Talk:Heal efficiency

I'd indicated some time ago that I desired to do something like this. For now, these are the only skills listed here, however I intend to add to this list if this form of an article is supported by the community. Otherwise, I don't wish to waste too much time if something like this is deemed unneeded. - Greven 17:45, 17 May 2006 (CDT)
 * Just a quick note before sleep. Shouldn't this be in the research category or something. I can't really remember the correct place, but see one of the previous similiar articles. --[[Image:Gem-icon-sm.png|User:Gem]] 17:46, 17 May 2006 (CDT)


 * Divine favor is actually a significant factor if you want to do a comparison like this. For example, with 16 healing and 9 divine, orison and other are much closer to each other than without the divine (20.6 versus 22).  At least, a discussion of a metric should include what it's not useful for, as reference.  --68.142.14.151 17:59, 17 May 2006 (CDT)
 * I'll make a note of this in the article, however I won't go in explicit detail, as Divine Favor also applies to any non-healing Monk skill. - Greven 19:16, 17 May 2006 (CDT)


 * This kind of research is supported, and you'll find some work done by Ollj on the matter somewhere. He had charts and stuff. I agree that ignoring Divine Favor does not make a whole lot of sense most of the time. A Boon Prot monk actually relies on high Div favor to boost his heals. Also, I don't understand any of the tables. If someone can kindly explain the correlation, I'd be very grateful. --Karlos 22:24, 20 May 2006 (CDT)


 * There's no way to chart the efficiency including Divine Favor. Additionally, not everyone is a Monk primary - and Monks aren't the only ones with heals.  I won't just add Monk skills as "healing" as you may have noticed by the inclusion of Vampiric Gaze.  I realize that some Monks rely on Divine Favor as their primary source of healing, however with the variable nature of Divine Favor and the additional compounding factor of variable attributes of the skills themselves makes things complicated.  That being said, I could add a chart detailing the efficiency of spells working with Divine Favor.  On another note, Karlos, it's simple division as to how the tables were made.  For instance, the efficiency of Healing Burst is that it heals 6 points of health for every 1 energy cost at 0 Healing Prayers.  - Greven 00:24, 21 May 2006 (CDT)


 * Alright, the Divine Favor charts are added in. Easy enough to add the correct value to particular skills to see which is more efficient. - Greven 14:38, 21 May 2006 (CDT)

I still don't like this article much. It's not linked from anywhere, either. --Fyren 01:36, 18 September 2006 (CDT)
 * Is it completely unuseful? I'm not sure where to link it from, personally.  I proposed long ago to add the information in a chart form to each healing skill page (like what has been done with Energy management skills), but was told it would be a bad idea, so this is why the article exists as it is. - Greven 01:53, 18 September 2006 (CDT)
 * The anonymous comment above was made by me. For this, I don't think considering only the cost, activation, or recharge is helpful.  Even if you combined all of them together (not that I think there can be a good way to do so), it would still be lacking information beyond just the numbers.  Gift of health sits pretty well in the middle of all the tables, but it's a very good skill for many monk builds.  Healing whisper is pretty high, but you can't really analyze the skill without considering the half-range aspect.  The divine favor information is vital, like I said while anonymous, but people aren't really going to use the table near the bottom to factor it in.
 * There might be some bias on my part about what I think is and isn't useful information. I wrote the original version of energy management guide while anonymous and dislike what it's ended up as.  I tried to write it considering the how, what, and why of the skills people were actually using in PvP for energy management (at the time I wrote it) and leave out theory.
 * Back to the point, I think putting a three line table in each skill (for heal versus cost/activation/recharge) isn't helpful at all. I'm not saying delete this, but the information needs to be findable.  Maybe link it from monk or the monk guide or any similar articles, if there are any.  Or maybe in the energy management article, too.  --Fyren 02:09, 18 September 2006 (CDT)
 * I've linked it from both Health and Heal. I'm not sure I want to link it to a Monk/Monk guide as it isn't strictly for that class. - Greven 13:31, 19 September 2006 (CDT)

Organization
Is the order of the skills (by highest at 0) useful? I don't particularly like this, but the question is... where do you place the starting point? At 16? Or perhaps 12? Any ideas? - Greven 21:43, 20 May 2006 (CDT)
 * Sorting by 12 and 16 would both produce similar tables. -- Gordon Ecker 02:27, 8 January 2007 (CST)
 * Yeah, although they'd still be somewhat of a mess, due to the interesting progressions and start points of some things. - Greven 03:04, 8 January 2007 (CST)

Recycle
Efficieny vs. recycle time (cast time+recharge time) is arguably more valuable than recharge time alone. &mdash; Shining 17:47, 8 October 2006 (CDT)
 * Good point. - Greven 18:13, 8 October 2006 (CDT)

Analysis
For the less observant wikiers, couldn't an analysis on certain healing skills (the strongest for cost, recharge, etc.) be written? Just a thought, thanks.
 * If I have some time to do this tonight, I'll probably restructure this and write up some notes as well. Likely, I'll create a couple templates so this can be more easily added. - Greven 17:49, 1 February 2007 (CST)

Progression template
You could make your own version of template:progression row. The real one checks the attribute to see how high it can go and not generate cells, you could stick in N/As instead. --Fyren 01:22, 9 February 2007 (CST)
 * Yeah, I'll probably end up doing that, I just didn't have time to fully complete that this afternoon. - Greven 02:10, 9 February 2007 (CST)
 * And the table cells aren't center aligned because the table doesn't have class progression, so the styles aren't getting applied. --Fyren 02:43, 9 February 2007 (CST)
 * Using the progression row was a quick and dirty way of getting this setup for change later ;P - Greven 02:48, 9 February 2007 (CST)

Informantion
I would welcome separation of direct targeted heals, aoe heals and self heals in spell eficiency, also i would remove highly condtional/unwiedly heals or give them their own section (i am talking about Intervention series, RoF which imho should be here.) Table is currently too bloated to be of any use when deciding for right skill (when working on healer build you are definitelly not interested in all self heals of various classes...). (divine favor eficiency table are nice btw.) Grima.worm@seznam.cz 13:56, 6 March 2007 (CST)

Practical concern
I think that perhaps spells with special effects like LoD, Ravenous gaze, and words of comfort should be separated from those with with straight effects like orison and Heal other. Its silly that loD is considered less efficient than Heal Other.

I.e., have a single heal versus party heal, or have two seperate "skills" e.g. LoD ideal situation, LoD unideal situation.(Not a fifty five 19:19, 12 March 2007 (CDT))

RoF
RoF doesn't heal directly, so it shouldn't be in here. 19:34, 12 March 2007 (CDT)

Ravenous Gaze
Ravenous Gaze should be given 2 values, 1 for the base steal, 1 for the base+additional if the condition is met, as this could be misleading to people to think it is a terrible heal --Gimmethegepgun 18:32, 4 April 2007 (CDT)

Conditional Heals
I believe that skills like ZB, WoH, etc should be included in the table twice; once without the conditional healing, and once with it. This drastically changes the values, i.e. WoH goes from 21 to 38 at 12 healing, ZB goes from 15 to 50 at 12 prot, etc. --131.216.134.155 17:57, 4 April 2008 (UTC)

Bugfix needed
breaks because of the 3/4 activation; the way this is done is that there's an explicit progression in the table, and that of course is an Ugly Hack, so for this one. --◄mendel► 00:04, 31 October 2008 (UTC)

Signets should be left out of heal to cost. activation time making 1 energy? thats hva, just leave sigs out of first table. Whats more the whole page is messed up looking and its a very useful feature, i would fix it if i was good at wikicode but im not so some1 plz fix-Rabus 00:24, 31 October 2008 (UTC)

Regeneration-based heals
I see they are not in the article yet. I was hoping to see if Troll Unguent was still the most/almost the most energy-efficient heal in the game. (T/C) 01:27, 31 October 2008 (UTC)
 * With Expertise it definitely is, but when you consider the fact it takes 13 seconds to give it all to you (and 3 seconds cast) it obviously isn't a combat heal. Now, if you want an extremely powerful, instant, fast cast, short recharge, AND low-cost heal, bring Spirit Light along with Attuned Was Songkai at 15 Spawning --Gimmethegepgun 02:06, 31 October 2008 (UTC)
 * Well adding AWS boosts any Ritualist heal efficiency by huge amounts >.> [[Image:Entropy Sig.jpg]] (T/C) 02:08, 31 October 2008 (UTC)
 * And adding Expertise improves any Ranger skill's efficiency by huge amounts (virtually identical amounts too :P ) --Gimmethegepgun 02:11, 31 October 2008 (UTC)
 * It is too bad Expertise and Spawning Power are mutually exclusive, though I bet Nerashi could do it. Cost reduction to the necessary binding rituals, cost reduction to all spells, take Whirling Defense for superlong block when recasting Life... [[Image:Entropy Sig.jpg]] (T/C) 02:13, 31 October 2008 (UTC)

Rework, starting thoughts...
I'm looking at possibly reworking this page, however before I do anything extensive (like axe the tables and start from scratch), I'm going to ask myself first: What information should the page present? And bugger all, but in the current form it only covers one of the five things I'd want most in a 'healing efficiency' page.

Of these, the page does 1. It comes close to 2, but omits after-cast delays, which are different for some skills, and cannot be neglected when calculating 3 & 4, which are the most useful of the 4 when aiming to initially tune a build, or compare different skills, for which 5 is almost a must.
 * 1) I'd want to know the points of healing per energy used (which for signets is infinity, not based off of 1 energy like the tables now handle it)
 * 2) I'd want to know the points of healing per cast -- which I can infer, but do not know based on the article in its current form.
 * 3) I'd want to know the maximum energy use over time, from casting the spell on recharge -- eg: does it equate to using 0.25 pips regen or 6 pips regen?
 * 4) I'd want to know the maximum healing use over time, from casting the spell on recharge -- eg: does it heal for 100 points per minute or 600 points per minute.
 * 5) I'd want a note about special characteristics of the skills (AoE heal, condition removal, requires enchantment, etc...)

All that being said, that's what I would want to see, but may not be what this page should offer (And I only listed the 5 things I'd most want to see.  I'd also like to throw in 20/20, 40/40, enchant mods, etc, but that gets...a bit towards information overload for anyone using the page). So, any thoughts before do anything rash? Yamagawa 07:55, 13 February 2011 (UTC)


 * Short story: please! I can't make heads/tails over what this page is trying to tell us.


 * Longer story: here's what I'd like to see:
 * What are all the resource costs to give back 100 health to the target? On the whole, the costs are energy, time (activation, recharge, After-cast).
 * For those spells that have the least costs, what is the impact of altering attributes (e.g. Divine vs Healing).
 * Are there mitigating factors (what you call, special characteristics above).


 * BTW: as far as I know, ACD is the same for all spells. It's 0.75s or 0s; I don't think there are any other possibilities. — Tennessee Ernie Ford ( TEF ) 09:18, 13 February 2011 (UTC)
 * I had been monking for like, two years at a high-end (top 500 GvG, HA, DoA HM, VQ as only heal, etc) level, if you need any help with this feel free to give me a call. &mdash; Scythe   16:31, 13 Feb 2011 (UTC)


 * Not all skills that heal are spells. Regardless, a 1/4sec cast is effectively a 1-second cast, where the heal occurs at 1/4 second.  A 2 second cast is really a 2.75 second cast, where the heal occurs at 2 seconds.  Come to think, the after-cast does come out of the recharge time... so for continuous use after-cast is actually moot.  It's only important when spamming several fast-recharge skills, or trying to chain skills -- "can I get that extra heal in before patient spirit ends?".  For the moment I'll leave after-cast out of it.
 * As to the cost to heal 100 hp or total 'cost/benefit' over 60 seconds... They are the same thing, just different formats  Let me sketch out a couple tables with a few rows, and we can compare.  I really like the 'costs x-pips energy' because that gives a clear idea of how much energy a build will use, if any, energy management is needed.
 * I'd rather do this right the second time, and not have a 3rd go-round at it. Yamagawa 17:04, 13 February 2011 (UTC)

. My point is: ACDs are not different; the skill has them (or it doesn't).
 * I completely agree with you that ACD is critical to calculating efficiency. I was responding to your comment that "after-cast delays, which are different for some skills"


 * Cost to give back 100 hp and cost/benefit over 60s are not precisely the same. The first compares costs directly and makes no claims about efficiency. The second calculates a ratio, which depends on the definition of cost and benefit. Tables of both will provide similar information, but presentation of data affects our understanding. Both methods are useful and valid; I have a preference for one over the other in this particular case: I want to know what I need to do to restore 100 hp because that's usually something that turns the tide of battle for me in PvE. Healing per second doesn't seem to matter as much. — Tennessee Ernie Ford ( TEF ) 17:34, 13 February 2011 (UTC)
 * Do you want to have that amortized or not? I.e. if a skill potentially heals 150, do you want its resources reduced to 2/3 for this comparison or not? And do you want to consider "continuous use" or not? If you want something that gives you a heal at a critical time, you might not care about casting the skill again, so any recharge might not be important. --◄mendel► 19:05, 13 February 2011 (UTC)


 * Right, having to invoke a skill 5x is a potential cost, so a quick-recharge-low-heal might be more efficient than a slower-high-heal, but that doesn't mean it's more useful. I'm not sure about amortization: skills that heal for more than the benchmark still have a higher cost.


 * Maybe we should rephrase from asking questions about the presentation of data to what in-game questions we want answered, e.g.
 * What skills prevent a near-death experience from becoming an actual death?
 * What skills allow the healer to keep party members (allies?) in generally good health, while preserving enough energy to handle emergencies and hex/condition removal?
 * What skills help the party retreat gracefully without leaving the healer behind?
 * Those might not be exactly the right questions, but they point at what I really want to know. In other words, efficiency isn't really the issue; winning the battle is. In other games, DPS and HPS trumps everything else. In GW, unless DPS is sufficient to kill right away, it's more complicated. — Tennessee Ernie Ford ( TEF ) 19:16, 13 February 2011 (UTC)


 * I've done up tables mostly how I'd first pictured them. I'll need a bit of a break before tackling more tables, but I do plan to do sample tables that chart costs to 'give back 500 hp'  (when some skills heal for over 100hp, the energy cost doesn't compare as well across skills...).  They might allow me to better condense information.
 * As a monk, and ESPECIALLY while skilling monk heros, I must consider not just the maximum healing ability, I need to consider how to maintain maximum throughput. This equates directly to energy management.  Looking at the 'maximum throughput' of Heal Party, it's a great skill (in a full party at 16 Heal, it gives 9360 HP/Minute), BUT it costs an ugly 11.25 pips of energy to use at that speed.  A hero monk will run out of energy using it, unless it has some form of energy management.  If it doesn't, you're going to find it capped at close to 1/3rd that number (3328 HP/Minute) as the hero runs up against his 4-pip regen -- assuming it uses no other skills.  It's for energy management tuning that I look at the cost as pips of regen when continually using the skill on recharge.  Yamagawa 19:23, 13 February 2011 (UTC)


 * Mendel, it makes sense to Amortize the costs. This page is about healing efficiency, not heal spiking, and that can be viewed as either 'doing so at a high rate' (time efficiency, or throughput) or 'doing so for the long haul' (energy-efficiency).  Were I looking for heal spiking, I'd be designing a rather different table, but to be fair, the 'Heal Skills' quick reference page already has most (all?) of what I'd want in such a table.  I don't see this being as the 'best single source' for deciding what heal skills to use.  I can see it being a quick reference for swapping out skills based on specific needs.  'I need a bit more energy efficiency, but can let the recharge slide a little.' 'More hex pressure here, lets drop a condition removal for hex removal.'  I expect players will start by looking up their favorite skills in the table, then use the table to locate comparable or improved skills.
 * As to the cost-to-heal 500 hp table, I don't have it finished, but at a glance, how I've got it:
 * I combine energy and time costs into a single table, losing the ability to sort on them separately. With the 'cost to heal' it's natural to want to combine the two.
 * Granularity becomes 'higher' the more effective the skill is, making skills with similar outputs potentially appear identical. This may or may not be a problem for comparing skills, but it also forces a variable scale on the results.  A 3 point difference on the table has a different significance when comparing 15 to 18, than when comparing 65 to 68.
 * Breaking out all the attribute levels will probably push the table off the right edge of even wide-screen monitors. While I'm already considering ommitting some of the columns, that will still leave this as a very wide table.
 * All that said, it's still a more condensed form of information that is fairly easy to understand, not just 'seemingly arbitrary' numbers.
 * Thoughts on either form of table? Yamagawa 06:20, 14 February 2011 (UTC)

Continuous Use Cost and Benefit, Benefit as HP/Minute
This can be read as a 'Maximum Throughput' table.

Continuous Use Cost vs Benefit, Benefit as HP/Minute/Energy Pip
This can be read as a 'Efficiency' table.

Cost to recover 500HP - B (work in progress)

 * -- Todo: Load real data. Decide if this is really clever or really silly Yamagawa 05:21, 15 February 2011 (UTC)

-- More tables may follow... Yamagawa 05:21, 15 February 2011 (UTC)

Post-tables notes

 * Why is Glimmer of Light yellow in every table? [[Image:Felix_Omni_Signature.png|link=User:Felix Omni]] 06:04, 14 February 2011 (UTC)
 * Elite Skill. I'd probably want the entire row higlighted for elites, but that gets a bit interesting to edit. Yamagawa 06:22, 14 February 2011 (UTC)
 * I don't see the problem. ;-P --◄mendel► 11:46, 14 February 2011 (UTC)


 * Our standard color for elites is #FFD700 (changed a row to show this), and it's in our CSS as class="elite". Unfortunately, it only works when applied to individual td's, not to the entire tr, but that could be changed easily.  &mdash;Dr Ishmael Diablo_the_chicken.gif 14:06, 14 February 2011 (UTC)
 * Also, these could be coded to use a variation of (without the explicit row markup, and with an exception for signets), since all of those numbers scale linearly with the attribute rank.  &mdash;Dr Ishmael Diablo_the_chicken.gif 16:47, 14 February 2011 (UTC)


 * I fully plan to use templates to implement the tables. Preferably so that the table won't need editing each time a skill is updated.  Before I tackle the how, I'm interested in tackling the what.  Thus the above examples.  I think the tables we want (or something like them) are up there somewheres.  I do like the look of the more condensed final table, but I rather dislike losing the sort capabilities in the process.  It remains a question of which figures to present.  I'll probably condense the columns a bit, listing something like attributes 0,3,7,10,11,12,13,14,15,16, and 18 (juggling around 9 to pick up 10 for the PvE skills)... but lets break this down into points we can agree on, then hash out any disagreements:
 * List Linked Attribute? Link-Y/N
 * List Energy Efficency and Time Efficency in the same table, or split them?  Join-Y/N
 * Rate efficency in healing/cost and healing/time, or in time/healing and cost/healing? Healing-over/under
 * Include a basic 'benefit from one heal' chart? Basic-Y/N
 * Break out by self-heal, Multi-Heal, and Other-Heal?  Typed-Heal-Y/N
 * A character looking for a simple self heal may want just the summary chart for those skills.
 * Multi-heal skills don't compare well with single target heals, so it may make sense to split them out to a different table.
 * List every attribute level (0-20) or, skip some/many intervals? skip-levels-no/some/many
 * There are more fine print items I could bring into this, but I think the above will decide the general composition of the table, and thus the templates needed to build them. I'd rather hash out these items first before we bog down on any other decisions.  I'll weigh in after I hear some input from others.  On the assumption that I'm doing the work, what I chose will be what I implement, but I want this to be useful to others so please, all you 'others', please let me know what would make this useful, and more importantly, why that makes it useful compared to the alternative.  Yamagawa 02:39, 15 February 2011 (UTC)