Usage Rate New Template.jpg

What’s up guys and gals, CaptainPlanet here to present the Overwatch Hero Tier List and Meta Report: Omnic Crisis Averted! Season 4 of Overwatch competitive play has just begun and the new era of our favorite game has already moved past its first crisis. I’m talking about the Second Omnic Crisis, also known as the week we just endured with the new, too-much-improved Bastion. This report does not include matches from that week (only a few matches of the Carbon Series were played on the patch), but players from pro to casual alike could attest to the strength of the beep boop. Don’t take it from me, take it from Jeff himself:


Shortly thereafter, Overwatch’s newest hero, Orisa, was announced and she and Bastion nerfs went up on the Ptr. Not wasting any time, Blizzard met the Omnic Crisis head on by porting those same nerfs to live before the weekend, luckily just ahead of the Strivewire Monthly Melee. What follows will be a report on the professional/tournament meta in a post-Omnic Crisis world, taken from a sample of EU only teams. Blizzard has also just announced further balance changes as I was writing this sentence – it’s hard to keep up with all the changes these days! Assuming this balance patch goes live in the next few weeks, this report should provide a comparison point for the impact they have, for the pro scene anyway. Speaking of which…

new tier template.jpg


S Tier (>=95% Usage Rate): Lucio (96%)

A Tier (>80% Usage Rate): Ana (81%)

B Tier (>50% Usage Rate): Soldier 76 (66%), Tracer (56%), Winston (54%), Genji (52%)

C Tier (>20% Usage Rate): Reinhardt (46%), Zarya (39%), D.Va (36%), Roadhog (30%)

D Tier (>5% Usage Rate): Zenyatta (16%), Pharah (14%)

F Niche Tier (<5% Usage Rate): Mercy (4%), Torbjorn (2%), Mei (2%), Bastion (2%), Widowmaker (2%), Reaper (1%), McCree (1%), Symmetra (1%), Hanzo (0%), Sombra (0%), Junkrat 0 Picks

Tiers/Usage Discussion

Help me make Tiers better

Recently Jeff Kaplan reflected on what the word “meta” means, and what he thinks people believe it means:


This made me think about how to better present my tier graphic in particular, to help curb some of the negative behavior that some have used it to justify. Jeff mentioned that for some people, a “balanced meta” refers to a theoretical world where every hero gets picked at the same rate. He concludes that he’s not sure that it’s realistic, nor that it would necessarily be good for Overwatch as a whole, but I (initially) thought that this “equal” rate would be an interesting baseline for comparison. My hope was that while this perfect average could never (and probably shouldn’t ever) be reached, it would help separate out both the “must-use” heroes and “niche” heroes more clearly. Jeff references an equal pickrate of 4-5% (100% / 23 heroes), but since my data is presented as usage rate, (100% * 6 lineup slots / 23 heroes), it comes out to roughly 26% usage rate. I took this average rate, then created a new Tier Chart. Each tier range is still chosen arbitrarily – but each cutoff is based on some fraction or multiple of the 26% usage rate mean. It looks like this:

new new2 tier template.jpg

You can interpret the chart as follows: “Soldier and Tracer were used more than twice the theoretical average rate (26%), while Zenyatta and Pharah were used between half as much and as much as the theoretical average”. Sounds confusing right? Having done this, I have to agree with Jeff. Even when basing around a theoretical “perfect average”, it’s still too hard to perform any meaningful analysis on heroes with such low sample size like those in the F Tier. All I can do is say things like “Symmetra was picked on only 3 out of a possible 55 maps, and played for just over 5 minutes total out of the 14+ hours of playtime in this dataset”, but that doesn’t really tell you if Symmetra is a good or bad hero with any certainty. I can only say that of those three matches she was played, only one of those resulted in victory. I leave it to you, the reader: how do you prefer your hero usage data? Do you want the tier graphic to stay the same? Do you prefer basing it off of the theoretical mean? Or something else entirely? Let me know in

This Strawpoll

The Strange Strivewire Situation

The Strivewire Monthly Melee was a tournament put on for the EU region that allowed teams to draft their own maps, contrary to Blizzard-licensed tournaments which generally require a map pool at minimum to be selected from. Left to their own devices, European teams picked King of the Hill or Hybrid maps 80% of the time:


This means that this week’s data should give us a good idea of a very specific cross section of the pro scene: what the EU region likes to play on King of the Hill and Hybrid maps. As it turns out, there were two competing lineup philosophies on display in the Strivewire Monthly Melee: the Dive comps of Bench Boys and Cyclone vs. the Quad/Tri Tank lineup of b0nkers (now Red Reserve). These three teams were the top three finishers in the tournament, and thus their lineups were the ones most represented in the data. What does this mean? Lots of tanks, and lots of Tracer and Genji. The top three most used heroes were actually Lucio, Ana, and Soldier 76, in part because all three of these heroes can be and were used in both dive and triple tank lineups. At least for the current, pre-Orisa, pre-balance change patch there is not a 100% consensus on which lineup is stronger: b0nkers swept their way to the finals leaning heavily on their tank-based lineup before losing to the Dive comp of Cyclone. For a more detailed look into what heroes worked on what maps, check out the chart below:


Direct Link

As always, you can drive down deeper into the individual data points by mousing over and clicking on the boxes representing hero usage for each map and game type. For example, we can see Tracer was very successful on Nepal in particular thanks to strong showings from Twoeasy and Narwak. Also noticeable: Bastion saw very little usage, but still saw greater than 0 usage. I guess that’s a win?

Since the upcoming balance patch seems to favor dive strategies in particular I’m going to take the time to unravel what makes that team composition tick.

Dive in to Dive Comps

What is this…Dive?

Dive lineups are more of a concept rather than a strict set of heroes, revolving around the idea of building tempo from striking first. The idea is that the team who engages the other first will have a positional and reactionary advantage having caught their opponents by surprise. Moving quickly into the back lines of the enemy in unison is paramount for dive comps on offense in particular, where defenses will usually have time to set up in advantageous locations. Even in the week of the Bastion domination period, the pro community speculated that Dive was one of the few non-mirror compositions potentially able to deal with entrenched Bastion defenses by attempting to catch the less mobile hero off guard.

This strategy can work on defense as well. The saying goes: “the best defense is a good offense” and the same is true in Overwatch. After all, the defense still has to brutally murder all attackers who wish to take the point or push the payload, so they’re still an offense at heart – just one that happens to be facing the other direction. Defending as a Dive comp has its own wrinkle of complexity as well: spawn camping/splitting. For any defense, preventing the cart from moving or point from capping is the ultimate goal; making sure your team contests and fights on the point is the means to that end. For defensive dive comps however, diving past the point to engage the offense is just as good, if not better. Back-capping points and/or payloads is next to impossible in Overwatch, so any fight won on or off the point is a definite win for the defense.

Getting first blood on defense can be just as effective as offense: anyone who’s ever been about to push on attack knows how deflating losing a Lucio can be, if the defense picks him off the instant before the “Go!” signal. Defensive dive comps then push the suddenly advantageous 6v5 to the limit, getting on top of the offense to try to wipe them all out - - charging their ultimates in the process. More often than not, defensive dives can stall out killing the last attacker (often a baby D.Va or support) to keep the clock ticking on this advantage. In recent times, teams like Selfless have begun to camp the attacking spawn, potentially forcing an extra team fight away from the point which again, works to defending teams’ advantage.

What makes a Dive lineup?

So you want to build a dive lineup? Let’s remember the main goal of Dive: make the first strike by getting into close proximity to your opponent. Heroes with the most mobility make sense then, bringing us to the traditional Tracer/Genji/Winston/Lucio Dive core. Teams treat Winston like a cannonball: his leap into the back line of the enemy is often the signal for the dive to begin.

Different teams have different triggers for their Winston initiations, depending on what flavor of flex heroes they prefer. Some teams, for example, will use a Widowmaker in a “slow” Dive lineup. These teams will bide their time, waiting for their Widow to get a pick, a body-shot, or even just force a mistake in positioning. Target acquired, the Winston will dive in with the Tracer and Genji in tow, forcing the enemy to make a choice between the lesser of two evils: stay put and fight the Winston and friends, or run out into the open into the line of sight of the Widowmaker. Other teams prefer using Pharah, when the map or player hero pool favors her. King’s Row and “the Mangachu route” is a popular opening for Pharah Dive comps, and it looks like this:

pharah 1.jpg

Pharah Dive comps will usually follow this sort of example: very fast engages that get the Pharah in a position where the opponent has to choose between focusing on her, or the rest of her team like the Widowmaker variant.

Dive lineups are interesting in that they require exceptional individual skill from their players and cohesiveness from the team in order to both get kills and execute the strict timing of the engages. Dives fall flat without kills, but dives fall flat without good post-dive positioning too. Dive comps are also a lot more flexible than their tri/quad tank counterparts, in part due to the greater number of DPS heroes to pick from in the first place. For example, Koreans will run a Tracer/Genji/Lucio Dive Core with a Winston, D.Va, McCree, Soldier 76, Pharah, or any combination of the group, or any hero they fancy at the time. As mentioned, Western teams tend to pick their flex heroes based on their personal strengths and/or the map geometry. With changes to Zenyatta on the way, I expect him to make his way into more Dive comps in place of Ana – who is slated to be nerfed as well.

Don’t Dive into the Deep End

You shouldn’t blindly follow everything that the pros do, but Dive comps require a special added emphasis to that warning. Dive comps rely on high individual skill and precise timing and teamwork. Both of which you’re not very likely to come across on ranked play on their own, much less together. If you think you’re up to create some Dive strategies of your own, I suggest watching as many Rogue, Bench Boys, Rise Nation, and Selfless VODs as you can to see which flavor of Dive comp fits your hopefully skilled group of 6 friends. Because trying to convince 5 pubstars to execute on a Winston leap with a Widowmaker providing long-range fire or being the Pharah Concussion-Blasting around corners is certainly not a task that I feel up to tackling. I hope you all are stronger willed than me!

Final Thoughts and Shoutouts

Shout again a second time to Noukky for organizing the Strivewire Monthly Melee and putting up some of her own money to fund the prize pool. Also shoutout to GosuGamers for allowing the tournament to be hosted on their Twitch channel, which I’m sure helped its viewership. EU Overwatch is not yet dead – so also thanks to all of you who tuned into the tournament and contributed to the crowd-funded prize pool. An announcement on my end as well: I’ll be on vacation (again?) this weekend and travelling out of town, so no Meta Report until probably after Orisa and the new balance patch releases!


Until next time,