What's the deal with Maps in Overwatch?
What’s up guys and gals, CaptainPlanet here with something a little bit different. Lately, pros have been pissed. Or at least, that’s what all of these reports have sprung up to announce. Here’s why: recently MLG announced their tournament ruleset for their Vegas-based Overwatch tournament, featuring eight of the best teams from across the NA region. With this announcement came a pre-determined map pool for each round, bucking the competitive community’s preferred system of map drafts and bans. This map pool also included an odd ratio of potential maps to be played – maps that are disliked by pros like Hanamura and Ilios leading the count with five and nine possible matches respectively. Pros and community members alike have voiced their dissatisfaction, but I wanted to give a little bit of history to their frustration. What follows will be a brief history of maps in professional Overwatch, a breakdown of MLG’s pool vs. historical draft-based pools, and suggestions for compromise.
Maps and Professional Overwatch
Like it or not, maps are an integral part to Overwatch competition at its highest level. Even in early closed beta, tournaments utilized and strategies revolved around map ban systems. The use of map ban/pick systems was in fact so entrenched that GosuGamers was forced to create their own third party application to support their weeklies and decrease the load on their admins, owdraft.com. I myself even had to use this application when competing at Overwatch’s first LAN tournament – EsportsArena’s Agent’s Rising. It is safe to say that giving teams the choice to pick and ban maps was an essential meta strategy that added another layer – and another mode of control – to their tournament success.
These early tournaments – the GosuGamers weeklies, opening week LANs, and the like – operated with relatively little Blizzard influence. As the budding competitive scene (and it’s prize pools) grew however, so did Blizzard’s design philosophies begin to inform how they were run. The Battle for the Atlantic – Overwatch’s first six-figure tournament – initially tried to mirror Overwatch’s Ranked Play rules of the time. When it was first announced, the ESL-run tournament included 0 hero limit and no stopwatch scoring on payload maps in a time before Timebank had been implemented but Coinflip still settled 3-3 ties. In the LAN finals, teams’ control of their map choices was also drastically reduced: the first map was picked by admins without player input, then the losing team was free to pick into whatever map they wanted without the worry of bans. The pro community vocally and vehemently decried these rules and ESL was quick to change their tune on the hero limit and stopwatch scoring decisions.
Since the Battle for the Atlantic, different tournaments have had different levels of player control over map pools. Most recently, Dreamhack Winter utilized a pick-ban system that was lauded by the teams that played there. Apex had a pseudo-predetermined pool, where one of each map type had to be played before duplicate types entered the pool. Blizzard’s own Overwatch World Cup had an entirely predetermined pool and served as an interesting case study for how predetermined pools can potentially influence the outcome of tournaments.
I believe most viewers would agree that South Korea was likely to stomp the rest of the world in the Cup regardless of their opponent or map, but they actually had a bit of an easier time getting to the finals than their competitors. The maps they played can be seen above, but let’s look at the possible maps they might have played, if they had actually lost a game:
-Group Stages: Temple of Anubis, Eichenwalde, Lijiang Tower
-Quarterfinals (vs Team USA): Temple of Anubis, Eichenwalde, Lijiang Tower
-Semifinals (vs. Team Sweden): Route 66, Eichenwalde, Lijiang Tower
-Finals (vs. Team Russia): Temple of Anubis, King’s Row, Dorado, Lijiang Tower, Hanamura, Eichenwalde
That’s a lot of Temple of Anubis, Lijiang Tower, and Eichenwalde. Once again, South Korea probably would have won anyway, but getting to play the exact same maps that they played and practiced during the group stages throughout the playoffs was lucky to say the least. Team USA in particular was not a fan of the situation and Ster made his distaste known on social media:
This brings us to MLG Vegas which (currently) sports an entirely predetermined map pool. Can the same “same-pool” situation happen with MLG’s map groupings? Judging from their maps per round page, it thankfully does not appear so. However, there is a different oddity at play this time – one that many writers before me have jumped on. What is the deal with how many Ilios, 2CP maps, and “less popular” escort maps in this potential pool? Not including tiebreakers, the potential pool of maps stands as follows:
-Ilios: 9 games
-Temple of Anubis: 5 games
-Route 66: 5 games
-King’s Row: 4 games
-Eichenwalde: 4 games
-Hanamura: 3 games
-Watchpoint: Gibraltar: 3 games
-Hollywood: 2 games
-Dorado: 2 games
-Numbani: 2 games
-Lijiang Tower: 2 games
-Nepal: 2 games
If we consider tiebreakers, things seem even more unbalanced because Ilios is the chosen tiebreaker map for group stages. Someone at Activision/MLG/Blizzard must really love Greece! Jokes aside, these numbers mean nothing without something to compare them to. But, what is there to compare to? Blizzard does not display wins, losses, or games played on their playoverwatch.com player profiles, so sites like Overbuff cannot make their usual nice summary data breakdowns. Luckily for you all, I have all sorts of map data from many of the Overwatch tournaments played throughout Season 2. Below, I have tallied up all of my “maps played” data in tournaments that had un-restricted pools and that had some form of pick/ban system. I then compared the map representation when players are allowed to draft to the proposed pool for MLG Vegas. The results are interesting to say the least:
And if you’d prefer a prettier version:
The Ilios explosion is staggering, but what really surprised me the most was the decimation of Hollywood’s share of the pool. Perhaps there are secret Blizzard metrics at play here that show that Hollywood takes up too much time and would cause scheduling problems for the MLG tournament. Maybe these metrics show the same for Nepal and Lijiang Tower, but not Ilios. Whatever the reason, there are demonstrable differences between the map pool at MLG Vegas and the pool that emerges when pros have control of the system.
Why does this matter? Why are the pros upset? Shouldn’t the pros know all the maps?? The answers to these questions are rooted in opinion, but I will attempt to explain these concerns from the pro’s point of view.
Map Picks and Bans add to strategy, complexity, and preparation for tournaments
In a tournament with a pick-ban system, matchups can be won and lost in the draft. Most Overwatch teams have “home maps” that they are most comfortable with and “away maps” that they despise. When preparing for big tournaments teams will develop map banning strategies to ensure they get their favorite maps and they deny their opponents theirs. This escalates into crazy mind-games during the drafting process as you might expect, as teams play chicken with leaving undesirable maps on the table and forcing their opponents into wasting their precious bans on them.
The outcomes of these drafts have an added bonus of granting additional talking points to the tournament talent who can use their knowledge of teams’ preferences to make better predictions of the match’s outcomes or to give context about potential lineup choices. For example, imagine a draft outcome where Complexity forced FaZe into a Watchpoint: Gibraltar opening match – a map Complexity is known for favoring. On the analyst desk, Flame could mention that it is interesting that Complexity continues to favor Watchpoint, despite losing two of their last three matches played recently.
He could then also note that FaZe is no slouch when it comes to Watchpoint either and that perhaps this drafting strategy was misinformed.
The error in thinking that pros should just “know all of the maps”
Let’s get this straight: the pros already know all the maps. They likely know the maps better than anyone even reading this article, probably even better than 99.9% of the player base. That is exactly the issue when it comes to having such a deep map pool and no way to control it: the pros have to play against their own peers. When preparing for tournaments, teams will hope to have multiple different strategies for every part of every map. Finding an edge on their opponents requires intense focus and time practiced on a select few maps because these are the best of the best that Overwatch has to offer trying to beat another team just as skilled working just as hard. All of those surprise flank routes, those quirky lineups, the exciting super-mega-secret-Bastion strategies – these are possible when teams are allowed to focus on their own personal map pool and given the freedom to draft into it.
The pros are upset because instead of this freedom, they are now staring at a fixed map pool that includes every single map in Overwatch less than two weeks before the event. They are now under an enormous amount of strain, having suddenly been forced to game plan for every possible scenario. In the past, this game plan has often resulted in one-size-fits-all cookie-cutter compositions that are just “decent” at every map, instead of wild, crazy, hype-inducing and map-specific lineups. They can look at the pool and decide that they should focus their bootcamping on Ilios – a map they never even consider under normal circumstances. But over this practice would float a melancholy haze because this meant their concerns about the fairness and map balance of Ilios – specifically the location and impact of its many death pits – fell upon deaf ears. As a final note, let’s not forget that reworked Symmetra is also releasing only four days before the tournament, completely altering the landscape of both the hero and map meta.
Now that I have successfully transformed my molehill into a mountain, let’s talk about what kind of compromises could be made. For example, while Ilios’ death pits are widely hated by the pro community, multi-kill Lucio boops are extremely hype-inducing for viewers – a feeling MLG’s tournament organizers likely wish to capture. I propose a re-balancing of the various king of the hill maps, taking some Ilios from the pool and giving it to Lijiang Tower and Nepal. Remember, there are still plenty of boop locations on Lijiang Garden and Night Market, as well as Nepal Sanctum. If Blizzard is concerned about the time it takes to draft and broadcast downtime, let’s allow teams to perform their drafts off-screen and only display the results. The casters and analysts can still use the draft results to fill dead air with discussion and the players will get to regain control over their destiny. If Blizzard is still steadfast on pre-selecting the pool, let’s at least make it only partially selected and have a draft/ban element to it. Perhaps only the first map could be pre-selected and the rest of the remaining pool determined via draft with loser picking next map. Maybe the group stages would be played on set maps, and the semifinals and finals have a draft and ban. And so on.
These are just a couple of ideas I have either heard around the scene or come up with myself, but what do you think? Should Blizzard allow players to pre-select their entire map pool? Should they do a hybrid selection/draft system? Should allow map drafts throughout the tournament? Let me know on Twitter, or with this strawpoll
That’s it for me! Until next time,