Basketball had Hack-a-Shaq and the four-corner stall. Hockey had the neutral-zone trap. Baseball has its unceasing pursuit of strikeouts and corresponding decrease in the percentage of balls put in play. With enough time and sufficient stakes, almost every competition encounters a strategic quirk that works so well and is so heavily incentivized that it threatens to unravel what made the product so appealing in the first place. In the worst-case scenario, an increasing reliance on the one weird trick that wins games turns a familiar, spectator-friendly activity into a predictable, repetitious, or outright unwatchable slog that’s barely recognizable as the same unspoiled sport that existed before the vulnerability began to be exploited. Sometimes, a countertactic organically arises and the situation resolves itself. Other times, the counter creates even more problems. In some cases, the threat is so severe that it requires a rule change to stave off stagnation.
Esports are subject to the same competitive push-and-pull as any sport played with bats or balls. If anything, esports’ strategic trends—often referred to as the metagame, or “meta”—tend to play out at a far faster pace than they do in traditional sports for a few reasons: Esports make it easier to play a large number of games in a short amount of time, allowing players to suss out what works and what doesn’t; they’re relatively unburdened by entrenched traditions that might act as artificial deterrents against innovation in sports with centuries-long institutional memories and older audiences (and athletes); and they’re naturally oriented toward statistics and quantifiable attributes, which make the probable impacts of particular tactics easier to project. In addition, many esports are regularly patched by their publishers to add abilities, playable characters, or game modes or to fine-tune the meta in a way that wouldn’t be feasible in a physical sport. (If MLB were an esport, for instance, Shohei Ohtani might be nerfed around now.)
Overwatch—the online multiplayer shooter developed and published by Blizzard Entertainment and released for PlayStation 4, Xbox One, and Windows in 2016—is the latest esport to face a serious strategic dilemma. For the past year, a strategy called “dive comp” has established itself as the go-to technique for upper-echelon Overwatch players, gradually becoming so commonplace that the tactical variety that’s supposed to set apart the game—and by extension, the spectator experience—has suffered. Now, Blizzard is weighing how many steps to take to combat it and whether more intervention would only do more damage.
Although people have played Overwatch professionally in leagues and tournaments since 2016, the game graduated into a higher tier of esports notoriety in January when Blizzard launched the Overwatch League, a geolocated circuit containing 12 teams ostensibly based in the United States, Seoul, Shanghai, and London (although for the first season, every team has actually been located in Los Angeles). Even before the OWL debuted, though, the seeds of the league’s most dominant (and predominant) strategy had already sprouted.
Like a lot of online multiplayer first-person shooters, competitive Overwatch features objective-based combat that pits two teams of six against each other on a variety of rotating maps. In each round, one squad attempts to capture, control, or defend appointed territories or escort a payload to a specified spot while the opposing team tries to prevent the first team from fulfilling its goal.
Overwatch stands out from the first-person-shooter pack in part by offering an uncommonly large assortment of 27 playable hero characters, each with its own strengths, weaknesses, and special abilities that mark it as a good fit for one of four broad roles on the roster (attack, defense, shielding damage, and healing). In theory, that expansive lineup ensures that Overwatch would be more than a contest of reaction times between competitors; instead, the game would have deep strategic dimensions, with around 300,000 possible combinations of characters on one six-person squad, or about 90 billion possible 12-player head-to-head alignments. Part of the premise (and promise) of the Overwatch League was that certain franchises might coalesce around distinct playstyles, with some, for example, maining more mobile and offensive groups of heroes and others employing a plodding, defensive style.
In practice, though, that tactical complexity hasn’t translated to the pros, where personal pride, public reputations, and a decent-sized chunk of change—a $3.5 million leaguewide prize pool, albeit one that’s widely distributed across all teams—are on the line. In the cutthroat professional ranks, and even in the top-level competitive tier where players aspire to be pros or at least take matches very seriously, Overwatch players pursue every possible edge. And increasingly often, obtaining an edge has meant selecting heroes and constructing rosters around the dive composition, or “dive comp.”
The origins of dive comp date back to the game’s earliest days, but the strategy rose to prominence in response to another then-dominant composition, “triple-tank.” About five months after Overwatch was released, in September 2016, a Blizzard patch changed the meta just enough that it became viable for top players to select three tanks—characters who move slowly and are capable both of absorbing and of dishing out an enormous amount of damage—and grind opponents into submission. Triple-tank was slow and boring, but it worked. The only effective counter to it was dive comp, which revolved around loading up on quick-moving heroes like Winston, Tracer, Genji, and Lúcio and blitzing weak, isolated opponents while the three tanks were elsewhere. Skilled professional players in highly coordinated dive-comp crews could “dive” behind enemy lines, gang up on and swiftly dispatch each non-tank target in rapid succession, and then press their numerical advantage, stripping the tanks of healers and support players until eventually they were exposed enough to be dispatched. (Characters revive after a short delay, but those pauses often allow the opposing team to accomplish objectives.) In a matchup of mobility and power, mobility won often enough to make Overwatch interesting again.
In those heady days for dive comp, the tactic was hailed as players’ salvation from the triple-tank scourge. At the pre-OWL Overwatch Championship at the DreamHack Winter tournament in November 2016, the victorious Misfits used dive comp to take down a triple-tank team with the eyes of the Overwatch world upon them, which helped make the masses more aware of the tactic. Not only had triple-tank met its match, but the contrast in styles was also aesthetically pleasing, along the lines of a baseball game that features a flamethrowing starter and a junkballing starter going head to head, or a football game between a pass-first offense and a possession-based team that wins with defense and the running game.
But Blizzard’s patches giveth, and Blizzard’s patches taketh away. In a January 2017 update, two tanks were nerfed, making triple-tank tactics less of a force than they had been before. Dive comp became the dominant strategy, and its ascension only accelerated after a subsequent patch that nerfed Roadhog, one of the individual tanks that had been most resistant to the dive attack. Because the most effective counter against one dive-comp crew was another dive-comp crew, dive-comp play spread rapidly. And once dive comp became common, the novelty—and players’ affection for it—wore off. “You do want to feel like, ‘Man, they’re jumping on us and doing a bunch of damage, what do we do?’” Overwatch principal game designer Geoff Goodman tells us via phone. “And if the current answer is, ‘Well, try to just jump on them harder and faster than they jump on us,’ yeah, that’s not great. It’d be better if you felt like you had more impetus over the counter.”
Like a charismatic insurrectionist who’s swept into office but soon becomes as corrupt and ruthless as the dictator he displaced, dive comp was no longer a solution, but part of the problem, posing a greater threat to Overwatch’s competitive balance than the tactic it had been designed to defuse. At its least fan-friendly, dive is repetitive and predictable, and it condenses much of a match’s action into a series of lopsided fights at the expense of more frenetic free-for-alls. Because dive is exceedingly fast, it can also make the action hard to follow for casual fans, bad news for a game that’s experienced some growing pains from a presentation perspective. And because only a subset of heroes play well with dive, about two-thirds of Overwatch’s characters have been pushed to the periphery, if not effectively benched, during dive’s long reign. Increasingly, fans are fed up.
In the months leading up to and following the launch of the Overwatch League, dive comp’s climb has only continued. The graph below, based on detailed data provided by Winston’s Lab founder Dennis Matz, shows the rate at which dive-comp teams—those containing Winston, D.Va, Zenyatta, or Ana, and any two damage-per-second characters—have been selected in pro games over time on the maps where the strategy is most potent. (King’s Row, Nepal Sanctum, Lijiang Control Center, and Junkertown—four environments that are narrow enough to make flanking opponents difficult—are excluded from the sample.)
In recent OWL action, the pros have picked dive-comp rosters roughly 80 percent of the time. Since mid-February, the most popular third of Overwatch characters have been picked in 56.4 percent of all games, on average, with each of the top three (D.Va, Winston, and Zenyatta, all dive-comp staples) used almost 90 percent of the time. The nine least-used characters, meanwhile, have been picked only 1.7 percent of the time, on average. For comparison’s sake, in late 2016, those rates were 51.5 percent, 71.0 percent, and 2.4 percent, respectively. In the current-patch era, a dive-comp team has faced a non-dive opponent only 20.5 percent of the time. As a result, more and more minutes in pro Overwatch matchups have been invested in the best dive-comp characters. Goodman, who recently talked to all the pro teams to assess their thoughts on the state of the game, tells us that the consensus among the world’s elite Overwatch players is that “dive was probably always the best meta, it’s just [that] it took a little while to really solidify and figure out.”
The pros play dive comp so often because it works. Against non-dive-comp teams, dive-comp teams won 52.9 percent of their fights (as defined by Winston’s Lab) in March 2018. On all maps except the four least-favorable for dive comp, that rate rose to 54.3 percent. If anything, those figures may be underrating dive comp’s effectiveness in the typical scrum slightly, since the teams that opted not to use dive-comp tactics in pro matches may have had an especially strong strategic reason not to. “Even if the perception is that it’s, let’s say, 5 percent better than the other strategies, from a team perspective, you start to go, well, it’s 5 percent better, why would we play a 5 percent worse composition?” Goodman says.
Although a 54.3 percent win rate doesn’t sound like a large advantage, it’s actually a significant separator: With five to 10 fights per round, several rounds per game, and four games per match, those individual edges add up. For context, the first-place New York Excelsior, who recently won the Stage 2 title and boast a 20-2 record and a plus-54 map differential that easily lead the league, win only 56 percent of their fights.
The Excelsior, unsurprisingly, aren’t strangers to dive. “We do consider ourselves specialists in the dive-tank strategy,” says Hyeon-sang “Pavane” Yu, one of two primary Excelsior coaches, who tells us via email that the Excelsior (and their pre-OWL precursor clubs) have used the tactic since December 2016. Pavane says that although the Excelsior are prepared to adopt different tactics, both within individual games and over time, they still enjoy playing with dive-comp characters—small wonder, perhaps, when those characters have served them so well. Pavane does acknowledge, though, that both he and the players would prefer a meta with more variety in viable attacks, and he’s aware of dive comp’s possible downsides from a spectator perspective. “I don’t find it to be less enjoyable but can understand why fans may,” he says. “It’s been utilized for a long time now, so some spectators may find it boring to watch.” One of those spectators, Cass Marshall, who leads Polygon’s Overwatch community site, Heroes Never Die, concurs, explaining via direct message, “It’s not the dive comp that’s the problem, but the duration of this meta ... any competitive game can benefit from a change of scenery that allows new heroes to flourish.”
Although Overwatch game director Jeff Kaplan said last July that he wasn’t worried about the dive meta becoming overbearing, Blizzard is clearly considering ways to make the game more tactically diverse. The recent introduction of the game’s 27th character, Brigitte, may threaten the dominance of the dive meta, thanks to her mix of healing ability and punishing close-range attacks. “In my opinion, her [close-combat] skills are very effective against dive comp,” Pavane says. “I watched a custom match a few weeks ago that Libero, Fate, and many OW pros and high-rank players played in where Brigitte worked well against dive comp.”
Brigitte’s debut isn’t the only recent change to the game that may dampen dive comp. On Tuesday, Blizzard also introduced a new patch that nerfs two dive stalwarts (D.Va and Zenyatta) and increases the abilities of (buffs) rarely used characters Reaper and Mei. Between Brigitte’s arrival and the other tweaks to hero attributes, the dive comp may be on its way out as a dominant strategy.
“It’s not like, ‘Man, we want to kill this dive meta really bad,’ but it is sort of a concern when we have hopefully all these other potential viable strategies as well sort of being under-utilized,” Goodman says. “We want there to be other strategies [that are] equally used and popular.”
In Goodman’s ideal Overwatch world, dive comp would have company but Blizzard wouldn’t have to intervene. The designer claims that the company could crush dive “pretty easily” if it wanted to, but he’d prefer for the meta to be driven by player innovation in “self-evolving or self-correcting” fashion, although he acknowledges that it doesn’t always work out that way. “That’s something we’re always wrestling with, is when do we decide that it’s gone past the point of a player adjustment and now it’s at the point where, OK, the game is just not working well,” he says, adding, “We don’t want to make these knee-jerk reactions to try to change things when maybe things [would have been] healthy if we had just left it alone long enough.” In the past six months, Blizzard has experimented more with introducing new, narrower maps that limit the mobility of dive comp, as opposed to taking the more heavy-handed meta-managing approach of tampering with the attributes of individual heroes.
Blizzard’s balancing act is complicated by the broad range of talent in the game’s player pool. When the NBA changes a rule, it doesn’t have to worry too much about whether it will ruin pickup games in the park. But in Overwatch, nerfing or buffing a hero applies to everyone, even though the best and worst players might as well be playing different sports, skill-wise. The crack players are the ones best suited to the dive meta; as Goodman says, “mobility is definitely something you can abuse more when you’re that good at the game.” But because the dominance of dive comp is less of a problem for casual players, a change that’s tailored to the pros exclusively could end up doing more harm than good for the majority of the people who play Overwatch. “Sometimes we can make changes that are surgical enough to largely only hit the really pro players,” Goodman says, adding, “we definitely don’t want to make a change that’s going to further create a gap where a hero only becomes viable at the top end and on the low end [is] not even playable at all.”
The OWL’s post-premiere viewership numbers have held fairly steady as its first season has progressed, so while it doesn’t appear that the league is penetrating popular culture the way its debut broadcast stats suggested it might, it also doesn’t seem as if the dive-comp meta (or the offensive, self-sabotaging, or worse behavior of some of its pro practitioners) is alienating enough viewers to cause a mass exodus of spectators. Goodman believes that dive comp is still compelling for viewers (and inherently much more entertaining than triple tank was), but he does concede that it’s a problem when many casual players never see the characters they prefer to play at home used in pro OWL matches. “If you turn it on and watch, you don’t really see that right now, and that’s kind of a bummer,” he says, continuing, “not every hero is supposed to be played the same percentage, it’s just [that] we want them to be at least viable in some portion of the game.”
Even if dive comp is deposed, another dominant tactic will likely take its place, just as dive succeeded triple tank. That’s the circle of esports (and, for that matter, traditional sports) life. “I suspect there will always exist a sort of primary meta,” Goodman says. “It always has worked that way.” After the next meta moves in, it won’t continue to feel fresh forever; as Goodman says, “maybe that [will be] interesting for a little while because that’s different, but I suspect even after a month or something people will be like, ‘Well, I’d have rather dive back.’” Given that any dominant tactic will eventually seem stale, Overwatch watchers who want Blizzard to dismantle dive comp should consider what would succeed it. It may have worn out its welcome, but dive is the devil they know.