Analyzing Performance With Logs

Now that I’m running heroics I’ve been spending some time using Warcraft Logs to analyze and improve my overall raid performance. Note, I said overall performance, and not just DPS. DPS is a big part of it, but it’s hardly the only thing that I look at.

That said, I’m going to begin by talking about DPS, because it’s what most Hunters tend to care about the most.

There are three major contributors that negatively impact your DPS. First is dying. Not much to say there, other than to remind you of the old adage, a dead player does no dps.

The second factor, and realistically the biggest factor, is not maximizing global cooldowns. During the heat of battle it can be hard to notice that you aren’t firing your signature shot as soon as it comes off cooldown, but you’ll be surprised at how challenging that can be, especially if you’re just learning an encounter.

Here is a look at damage done for a Thok 10-man normal I did recently. I love this parse, because it provides me with a fantastic baseline with which to compare all my other encounters.

Our strategy was to separate the raid into three groups and strategically place them around Thok. We then proceeded to DPS him without forcing a second phase.

I got to stand in one spot and just DPS the boss without having to do anything else. This was more of a Patchwerk style fight, than Patchwerk itself.

Here’s an overall look at the damage done for this fight. As you can see, even with my mighty Sporebat Kitteh at my side, we were destroyed by, as BigRedKitty would say, a cockroach Pally.

thok_damagedone
Thok 10-man Normal Damage Done

What I like to pay attention to when looking at overall damage is,

  • Active – This is the percentage of the fight I spent actively casting abilities and DPS’ing the boss. You want this to be as high as possible. 98.98% is not bad. What this tells me is that I was probably a little slow out of the gate starting my DPS.
  • DPS(a) vs DPS(e) – DPS(a) is an average of your active DPS, while DPS(e) is your effective DPS. You want these to be close, but the more important number is the DPS(e).

So nothing too alarming here, other than I could have been a wee more active. Nothing a shot of espresso before the pull can’t fix.

That’s just a high level look. To truly evaluate things you have dive into the details. Here’s a look at all of the abilities I cast during the fight, and how many times.

thok_abilitiescast
Thok 10-man Normal Abilities Cast

As I said, one of the major reasons DPS suffers is poor GCD management. Let’s see how I did.

Kill Command is Beast Mastery’s signature shot. It does more damage than anything else, so it’s important to cast as many as the fight will allow.

This fight lasted 3 minutes and 21 seconds, or 201 seconds. Kill Command has a six second cooldown which means you can cast 10 per minute.

Over a 201 second period I could have gotten off 33 Kill Commands. Since this was literally a stand still fight where I had nothing to do but focus on rotation, getting off 33 Kill Commands was entirely possible.

As you can see I cast 31 Kill Commands during this fight, So I missed a couple. The horror. What could have possibly gone wrong? Was it an unruly pet? Ooh, I know, it was a boating accident! Let’s see if the logs corroborate either.

Here is a look at my Kill Command cast intervals.

Thok 10-man Normal Kill Command Casts
Thok 10-man Normal Kill Command Casts

I’ve highlighted three gaps where I waited longer than six seconds to cast Kill Command. Not the pets fault, and not a boating accident.

What you’re seeing here is some of the decision making that goes into rotations. In this case it was fire Kill Command and delay Bestial Wrath which was coming off cooldown, or delay Kill Command in order to line it up with Bestial Wrath. In this fight I chose the latter.

thok_bwrath_casts
Thok 10-man Normal Bestial Wrath Casts

My Bestial Wrath is on a 40 second cooldown. Five was the maximum I could have used this fight, but I probably had room to delay it a bit in order to squeeze in those two Kill Commands. Something for me to think about as I try to get better.

Another thing I like to look at is Serpent Sting. I check to see how many times I cast it, and what was the uptime.

Now there are times when you want to recast the Sting to really maximize output. I don’t always do that, but I know that there are occasions when when it’s beneficial to do this.

Thok 10-man Normal Serpent Sting
Thok 10-man Normal Serpent Sting

Looking at this fight, for whatever reason, I did not cast it until almost 11 seconds into the fight. Definitely not recommended behavior.

Because it was a standstill fight it never fell of, but this was a missed opportunity for sure.

A couple of other things I’ve learned from my logs is that once upon a time my Bestial Wrath had a one minute cooldown. This was perfectly aligned with my Engineering Synapse Springs. To keep things efficient I use a macro that casts Synapse Springs with Bestial Wrath.

I forgot to change that once I got Assurance of Consequence, and thus I’ve short changed myself out one or two of these in certain fights.

The other thing I need to do is take control of Rabid. Auto-cast is not a good way to go with this ability, and in this particular fight, I ended up missing one.

So as you can see even under ideal circumstance it can be difficult to execute a rotation with 100% perfection. All in all this was a pretty good parse. Let’s a take look at what happens when you’re still learning a fight.

Here’s a look at my Kill Command distribution from Heroic Malkorok. This was a 4 minute and 20 second fight, or 260 seconds. It was wipe, and, while I died after it was a wipe, the fight continued for 10 seconds while I took a dirt nap.

hmalk_kc_casts
Malkorok 10-man Heroic Kill Commands

You can see that I cast a total of 27 Kill Commands. Even if you subtract the ten seconds that I was dead that still leaves room for a total of 41 Kill Commands. So I’m still short 14.

It should be noted that I finished second on the meters, not that it means much. I say that merely to point out that I wasn’t dragging the bottom here, and yet lots of room for improvement.

Reasons for the missed Kill Commands have to do with a focus on surviving. I’m dodging orbs, running to stand in Implosions, running to safe spots, and soaking orbs with Deterrence.

Lots of room for improvement, and the logs help me focus on specific areas where I can do better.

Now as I said in the beginning evaluating logs isn’t just about DPS. Many fights have adds that are a priority, so it’s important to see if I’m being carried or carrying my weight here.

Sticking with the Heroic Malkorok encounters, here is the overall damage done to the Living Corruption adds.

hmalk_livingcorruption
Malkorok 10-man Heroic Living Corruption

I’m not the worst, but I’m not the best either. We had two corruptions spawn and killed after I died. I obviously did no damage to those.

What’s great about Warcraft Logs is you can look at your damage on each individual add.

By doing that I discovered that my DPS on the adds is hit or miss. I either recognize the add is up and crush it, or fail to see it entirely. Not a lot of middle ground here.

Next time we face this boss, I’m going to super emphasize the add in my boss mods, as I clearly need that.

Adds are just one mechanic that you can measure in fights, but as I continue to say it’s not just about DPS.

Here’s a look at Sha of Pride heroic. One of the things you have to do in this encounter is run around and close rifts. When you close a rift, you get a debuff that prevents you from closing another one for one minute.

sha_rifts
Sha of Pride 10-man Heroic Rifts

I closed four rifts during this encounter, but probably could’ve gotten five.

What stands out to me is the late start I got in attempting to close the rifts. I started out slow, but then got into a rhythm, but I need to try a bit harder next time.

Now one other thing I like to look at on this fight is the Mark of Arrogance debuff, because you don’t want to close Rifts when you have it.

markofarrogance

Nothing too out of the ordinary here, other than it took a while to get that last Mark dispelled.

For the most part the healers did a good job here and the debuff likely didn’t hinder my ability to close rifts.

Thus fare most of the analysis has focused and dealing damage. It’s also important to look at damage taken. Here is a look at overall damage taken from our Heroic Nazgrim fight.

hnazgrim_damagen_taken1
Nazgrim 10-man Heroic Damage Taken

Unless you’re a tank, you always want to be towards the bottom of this list. I did a pretty decent job, but I always like to see what I took damage from. Here you go.

hnazgrim_damagetakent
Nazgrim 10-man Heroic Damage Sources

I did well, but there’s room for improvement. Heroic Shockwave can also avoided. This was a player proximity issue.

Backstab is an Assassin ability, and it only hits you from behind, which is a big no-no in this fight. I’m lucking I didn’t end up dead.

I know what happened though.

hnazgrim_assassins
Nazgrim 10-man Heroic Debuffs

I was targeted by Assassins three time, and each time I made a bee line for our Warlock’s Demonic Gateway. One of those times I accidentally turned my back to him.

I think next time I’ll just disengage everywhere.

The final thing I like to look at is healing done. That’s right, healing done. Again, here it is from Heroic Nazgrim.

hnazgrim_healingdone
Nazgrim 10-man Heroic Healing Done

Remember that Backstab I took? At least I had the wherewithal to pop a cookie.

I also gave the Glyph of Liberation a try, which as you can see, I gained no benefit from whatsoever. I also probably didn’t disengage as much as I could have this fight.

The last thing I would suggest if you’re looking to improve your raid performance is to video your encounters. This can be incredibly helpful to diagnose what it is you do wrong.

I like to watch my movement, my use of globals, as well as watch boss mode timers. Having a video gives you the opportunity to replay the encounter with a relaxed and objective frame of mind.

Some things I’ve discovered by watching my videos from Heroic Malkorok are,

  • Not DPS’ing an add because I didn’t know it was up, even though there was a boss timer warning of the add two seconds ahead of time.
  • Not getting out of the way of spawning orbs in time. I thought I had avoided all of them, but that wasn’t the case.
  • Not maximizing globals due to movement.
  • I Need to emphasize certain events in my boss mod so as not to miss them.

I find that logs and videos combine to provide a wealth of feedback that I can use to help increase performance, and if you’re looking to do the same I encourage you to give it a go.

One Comment on “Analyzing Performance With Logs”

  1. Thanks for the detailed write-up of your analysis. It’s good to see how to use tools like Warcraft Logs to dig through our performance.

Comments are closed.