Life in Group 5 – A Resto Shaman Blog
A resto shaman perspective on raiding


Healing

March 10, 2010

The Next Step with WoL: Healing Analysis

More articles by »
Written by: Vixsin
Tags: , , , , ,
WoL Healing Analysis

So by now, you’ve become acquainted with World of Logs, played around with its tools, and maybe uploaded a log or two to your guild’s site. You have a basic faculty with the interface and feel comfortable navigating through the various layers of information. But, if tasked to provide feedback on a new healer in the guild, or critique your own performance, do you know where to look? Given that I’ve been spending the whole of my raiding time in hardmodes lately, and accordingly more time picking through WoL to identify anything that could make the difference between a kill and a wipe, I thought I’d pass along some of my tips for Healing Analysis.

In comparison to the Introduction to WoL, this healing analysis is going to be relatively brief, and only cover some of the more basic tools for evaluating performance. You can go absolutely nuts reconstructing encounters through Expression Editor, so in consideration of your sanity, I’ll leave that alone. The point here is to get you familiar with what an outside evaluator (maybe an officer in the guild you’re applying to, perhaps) might look at when combing through your parses. In the examples below, I’ll be looking at one of my own parses, of a recent normal-mode Putricide kill, from as close to an objective perspective as possible.

* When describing how you can access the various tools (1) will represent the first drop-down (or “data category”) menu, while (3) will represent the third drop-down navigation (or “player detail”) menu.

Overall Encounter Tools

When evaluating a healer’s performance, it’s oftentimes our natural inclination to dial down into the micro level right away, forgoing even the most basic summaries of performance in favor of critiquing buff uptime or flask usage. But, it’s important to remember that fight time, incoming damage, raid composition, etc., are all major determining factors on how easy or hard the encounter is on the healing team, and accordingly, how easy or hard it is for that team to perform well. So, as a starting off point when evaluating parses, I generally recommend that you look at the data from a macro level first, taking in mind that all the parsing tools out there will not explain strat or healing assignments.

Overall Healing

(1) Analyze --> Healing Done

Accessible through the main category menu, the Healing Done summary will provide you with a brief overview of the healers’ performance through the course of the encounter. However, I much prefer to use one of WoL’s recent additions—the Analyze: Healing Done tool, accessible through the same main category menu. In contrast to Healing Done, the Analyze tool will show me values and percentages for the following categories:

  • Effective, raw and over-healing: will tell me how much or how little a player was spamming heals during the course of an encounter
  • Effective and raw hps: similar to the above, the difference between raw and effective hps will tell me how much or how little a player was healing.
  • Activity: will reveal how present the player was in the combat logs, meaning that boss effects and casts were recorded for the player in question. (It is important to note that this does not mean that the player was actively casting the entire “Activity” period.) Lower times/percentages of activity typically mean that the player died.

But, as so many before me have observed, the overall healing meters simply gives general feeling for the fight. Although I will note how a player performed in comparison to the rest of the healing team, I also know that overall meters are only a small part of the puzzle. To get more information regarding a healer’s performance, I need to dig a little deeper.

Verdict on our sample parse: (WoL sample) Hrm, this doesn’t look too good. I’m 4th on the overall healing meters and my hps is pretty low. Now granted, the healing from my Healing Stream totem is not included in my total, but that wouldn’t make up the difference. So there must be something else at work.

Healing by Target

(1) Analyze --> Healing Done --> By Source: [Player]

The second place I look when evaluating healing is the customizable “by Target / by Source” listing. By selecting the “by Target” option on the left navigation menu, and then specifying “Vixsin” as the Source, I can see all healing I did on all friendly targets. By doing so, I can see if I stuck to my assignment (which in this case, is tanks + raid) or if I possibly wanted to fluff the meter by healing the Abom to the exclusion of all else. If I wanted to further evaluate healing on a target, I’d select the “by Spell” navigation option and then specify “Vixsin” as the source and the target in question as the target. Unfortunately, the target listing only shows players and not friendly targets, so I can’t confirm whether or not my healing on the abom was incidental or intentional.

Verdict on our sample parse: (WoL Sample) It looks like I stuck to my tank assignment pretty well, and contributed a good amount of healing to the second and third tanks (along with the abom). But the one anomaly I see is that I healed myself a significant amount—that’s something that bears more investigation.

Healing by Spell

(1) Analyze --> Healing Done --> By Spell [Any Source, Any Target]

As any healer will acknowledge, good healing is about finding the right tool for the job; proper selection can lift a healer’s performance or sink it quickly. So, to get a better idea of what spells are “right” for each fight, I use another subset of the Analyze feature—the Analyze Healing Done: By Spell listing. By leaving the Source and Target fields empty in this search, I can see the effective HPS value of all of the healing spells used during the encounter. In the past this has proved incredibly valuable information in helping me what sort of healing flow I should adhere to. Spells ranked higher in the list are spells that should be prioritized, while those at the bottom could be either ill-timed or ineffective given the fight’s circumstances.

What’s additionally important to note here is that the graph above the table of spells doesn’t just provide a pretty illustration—the peaks of any tracked spell will show when its highest HPS.

Verdict on our sample parse: (WoL Sample) With Chain Heal second to the top of the list, I get some reinforcement on the fact that I was using this spell more often than not. However, the high overheal percentage associated with Chain Heal, and all the top HPS spells for that matter, means that there likely wasn’t much to heal during the majority of the encounter. The plots for all the top-performing spells reveal that to be the case, along with demonstrating that the end of the fight was where the healing output became more intense.

Survivability

(1) --> Survivability

Just like with DPS, a dead healer doesn’t do anyone any good, so the next step in my healer analysis would be to look over the Survivability category to see if low activity times might correlate to early demise. WoL provides a quick overview of each player’s activity for the fight in the “Survivability” breakdown (which can display for as many or as limited number of fights as you designate, even going so far as to look at a specific time period). So, while “Activity” will display the total active time on the encounter, Survivability will indicate when a player met his demise. One thing to keep in mind when reviewing survivability numbers is that it will only display information pertaining to a player’s first death on an encounter. If the player was unlucky or unwise enough to die multiple times, only information from the first death will be shown. From the Survivability page you can also quickly access any player’s death log, by clicking on any value which is less than 100%.

Verdict on our sample parse: (WoL Sample) Ah-ha! Whereas my activity for the fight was 93%, my survivability was only 66.9%, indicating I had a slight mishap during the encounter. Looking over the death log, I can see that I was afflicted with Volatile Ooze Adhesive a couple seconds before a Malleable Goo hit me for 17k. So, it looks like I was rooted in place and in a fit of bad RNG, finished off by a Malleable Goo. But, this slightly explains the excess of healing on myself and could further indicate that I might not have had as much mana for the rest of the encounter, also explaining why I didn’t put up the HPS that I normally do.

Player-Specific Tools

After I’ve gotten a feel for the fight, I then dial down into the player detail to evaluate personal performance. Player-specific information is valuable in assessing management of personal cooldowns, supplemental contributions (like interrupts, incoming damage, etc.), and for checking player buffs and debuffs.

(At any point while you are perusing WoL, you can access information for a specific player by clicking his name in whatever table or breakdown you’re looking at. Alternately, you can access player details by selecting Players > [Character] from the detail menu.)

Healing by Spell

(3) --> [Player], Healing by Spell

To get a better idea of the spells a healer uses during the course of an encounter, as well as crit, overheal and uptime percentages for hots, the best place to go is the “Healing by Spell” window, found on the Player’s detail page. Clicking on this tab will display all healing done and received for the fight, including healing done and received through passive means, pots, and healthstones. Because there’s a good deal of information on this page, here are some tips to keep in mind:

  • For any class with splash healing or multi-target spells, it is important to note that the total number of heals (hits + crits) for that spell is not indicative of the number of casts.
  • WoL will record a critical as a critical, no matter the amount of effective healing. So, for this fight, my crit percentage on Chain Heal was 49% (213 crits / 438 total hits)
  • To get your effective healing you do not need to multiply your healing done by your overheal percentage—all stats shown under the Healing breakdown are effective numbers, not raw. (Raw data can be found under Analyze –> Healing Done)
  • Absorbs are shown in yellow because their value is estimated, but based on my own WoL usage, I would posit that the calculation of effective absorbs is fairly accurate.
  • To view a detail of who was healed by a particular spell, click on the spell’s name in the “Healing Done” detail window.

Verdict on our sample parse: (WoL Sample) First and foremost, by looking at my healing breakdown, you can see that I didn’t use much aside from Chain Heal on the Putricide encounter. A high overheal rate, combined with the high number of hits, likely means that prior to my untimely demise I was chain casting. To further verify the distribution of my casts, I could also narrow down the fight timeframe and compare the number of CH’s per second before my death and after it (but I think I’ll cheat and recall that yes, my mana was sparse after the death). ES procs look pretty low; further investigation shows that ES was set to apply on the 2 OT’s instead of the MT. Definitely something I should have caught. I can also see that I used a healthstone at some point, so I get bonus points for using the tools at my disposal to stay alive. (WTB Desperate Prayer for non-spacegoat Shamans, kthx.) So in all, there’s room for improvement but there are no major red flags.

Damage by Spell

(3) --> [Player], Damage by Spell

One of the great things about being a healer is that you have the tools to keep yourself alive when the stuff hits the fan, but when reviewing logs, high self healing can also be a sign that they player is covering up incoming damage that he could have avoided. So, to check and make sure that a healer wasn’t standing in slime and healing themselves through it (yes, I have seen logs where this is actually the case), “Damage by Spell” is the best place to go. The one thing to keep in mind is that the numbers presented in this tab are relative—they do not portray how much damage the player in question took in relation to all damage done (that information can be found under “Damage Taken”). It’s additionally important to review this tab with the fight in mind—there are effects which can be avoided through good play (Marrowgar’s Cold Flame, Death and Decay, Malleable Goo) that simply shouldn’t appear on incoming damage without good reason.

Verdict on our sample parse: (WoL Sample) As I noted in Survivability, my bad luck netted me a fixated green slime and a Malleable Goo at the same time, so the damage from those two spells is something I expected to see. My mutated slime damage is moderate, nothing too severe, but the one thing that does stand out is that I almost took the prize for most damage from the Abom’s mutated transformation! While not a huge issue in terms of damage (~2k/5seconds) it does show that I might be lurking in the Abom’s shadow a bit too much; a tendency which might have more severe consequences on hard mode.

Buffs/Debuffs Gained

(3) --> [Player], Buffs Gained

Although I’m not the type to hold up a ready check while I scream for Kings, I will acknowledge that group buffs are an important factor in beating an encounter. So, to evaluate the raid’s buffs over the duration of a fight, I’ll look at the “Buffs Gained” tab of the player detail window. Selecting this view will show me all buffs and debuffs applied to a player through the course of the fight, by any one of the player’s teammates or adversaries. The tab does includes auras and class-specific buffs (including mana tide totem), though you should understand that uptime on these effects may vary as range becomes an issue. For example, Moonkin Aura at 74% doesn’t necessarily mean that our laser chicken spent 26% of the fight in an alternate form; it’s more likely that I was too far away to gain the effect for the duration (this can be a red flag for bad healer positioning, though).

Verdict on our sample parse: (WoL Sample) There’s not much to note here other than passive buff uptime—Demonic Pact was at ~71% while Elemental Oath was at ~85%.

Personal Buff/Debuff Management (Buffs Cast)

(3) --> [Player], Buffs Cast

The most controllable aspect of any healer’s performance is his/her own self buffs, so the “Buffs Cast” tab of the player detail can sometimes be the tab I skip to first when reviewing a log (though sometimes I leave it until last because I know abysmal WS uptime awaits me.) For any class, this is an excellent way to review trinket, totem or buff uptime, mana return (eg: how many IED ticks you got in) or check for proper execution of a CC assignment. Bear in mind that the “debuffs” section details which debuffs the player applied, not what debuffs were applied by other friendlies or hostiles. When reviewing the information presented it is important to distinguish between procs and constants, and to remember that some personal buffs are significantly more important than others.

Also, it’s very important to note (and I almost made this mistake when commenting on an application a couple weeks back)—food and flask buffs (along with Inner Fire, unless you recast it during the fight) will only be displayed when you zoom out to review the entire report.

Verdict on our sample parse: (WoL Sample) Overall, uptime looks okay but not outstanding for my necessary buffs. ES was at ~97% uptime (albeit on an offtank) while WS came in at ~78%. “Spiritual Trance”, the buff from the Shaman T9 totem, had 81.5% uptime, which given the fight’s multiple transitions and my death, would seem to be about right. I can also see that my dual Solaces are averaging about 88% uptime, so when I go through future mp5 calculations, I might want to adjust for a reduced amount of mana return. I used Berserking and Nature’s Swiftness one time apiece, which given that it was a 7+ minute fight, means I wasn’t using them to their fullest potential.

What to do with all this information?

First and foremost, it’s crucial to remember that the log you’re reviewing reflects a player’s performance on *one* night, on some selection of data. And as any good analyst will acknowledge, a single data point does not constitute a trend. Not only that, (though I hate to admit it,) we’re all human; we all have off nights. What you should be concerned about is when trends start to manifest over multiple parses, on multiple nights.

Secondly, logs are the first step to improvement, but they’re only a step. To improve healing performance you need to act on what you’ve learned (or help others to act). If you have incredibly low ES uptime, figure out how to alert yourself that your ES has fallen off—maybe having it as a small corner dot on Grid isn’t the in-your-face alert you need. Likewise, if that strat can be improved to optimize healing output or limit incoming damage, it’s on you to make the suggestion. (On several hardmode fights, it’s made a significant difference in my HPS when I was permitted to stand in melee and thus avoid a good amount of movement. But no one will know this unless I speak up.)

Lastly, and as I mentioned previously, meters don’t reflect strats or assignments. The healer who is assigned to running around after the Council Orb Tank will likely be fairly low on meters if he/she sticks to the assignment. Likewise, a raid who clumps up during Marrowgar’s bone storm will have healers with higher HPS than raids who disperse every which way. But meters have no way of showing these two factors of performance.

So in closing, I’m inclined to remind you of the tale of the blind men and the elephant—don’t fall victim to thinking you have an understanding of the whole when you’re only looking at one part.






15 Comments


  1. Alex

    Nice read!

    When analyzing buffs cast / buffs gained section you can additionally klick on the little hash key (#) behind every buff. this will show the fight timeline + the timeline of the active buff. this can help you to understand why there’s a lower uptime than estimated before.


  2. Wow, this is incredibly helpful. Some of these things I was already doing, but others I was not taking full advantage of. This should help my own self-analysis, and allow me to better advise others.
    .-= Wugan´s last blog ..Pimp My Resto Shaman: Shimy =-.


  3. I agree, this is amazingly helpful. I tended to disregard logs if the fight went well and analyse them only if problems arose at which point the combatlog tool promptly refused to work for me.

    Great read!


  4. Ateve

    One thing I was interested but can’t figure out how to get at is pulling out data about the Emerald Vigor buff on Dreamwalker.

    Any suggestions?


  5. I would hypothesize that the problem with tracking Emerald Vigor is that the logger typically doesn’t go into the portals, thus the combat log does not record what went on in the “dream” state. I went back through a number of our normal mode attempts, and I couldn’t even query it based on the spell name, so it simply wasn’t being recorded. Conversely, in hardmode, because the Twisted Nightmares debuff actually does constant damage based on your stacks, both inside and outside the dream state, it does show in WoL. An alternate way to be able to evaluate stacks (and who let theirs fall) is through the graphing function. If the healer’s stack dropped or they didn’t collect as many, they will simply have less healing output and their peak HPS will be noticeably less.


  6. Ateve

    Good to know. I am the one logging the raids and I went in but still can’t find the buff on the site. Maybe the combat log doesnt record it at all?


  7. That could be possible, but it doesn’t seem to make sense since the buff is applied and tracked through normal game operations. I think you might simply have to have your healers on the honor system when it comes to stack reporting. ^_^


  8. Good stuff. I’ve just started keeping a record of my logs for some PuG groups (I don’t get a chance to run challenging guild content atm) so would be helpful for a quick analysis of my general playstyle.

    This and the original Intro to WoL posts are enjoying time together on my bookmarks bar.


  9. [...] in Group 5, a resto shaman blog, teaches you how to analyze combat logs for [...]


  10. [...] in Group 5, a resto shaman blog, teaches you how to analyze combat logs for [...]


  11. [...] in Group 5, a resto shaman blog, teaches you how to analyze combat logs for [...]


  12. http://www.worldoflogs.com/reports/rt-55grlye99hvvbh91/spell/71941/?s=8906&e=9202

    This is our parse from heroic Dreamwalker this week, counting the number of Twisted Nightmare stacks.

    On normal mode, the Dream Clouds don’t seem to log anything. It’s very strange. My only thought is that it doesn’t log because the clouds do no damage to Dreamwalker or to you when the stacks build up. They might not be actual creatures.


  13. Tip for keeping your stacks rolling: Have your healers travel in pairs.

    Apologies for the double-post, Vix, I hit submit a bit early.
    .-= Borsk´s last blog ..ICC25 Hard Mode: Power Overwhelming =-.


  14. [...] in Group 5, a resto shaman blog, teaches you how to analyze combat logs for [...]



Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

CommentLuv badge