like this. People seem happy to just check WoL or something similar for specifics if there are healing issues within the group, and don't really feel the need for a single number that summarizes. Fair enough.

I haven't really raided at a high enough level before that detailed analysis of WoL was a common practice.

To be honest, I'm just a nerd who enjoyed coming up with the idea and basic algorithm behind a better healing metric. I figured it was worth sharing to see if the community thought it was useful :P.

@Urteil

Although it doesn't account for things like healing assignments, it could account for things like absorbs, dmg reduction CDs, cast times etc.

It also does NOT reward a healer for letting someone get low on purpose before healing them. The reaction time is calculated with respect to when you first COULD have healed them (i.e. when they first had that much empty space in their HP bar), not when you CHOSE to heal them.
Reply Quote
90 Troll Shaman
17270
3 issues:
1) Spirit Shell/Absorbs, if the damage burst is big enough to bring people to critical, does it count as bringing them from critical to what they are in 0.0 seconds.
2) How do you measure how much time you have to bring them up. You'd have to measure exactly when the next damage event COULD hit them, and whether it would be lethal. (not just damage that does hit them)
3) You say overheal shouldn't count, but it already doesn't on meters.

Absorbs are the hardest part. The other half is that taking the raid from 80% to 100%, means that when someone takes 60% of their health in damage they are at 40% health instead of 20%. So in a lot of sitautions, raw oomph is just as good as "clutch" healing.
Reply Quote

Please report any Code of Conduct violations, including:

Threats of violence. We take these seriously and will alert the proper authorities.

Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.

Harassing or discriminatory language. This will not be tolerated.

Forums Code of Conduct

Report Post # written by

Reason
Explain (256 characters max)
Submit Cancel

Reported!

[Close]