Opinion: The Overwatch League is hurting the game for regular players

The Overwatch League (OWL) began playing the second stage of its second year of operation over the weekend, with all the pomp and pageantry we’ve come to expect from Blizzard’s premiere esport. The Blizzard Arena in Burbank saw sellout crowds every night, with fans from all walks of life cheering on the six-man squads. But the game these pros are playing is having trouble retaining players, calling into question its viability as both a live service and an esport.

Overwatch’s development team has done an amazing job keeping the game relevant over the last four years, and it’s still a lot of fun to hop in for a few rounds of Quick Play or Mystery Heroes. What’s become much less fun is the game’s competitive mode, and I believe that’s due in large part to how Blizzard has decided to try and balance the cast of 30 heroes players can choose from.

Balancing Overwatch is a thankless task, and it’s impossible to make everyone happy. But by focusing their development staff’s attention on balancing from the top down, Blizzard is actively making their game worse for the majority of players.


Most of Overwatch’s recent patches have dealt with what’s commonly referred to as the 3-3 or GOATS composition, an acronym for the semi-pro team Greatest Of All TimeS, who were the first to experiment with the idea. GOATS play involves picking three bulky Tank characters with lots of defense and hit points, and teaming them with three Support characters, capable of sustaining the tanks, allowing them to move more quickly and healing damage done to them.

It’s a symbiotic relationship, as the Tanks keep the Supports alive by providing barriers that block incoming damage from the opposing team. When run correctly, it’s extremely difficult to single out and eliminate any single character in a GOATS comp.

The main problem here is that Overwatch is an objective-based game, so a team which can take and hold territory will always have an advantage over one which can’t. The GOATS comp succeeds by bullying its way onto an objective and sustaining itself until the other team makes a mistake.

This has led to a meta where teams have no choice but to play some variation of a GOATS comp if they see the other team is running one. Trying to do anything else will almost invariably result in a loss, and stage one of this year’s Overwatch League has been stagnant, with every team running slight variations on the theme.

It’s not especially entertaining to watch due to GOATS’ reliance on slow-moving, melee-heavy brawling characters rather than dynamic, skill-based heroes who can do damage at a distance. It’s not much fun for the players either. Fan favorites like the New York Excelsior’s Do-hyeon "Pine" Kim and Jong-ryeol "Saebyeolbe" Park have seen little play in stage one; they’re among the best in the world at what they do, but there’s no place for them in the current metagame.

It didn’t matter what game you watched in stage one; they all looked the same.

In fact, there’s very little for any high-skill hitscan players to do here, so some of the best in the world are forced to ride the bench since their particular skillsets aren’t useful in this meta. Even fans in the Blizzard arena have become bored by the GOATS comp, booing the Houston Outlaws when they appeared to be playing a damage composition, but ended up returning to their spawn point and switching to a mirror of their opposition.

This weekend marked the first time pros began playing under the new rules meant to curtail GOATS’ dominance, and we did see a little more diversity. But most of the teams were still relying on the composition as a fallback, making minor changes such as swapping a plodding Reinhardt for a more mobile Winston. Only the Chengdu Hunters seemed willing to shake things up, running a Symmetra-based mobility composition against the Paris Eternal’s bunker on Hanamura and camping the Eternal’s spawn on Blizzard World with Wrecking Ball.

It’s hard to say what kind of effect this is having on viewership, since the only hard data we have to go on is the number of concurrent viewers watching the games on Twitch. This has remained relatively consistent, but no one can say whether viewers are actually watching, or just leaving Twitch open to earn tokens which can be redeemed for cosmetic items in-game.

Blizzard has said viewership is up since last year, but the many changes made between the end of stage one and this weekend’s stage two opener imply a strong desire to shake up the tedious matches we’ve seen so far. Unfortunately, the changes don’t seem to go far enough, and we’ll likely see the GOATS composition dominating pro play until at least the third stage of OWL’s season two.

One little shield makes all the difference

So why is GOATS so good? A lot of it has to do with the interaction between most characters and the translucent barriers used by Winston, Reinhardt, Orisa, and Brigitte. These barriers are used to block incoming attacks, and there are very few characters who can get damage past them before they’re destroyed. With the exception of Moira’s Biotic Orb and Reinhardt’s Fire Strike (and a few Ultimate abilities) only close-range attacks and abilities can penetrate an enemy barrier.

It’s strange to say, but the best composition in what’s supposed to be a shooting game actually involves very little gunfire. Melee attacks ignore barriers, so once Reinhardt and Brigitte get into range they can overwhelm an enemy team simply by moving forward and attacking indiscriminately. This has two consequences: first, the GOATS composition is at its strongest when enemies are within melee range. Second, Lúcio is another key component due to his ability to move the team into position rapidly. Every GOATS team will include Brigitte and Lúcio, and usually includes Zenyatta to help the team focus fire on a single target.

GOATS became dominant after support character Brigitte was introduced in March of last year, and the character is seen as the enabler for this style of play. Brigitte was intended to curtail the dominant meta at the time she was introduced, the mobility focused “Dive comp” which wins by swarming a key target, then using movement abilities to escape and move on to the next victim.

Her kit was specifically designed to punish this playstyle, making her a hard counter to mobile damage sources like Genji and Tracer. She was also designed to be difficult to kill, able to sustain her own health with her passive Inspire healing ability and block frontal attacks with her personal shield.

She’s still fun to play, but in trying to make Brigitte a meta-changing character, the design team did their job too well. Brigitte’s abilities are so useful she’s completely supplanted the need for damage characters in a composition. Why play a character who can only shoot faces when you can play one who has the best burst heal in the game, heals multiple targets passively, can block damage for herself and others, has two health pools that both regenerate, can knock enemies away, and has the best interrupt ability available (which can even cancel some enemy Ultimates)?

She can even serve as a mini main Tank in a pinch, not to mention the fact that her Ultimate ability can swing a team fight all on its own by making her entire party more durable. As it stands, no other character in any category currently offers as much value as Brigitte can provide.

It’s telling that her kit has had its effectiveness reduced in every patch since her introduction, but most pros still consider her to be a must-pick. Her cooldown on Shield Bash was increased, the damage she deals with a bash was reduced by 90%, Bash no longer goes through enemy barriers, the health of her personal barrier was reduced, and Rally’s function was changed to provide less armor to teammates (and later, to degrade after a short time).

Additionally, many other changes meant to reduce her effectiveness have been made to other characters. Her father Torbjörn got a complete overhaul to make him particularly effective against armor, and recent changes to Reaper and Junkrat seem to be specifically designed to make them more viable against GOATS. The entire armor mechanic was reworked recently in an attempt to make damage characters worthwhile again, and fully two-thirds of the cast got some sort of change to their abilities in the most recent patch, all with the goal of disrupting GOATS. Nothing seems to be working.

Players and viewers have become increasingly frustrated with the attrition-based slugfests that have come to define high-level Overwatch. It’s not hard to find videos complaining about the current state of the game, begging for some sort of solution. A grassroots campaign encouraging Blizzard to #DeleteBrig began gaining traction in November last year.

Unfortunately, some people took it too far, directing their misplaced frustration at Brigitte’s voice actress, Matilda Smedius. Removing Brigitte from the game isn’t likely, but we’ll look at some other options Blizzard might use to fix the situation a little later on.

OWL pellets

Part of the appeal of the Overwatch League is the fact that the pros are playing more or less the same game you can purchase and play on your home PC or console. Theoretically, the only difference between a pro player and anyone watching from home is their ability to exploit the game’s mechanics.

Pros have superior reaction times, game sense, and conditioning, but they have the same toolset available to them as any kid who picked up a brand new copy yesterday. There’s an aspirational aspect, sort of like watching an NFL game, then throwing a football around in the backyard during halftime.

Unfortunately, this means any changes intended to affect how pros play also trickle down to everyone at every skill level. Blizzard’s in a tough spot here. GOATS isn’t especially fun to watch or play against, and that’s hurting not only OWL viewership, but the interest of the more casual playerbase who play the game at lower skill tiers. The problem here is that the game plays completely differently depending on the proficiency of the player and their teammates. YouTube channel Your Overwatch calls this the “Multiple Meta Problem,” and it’s been an issue ever since the game launched in 2016.

Take for example Reaper, the shadowy assassin with twin shotguns and a short-range teleport. His abilities make him particularly deadly against Tanks, so Blizzard increased his sustainability just before OWL’s stage one kicked off back in February. This didn’t have the desired effect. Reaper can’t shoot through barriers, and he’s most effective at close range, which is where GOATS wants to fight anyway. Some pro teams tried using him during the first week of OWL play, but it didn’t work at all. By week two it was all GOATS, all the time.

Meanwhile, more casual players on PC and consoles had to deal with Reaper’s increased lifesteal, which made him much harder to deal with at lower ranks. Reaper isn’t an especially hard hero to play, and players at the lowest skill levels found the character oppressively difficult to take down. The next patch reduced his lifesteal, though it’s still higher than it was before the Lunar New Year event began.

This is far from an isolated incident. More recent changes to Junkrat have increased the damage dealt by the mad bomber in an attempt to reduce the effectiveness of shields. While it may have the desired effect, it also makes him a terror at lower tiers of play, particularly where players don’t have the benefit of coordination. The game becomes a lot less fun when any single character becomes overpowered, but this is something Overwatch players have had to deal with since the very beginning.

Online comic Nerf Now!! illustrated one of the problems with this philosophy back in 2010. “Imba” is short for imbalanced, while “nerfing” something means to make it less effective.

The strangest changes seem to come when Blizzard attempts to balance around a specific circumstance. An adjustment made in March 2018 seemed to target New York Excelsior player Bang “Jjonak” Sung-hyeon specifically, leaving average players confused. The most recent set of changes seem to be designed to drastically reduce the effectiveness of teamwork, which is odd for a game so heavily focused on working together with your squad.

A big part of the problem is how Overwatch rolls its updates out. Instead of balancing the game frequently or iteratively, multiple changes are packaged into monthly patches which tend to either overcorrect or do little to change the current state of the metagame.

Even if something doesn’t work or a character becomes overpowered as a result of a patch, Blizzard won’t correct itself until the next major update goes live, waiting as long as eight weeks to make changes. The prevailing theory is that this is due to the console versions of Overwatch, which require payment to the platform holders (e.g., Sony and Microsoft) every time a patch is uploaded.

It’s even worse in the Overwatch League, as pro players don’t receive any balance changes whatsoever after one of their six-week stages has begun. This is intended to maintain parity between teams, ensuring that pros can practice with the same ruleset they play on for at least a month and a half. For good or ill, the version of Overwatch OWL viewers saw over the weekend won’t be changed again until mid-May at the earliest.

What’s especially galling is the short shrift given to the game’s Public Test Realm (PTR). This separate build of the game is only available to PC players, and is ostensibly intended to collect data about upcoming changes before they go into the live game. In reality, it seems as though Blizzard just uses the PTR to test stability before pushing through changes that would have been made regardless of player feedback. Unless something is truly game breaking (such as the damage exploit that preceded Symmetra’s second rework), it seems to go through to all three platforms without additional comment.

One suggestion that might help make for a healthier metagame would be to use the PTR more dynamically, and to use it for more than just bug testing. Other competitive games such as Fortnite and League of Legends update on a weekly basis, making changes as needed in the moment. It makes sense to freeze the pros on a specific set of rules, since it’s hard enough for them to deal with the shifting state of the metagame.

But for more casual players who just want to unwind after work, it would be much more fun if they didn’t have to deal with the overpowered character du jour for months.

Fix plz

A big part of Overwatch’s appeal is the ability to change your hero character at any time from the spawn room to deal with an enemy team’s composition. One of the major challenges for balancing Overwatch is making Tanks and Supports fun to play, since most players tend to gravitate towards the more diverse damage characters by default.

As a result, most of the Tank and Support classes are tuned to be a little more effective than damage characters, which has led to the current state of affairs. So how could Blizzard change things to keep this from happening again?

What can be done?

One suggestion which has gotten a lot of traction recently is a locked composition, restricting teams to contain two Supports, two Tanks, and two Damage characters. This is already how the majority of Overwatch’s players tend to build their compositions in the game’s casual and competitive modes.

It’s even gotten support from prominent OWL figures such as professional analyst MonteCristo. He points out that such a change would stabilize the state of competitive Overwatch, and allow individual experts to shine no matter what the game’s meta shapes up to be.

The benefit of a locked 2-2-2 composition is that it makes Overwatch much easier to balance since developers will have a rough idea of which characters will make up a team. Such a sweeping change to the game’s rules isn’t without precedent, either. At launch, you could have multiple copies of the same character on a team, an idea that was quickly changed before the game’s competitive mode went live. Anyone who played though games containing five Bastions and a Mercy will be happy to tell you the current state is a lot better.

The idea of a locked composition does raise some other issues though; Overwatch has many hybrid characters who don’t fit into a specific role cleanly. Determining which character should fall into which role could be sticky. Should Soldier: 76 be a Support since he has a healing ability? Should Brigitte and Mei be Tanks since they can disrupt enemies and block damage? Maybe Roadhog and Zenyatta should be Damage characters since they can put out so much firepower. Locking roles would require uncomfortable questions to be asked.

Another proposal (and one I think might be much more viable) is the idea of allowing a pick/ban phase before a match starts. This would work something like hero selection in games like League of Legends or Rainbow Six Siege, with each team choosing their composition after protecting certain desired heroes and allowing the opposing team to ban their use of a certain character. The Dallas Fuel’s assistant coach Jayne has done some experimenting with this idea, and it seems to result in much more diverse compositions.

This would have to require some additional rules, of course. Bans would need to be applied to both teams, and this might require a deeper bench since not every player can play every character. Banning more than one character from a certain class would probably be a bad idea as well, especially since the Tank and Support classes have far fewer options than the Damage category. Teams might need to substitute players after the bans came through, instead of swapping between maps as they do currently.

The main reason this hasn’t been considered so far seems to be the comparatively limited hero pool found in Overwatch compared to games like League. As Overwatch’s roster expands, it’s more likely that this idea will be given serious consideration by the game’s developers. One other possibility is that a pick/ban system has the potential to make Overwatch even more inscrutable to casual viewers; FIFA and the NFL don’t allow teams a say in determining the opposing lineup due to a bad matchup.

Between a hammer and a hard place

The Overwatch League is worth at least $500 million, but it requires both a healthy playerbase and an educated viewership to remain sustainable. Watching OWL requires viewers to have a fairly deep understanding of the game, otherwise it looks an awful lot like a bunch of plastic toys bouncing off one another.

 The team of “shoutcasters” who provide color commentary do their best to explain what’s happening during a given match, but it’s hard to explain in the moment nuances such as how the angle of Reinhardt’s shield determines whether McCree’s Flashbang will disorient him. To enjoy watching professional Overwatch, you pretty much have to have played Overwatch extensively.

What’s more, pro players tend to have fairly short careers in any sport, but it’s especially true in esports. Reaction times lengthen and senses dull as players age, and it’s not uncommon for players to retire in their mid-twenties. This means a professional league such as OWL will need a constant stream of new talent to replace players who retire.

OWL does a better job than most leagues of making sure its players are well-salaried for their skills. But for some players, the restrictions aren’t worth the compensation. Several  popular players like Brandon “Seagull” Lamed, Felix “XQC” Lengyel and Daniel “Dafran” Francesca have all retired from the Overwatch league to focus on their personal brands (though in Lengyel’s case, the parting was accelerated by multiple behavior problems offstage).

Ironically, other content creators have noted that Overwatch isn’t worth streaming for them since Blizzard is already showcasing the best players in the world, four nights a week.

GOATS has hurt the Overwatch League, and the game as a whole by making Overwatch less fun to watch and to play. Without a healthy playerbase, it may become increasingly difficult to find players skilled enough to compete in OWL. It’s a long-term problem, but this threatens the lucrative advertising and sponsorship deals made with partners like Toyota, T-Mobile, and HP, as well as the broadcast deals made with ESPN, Disney XD, and Twitch.

For now, Overwatch still has plenty of players. But gamers are fickle, and every new game that comes out siphons attention away from Blizzard’s shooter. Last month it was Apex Legends, this month it may be Days Gone or Mortal Kombat 11.

And if Overwatch’s meta has another stale period, it’s not hard to imagine people uninstalling the game and watching something else when given the opportunity.

Time to Rally

In the year since Brigitte released, Blizzard has added three additional characters to Overwatch, and we can see here how the company is attempting to avoid the mistakes they made with Reinhardt’s squire. Wrecking Ball, Ashe, and Baptiste are all characters who have a harder edge on them, much more difficult to play than Brigitte. Players with mechanical skill are able to gain far greater value than those who lack this aptitude. All three look great in the hands of a pro, but each is quite difficult for casual players to master.

Overwatch in general has become steadily more difficult to play since it released, with easier-to-play characters like Mercy, Symmetra and Torbjörn reworked into more skill-based heroes. This is great for players with the proficiency to adapt to these reworked heroes’ playstyle, but raises the bar of entry for those who can’t. The recent debate about accessibility in games applies here as well; at what point do casual players decide the game isn’t for them anymore?

Balancing Overwatch seems to be a Sisyphean task, and it’s literally impossible to please everyone. But it seems that by tuning the game for the most elite of players, Blizzard risks alienating fans and endangering the premier esports league they’ve done so much to build up over the last couple of years.

I’m hopeful that the developers will be able to find a happy medium before it’s too late.