Over the past year, I’ve authored a couple posts about how to evaluate combat logs, what to look for, and what to focus on. And second to those emails about gearing, one of the most frequent types of requests that I get from players are to look over their logs and provide some feedback. But, when it comes down to it, reviewing logs is more art than science, and coming to good conclusions oftentimes takes more knowledge than is presented in the litany of tables that characterize any parser. So today, instead of giving you vague or generalized advice about how to improve your performance as a resto shaman, I thought I’d invite you all into my own head and describe just what it is I look for when I turn that scrutiny inwards.
Before We Get Started
If you’ve been reading LiG5 for any measure of time, you’ve no doubt noticed that I put a lot of weight into meters. And I’m definitely not shy in stating that I think healing meters are a valuable tool. Unlike other players who might consider any kill a win for the healing team, I’m rarely happy with my performance on any fight unless I’m doing two things: 1) improving on my previous week’s performance, 2) doing more than my fair share of the healing. Now I realize that this may not be your perspective, but let me at least present one reason why you should listen to my crazy meter ramblings anyways—because everyone can improve. There is no skill cap in WoW.
While I’m always interested how I compare to the rest of my healing team, I’m generally more interested in my own week-to-week performance on encounters. So let’s take a look at some of my parses over the past couple of weeks, highlight some of my successes and failures, and hopefully come away with some conclusions about how you can increase your own performance.
Example 1: Heroic Halfus
- 3/29: 14,046 HPS / 4.2M Healing done (WoL Link)
- 3/15: 12,692 HPS / 3.6M Healing done (WoL Link)
- CompareBot! comparison
One of the first things I look for when comparing parses is to make sure that the duration of the two sample fights is about equal. In the case of these two Halfus parses, they’re separated by a mere 15 seconds, which tells me that I’m at least looking at parses with potentially similar damage patterns. (This requires some knowledge of the fight and requires that you understand how small time shifts can impact stats). To confirm that the parses do indeed have similar damage patterns and potential healing, I’ll also take a look at incoming damage as well. Given that the difference between our two samples is 3.5M damage taken, I know that in one fight, there was an increased threshold for healing (meaning, shockingly, I had the potential for higher HPS in the one where there was more incoming damage).
Now, let’s get into the parses themselves. Throughout this entire post I’m going to be using a great tool called CompareBot! designed by Seriallos of Roll Initiative. This tool will allow me to look at up to 3 parses side-by-side, so that I can evaluate the differences in my performance or the effects of any stat changes. (Suffice to say, CompareBot is one of my “go-to” tools these days, and the guy behind the tool has been nothing but accommodating with the assorted mass of questions that I’ve sent over the past month or so.)
So onto the Halfus parses that I’m going to be comparing. Like I mentioned previously, the parses are separated by 3.5M incoming damage and about 15 seconds of time. Both fights had 7 healers, and about the same healing comp (one pally was replaced by a disc priest in the second parse). When looking a bit deeper into the comparison, I note:
- My uptime on ES is much better on the higher HPS parse (93.4% versus 82.4%)
- There was much better PW:S uptime (shown as damage absorbs) in parse #2, at almost a 1:2 ratio, which would explain why my HPS was lower and why less damage was incoming.
- There was equal TC return in both the parses, but WS gains were greater with worse uptime (84% versus 92%). The difference? More crits on more CH casts. Interesting …
- AA’s benefits are horrible when you don’t cast a majority of single-target spells, which leaves me wondering—maybe a secondary resto spec, sans AA, would be worthwhile in 4.1 when I need to accommodate Spirit Link Totem.
Now while the above observations are all well and good, there’s one bit of information that’s not shown on the parses that might cast a different light on things—on the higher HPS parse (parse #1) I had about 600 more haste and 4% more crit than on the lower HPS parse (#2). Now you might say, “well you had higher HPS because you edged out other healers with all that haste” but in fact, the healing distribution doesn’t support this either—parse #1 I did 14.41% of the healing while in #2 I did 14.31% of the healing. In fact, if you take the percentage of healing that I did (~14%) and us that prorate the additional healing (3.5M), you’ll see that I should have increased by ~500k healing, which is actually more than the increase in healing that I actually saw. So what made the ultimate difference in upping that HPS threshold? One more breath from the Proto Behemoth and a slight bit more melee attacks.
So what does this parse tell me about how to increase my hps? Not much, in my opinion. While it does point out a number of potential areas of improvement (grrrrr, shield uptime), the biggest thing I take away from it all is that while the haste might have helped up my HPS, in the end it’s brute force healing that wins out. The source of my big HPS boost? 2/3rds of it was due to an increase of the hits per minute of HR (243 versus 185 per minute) and 1/3rd of it was due to an 8% increase in crit on CH.
Example 2: Heroic Cho’gall
- 3/23: 11,535 HPS / 6.6M Healing done (WoL Link)
- 3/31: 10,267 HPS / 6.0M Healing done (WoL Link)
- CompareBot! comparison
So for this next comparison we’re going to be taking a look at two Heroic Cho’gall parses of mine, which occurred on back to back weekly kills. Again, it’s important to note that while the time on the parses is relatively equal, there’s one glaring discrepancy—I did almost 600k more healing in Parse #1. So, with only a 7% overhealing difference and a mere 11 seconds in duration, the question becomes, where did all my mana go? One look at my healing distribution should give you the answer: into more expensive single-target spells—HS and GHW. The table below shows the distribution of my healing done for the fight, by multi-target, single-target and healing-over-time spells:
|Parse #1||Parse #2|
To me, this begs the question of: why was more single-target healing needed? Well, let’s take a look at tank damage for the two encounters:
|Parse #1||Parse #2|
|Total Tank Damage||15,738,140.00||18,399,083.00|
|% of damage for fight||30%||36%|
So, while Tank 1’s incoming damage increased slightly, Tank 2 saw a substantial difference in incoming damage. Because Tank #2 is under my watch (that poor, poor bear) this means I had to spend a good amount more of my mana keeping him up with powerful direct heals, like GHW and HS, instead of using him as a CH bounce target. What the meters don’t tell you is that his sudden squishiness was due to a change our tanking strategy to account for the fact that Grounding Totem no longer mitigated Flaming Destruction. So, tank #2 stepped up to be much more of a punching bag.
Now another interesting thing that crops up on this parse (and this is something that’s been on my radar for the past month or so) is in regards to AA’s healing. Arguably, with my low crit levels, my AA performance is fairly low for most encounters, which is to be expected. But in this parse comparison, we see something slightly odd—in parse #2, where I clocked almost 500k more single-target healing, my AA performance actually experienced lower overhealing, even when my total overhealing for the fight was 7% higher. So what gives?
I think the answer lies in timing. Where it would be normal to see AA clock 10-20% overhealing on most encounters because tank damage typically doesn’t align with raid damage, Heroic Cho’gall is one of those fights where it does, in spades. And so, when I switched to single-target healing while AOE was going out, my AA procs all went straight to effective healing. But then, when I tried to catch up with AOE healing, players were already at levels of HP where they didn’t need that much assistance. So what this tells me is that I was playing catch-up a little too much in Parse #2, and that if I knew more single-target healing was needed during AOE (which I did), I should have laid down a HR earlier rather than later.
Example 3: Heroic Omnotron Defense System
- 3/22: 12,231 HPS / 4.4M Healing done (WoL Link)
- 3/29: 11,885 HPS / 4.6M Healing done (WoL Link)
- CompareBot! comparison
The last fight that I’d like to take a look at is Tron Council, for the simple fact that of all of the fights out there, it’s one that for me requires much more single-target healing. The two parses that I’ve pulled are both from the end of March, and both were logged while I was in my Mastery gearset (~17 Mastery Rating). Starting off, there are a couple things that stand out at me:
- WS uptime for both fights was very poor—75% and 65%, respectively. ES uptime was a slight bit better at ~85% for both, but this still can be improved upon.
- Chemical Cloud damage was insanely high in Parse #2—27k to 195k in total damage done—which leads me to think “WTF was going on with me that night?” Whether due to sloppy play or bad positioning, this type of damage is totally unacceptable. (It’s worth noting that in Parse #1 I cast Instant GW a total of 7 times, versus 1 time in Parse #2, suggesting that fleet feet played a vital role in reducing that incoming damage).
- I used TC very well in Parse #1, racking up almost 4k return per LB cast (factoring in the cost of the 9 LB’s cast over the course of the encounter, it becomes ~2500 mana return per LB). This was a definite improvement on Parse #1, where I simply didn’t use TC at all.
This parse comparison also happens to be a great example of something that’s been more on my radar as of late—Riptide usage—and more specifically, how often I overwrite an existing RT hot with a new application. Because I’m glyphed for extended RT duration, it follows that I’ll get the most out of my RT hots when I can spread them out as much as possible and keep from overwriting them with new applications before the 21-second window is up. So, there are a couple indicators that I can evaluate in order to determine how well I’m doing with RT. The first one is Surging Tides uptime, as represented in the “Buffs” tab; Parse 2 had ~57% uptime for the 4pc bonus, while Parse 1 had 45%. This means that we can approximate the number of RT casts for each parse, and using the number of recorded ticks, calculate the average number of ticks per application.
- Parse 1: 458k healing, 126 ticks over 27 casts = 4.6 ticks each (Query for Riptide targets)
- Parse 2: 544k healing, 200 ticks over 38 casts = 5.3 ticks each (Query for Riptide targets)
Now remember these are glyphed RT hots, which should, at my level of haste, tick a total of 8 times before expiring. So the first thing we can recognize is that I’m overwriting frequently, and losing out on about a third of the benefit to the RT glyph. If you look at the target queries linked above you can see why—both the druid and the DK who appear more frequently than everyone else combined are our two tanks for the encounter. So that trap that I’m falling into is reapplying RT on tanks to keep my Tidal Waves uptime high. In Parse #1, it’s evident that I was keeping RT on tap for tanks; in Parse #2 I spread the love out a bit more, and was rewarded with an almost 18% increase in healing done.
The last thing I wanted to call your attention to with regards to this parse comparison relates to mana usage. Although there was only a 1% difference in the total single-target healing done for the two encounters, the spell distribution was drastically different. Parse 1 had me casting 20 HS, 11 GHW, and 15 HW, while Parse 2 recorded only 5 HS, 25 GHW, and 10 HW. But the really interesting thing, at least in my opinion, is that the difference in mana cost for the two rotations was only 2k mana–the cost of a single HW.
Hopefully, the preceding analysis has helped open your eyes a bit into just how much you can learn from your own combat parses if you only spend the time to go through them. Obviously, everyone in this game has room to improve, even us snooty blog writers, and I do hope I’ve proven that to you as well. Even though I’ve racked up a handful of solid parses, the fact of the matter is that I can still fail at the basics (like *ahem* shield uptime).
In terms of key “findings” from the above parse comparisons, I think there are a few notable ones:
- Make more use of Instant-cast GW. Perhaps the best thing I ever did was bind GW to mouse wheel—one scroll and I’m zooming out of all the bad.
- Pay attention to your RT targets and your hots. If you can’t see Riptide in Grid/Vuhdo/Raid Frames, make sure you can, so you don’t inadvertently overwrite a previous application. And in the event that RT comes off CD and there isn’t anyone to heal, then that’s a point where you should simply be reapplying it to a tank or other raid member who’s likely to take damage in the near future.
- Understand how your “rotation” choices will affect your mana. The more you think you need the fast, expensive heal, the more you have to be willing to pay for it. While my effective healing was pretty close in the Tron Council example, it could have played out much differently if my overhealing on HS had been higher.
And of course, in the end, I’m left with the thought that I always have when reviewing logs—your team’s and your own strategies play a significant role in how “well” you parse. So it’s important when you compare yourself to other players and other parses that you make sure you take a look at the whole picture. A priest who does nothing but spam PoH all day makes a bad comparison against one who’s using Penance on CD. Or in Resto Shaman terms, looking to Paragon’s shaman and their first kills of heroic bosses, and modeling yourself after them without care for strategy, will inevitably lead you down the wrong path. But take a step back, understand the circumstances, and identify where your similarities begin and differences end, and the world (of warcraft) will become your oyster.