Proof you can play any spec and perform well

When players start a new class they all have the same question: “What spec should I play?” After analyzing data from hundreds of thousands of data points, I’ve come up with the answer.

Play whatever you want. Then practice being good at it.

I don’t think anyone would disagree when I say that being more skilled at the game can increase your damage significantly. But what I’ve found is that focusing on mastering your class increases your damage so much, that the spec you play isn’t a big concern. I’ll show you why with the magic of numbers!

Master your spec


I gathered data from Warcraft Logs and cross referenced item levels with Ask Mr. Robot’s specific gear data from logs. Data points included the following parameters:

  • Spec
  • Hellfire Citadel bosses
  • Difficulty
  • Gear
  • Skill levels
  • Boss kill times

Data consultants: I also consulted with Kihra, the creator of Warcraft Logs, to ensure I had all of the right data and interpreted it correctly. I also consulted with a math PhD and a statistician to verify that I was doing ‘good math’ and analyzing the data correctly.

50th percentile: Lastly, it’s important to know that when I talk about the 50th percentile player, that’s the average player. This isn’t like getting 50% on a test. This is a distribution, like a bell curve, most people fall in the middle. By definition, someone who ranks at the 50th percentile is better than half of the people who kill a boss.

Discarded data: Some specs really had to be left out of the analysis due to so few data points, including Protection Gladiator Warriors, Demonology Warlocks, and Survival Hunters (on Mythic). I also threw out the Kilrogg fight because of the damage buff. Iron Reaver actually did not need to be thrown out, since the very small number of sub-minute kills barely impacted the massive amount of data from everyone else.

OP Fire MageMastering your spec: 30-40% increase to damage for the average player

If you’re an average Heroic player and focus on improving, you’d gain 40% damage when you reached the 95th percentile. For Mythic, that DPS gain is 31%. As you can see, even with Mythic players, there is still a large skill gap. Mastering a spec takes a lot of research, practice, and continued motivation.

The different skill gaps for each spec are also an indicator of how hard that spec is to master. Almost all specs hover close to the average. But a few really high and low ones stand out.

Heroic skill gaps:

  • Harder to master – Fire mages (57%), Arcane Mages (52%), Destruction Warlocks (50%)
  • Easier to master – Frost and Unholy Death Knights, Balance Druids, and Enhancement Shaman all have 31-33% skill gaps

Mythic skill gaps:

  • Harder to master – Fire Mages (65%), Arcane Mages (43%)
  • Easier to master: Frost Death Knights, Frost Mages, Shadow Priests, and Assassination Rogues all have 24% skill gaps.

The skill gap differences were calculated as follows: (damage of 95th percentile – damage of 50th percentile) / (damage of 50th percentile).

Spec Balance

How much more damage would you do if you played the best spec?

Not nearly as much as if you focus on mastering your spec. But you’ll see at least some of an increase if you aren’t playing one of the best specs.

For an average player, if you are playing the spec that’s smack in the middle, on Heroic, you  are 7% away from the best spec. If you’re playing the worst spec, you’re 7% below the average spec, and 14% below the best spec.

On Mythic, the average player sees a difference of 9% from the middle, and 18% from the best to the worst. If you’re in the 95th percentile on Mythic, the difference are larger, likely because the best players are able to squeeze every last trick out of specs that can outperform, widening the gap to 12%.

But remember, one spec isn’t the best on every fight.  And of course I gathered data to paint an accurate picture. Is this literally paint by numbers then?

I looked at how many times each spec was in the top 3 ranks or in the bottom 3 ranks, on each fight. I looked at the 50th and 95th percentile rankings, and if a spec showed up in top or bottom 3 for either skill level, I counted it. If the spec showed up in the top or bottom 3 on both skill levels, they were counted, but only once.

Specs averaged being in the top 3 on three boss fights, out of 13 Hellfire Citadel bosses (I included Kilrogg for this). The average for the bottom 3 is also three times. But it’s interesting to look at the actual distribution, so here’s a graph.

compare spec balance


All conclusions in this section are based on data filtered to appropriate gear levels. This is very important to know, because over-gearing is a different story.

Over-gearing destroys the balance

“Fire Mages are OP!”  How many times have you heard that (or thought it to yourself?). Well, it is and isn’t true.

When it isn’t true:

When appropriately geared, their damage is not exceptional. In fact, for the average Heroic player, 7 specs perform better, when averaging across all boss fights. Balance Druids and Affliction Warlocks actually take the top 2 spots. However, when looking at the 95th percentile, Fire mages do perform better than all of the other classes, so if you master it, it is strong, but not over powered.

When it is true:

When people out-gear the content, things definitely lose balance. Fire mages in particular become pretty ridiculous, pretty quickly.

Remember back in November when Valor upgrades launched again? Within 2 weeks, people quickly out-geared the content and we saw huge damage increases for a bunch of specs, but Fire Mages shot way ahead of everyone else.

Rankings over time

WarcraftLogs damage per spec, over time, for all gear, heroic difficulty.

So what spec should you play?

Look up your logs. Both Ask Mr. Robot and Warcraft Logs show your percentile ranking for your item level. While you’re at it, look at logs where you had an appropriate gear level for the fights (so you might have to go a ways back).

If you rank at 75% or below on most fights, you’ll get more damage out of playing a spec that gets you excited enough to practice and become better. That 30-40% skill gap will get you further ahead that being average at the best spec.

If you consistently rank in the 95th percentile, you’ll see a damage increase if you play one of the ‘best’ specs, assuming you can master it. If your whole team isn’t in the 95th percentile though, don’t feel pressured to switch, because a skilled player does 30-40% more damage than an average player of the best spec.

To put this all in perspective, here’s a chart that shows why skill trumps playing the best class.

Skill trumps spec

Pre-Patch & Legion

No one knows what the best specs are for Legion, especially since tuning isn’t done yet. My advice: see what feels the most fun and pick the spec that interests you the most. Then practice! No one even knows the besT way to play each spec yet, so have some fun and experiment.

Join the Discussion

Blizzard Watch is a safe space for all readers. By leaving comments on this site you agree to follow our  commenting and community guidelines.