Has Blizzard figured out the optimal match format?

August 08th
written by Karahol


Finding the optimal match format in any esport is not an easy task, especially during its first years, where developers and community are on their honey moon and everything looks shiny and on roses. However, the time of realization comes sooner or later and for professional play, choices need to be made that promote a challenging but at the same time, solid environment.

This time is now for Overwatch.


The challenges of Bo3, Bo5 & Bo7 

Tournaments for Overwatch started early, from its beta phase, and a great deal of them featured Bo3 matches. They bypassed quite easily the decision to drop a game mode or anything of the sort by using pick/ban or veto map drafts that allowed teams to decide the maps (and consequently the gamemodes) they wanted to play. They could even play the same game mode for all three maps if they wanted. The same format could be used for Bo5s and Bo7s, although some tournaments from the get-go (like Overwatch Atlantic Showdown) effectively limited the map pool down to 7-8 maps. This decision was dropped for a short while, but it seems like it’s favored by Blizzard and pro players alike because it resurfaced lately.


But then APEX Season 1 came about, where matches had a Bo5 format (except for finals) with pre-arranged game modes and limited map pool, and challenged this notion we had that teams were the ones to decide which game modes are to be played and which maps. At the same time, this different match format showed a weakness of the pick/ban map drafts, that teams avoided playing certain game modes and therefore they were far behind South Korean teams.


So, the question raised was: Should all game modes be represented in the matches? And if they do, the game modes are 4, which one do you drop in a Bo3 and which ones you double for Bo5/Bo7? Especially for Bo7 things get even more interesting, because three game modes must be doubled and the choice can easily favor one over the other.

This question is challenging also due to the fact that apart from decisions made on the basis of game mode representation, one should keep in mind that there are other limitations deriving from the broadcasting schedule. A Bo3 game can be too short, but a Bo5 with pick/ban map draft can drag too long if there are 2 (or more even) Bo5 Control maps in it.


The “4 Game Set” format

Blizzard has shown their intention for all four game modes currently in-game to be represented in the professional scene of the game and the solution they came up with was the 4 Game Set.

In this match format, both teams are required to play all 4 game modes once, despite if one of the teams manages to reach a 3-0 score. Although it counts as a win (something we learned in Contenders S0), the teams are required to play the 4th game mode as well for tiebreaker purposes and points. The game modes are pre-arranged again and people have witnessed this format in action in Take Over 2, Overwatch Contenders Season 0 and currently in the Overwatch World Cup matches. From some very fast calculations, the 4 Game Set matches have a low variance in terms of broadcasting time and now that Control maps are played as Bo3, their in-game time is maximum 1 hour, which is considered good.


What one should notice, however, is that this match format is only applicable in Group Phases, where there are no elimination brackets and the extra maps played can act as points tiebreakers in the case that two teams have the same wins, losses, ties etc. For the playoffs a different match format was needed that ensures that there are no ties. This match format is a Bo5 made up from the 4 Game Set format with the addition of another Control, where teams can’t tie on and it’s used as a tiebreaker (the Control is played as a Bo3).

In APEX, Hybrid is doubled, whereas for the Blizzard format, Control is doubled. Which game mode should be doubled, though, and why?


Double trouble

Doubling a game mode doesn’t create the problem by itself. It’s specifically when you have to choose only one game mode to double and therefore the decision weights a lot more. If we had 3 eligible game modes for professional play, it would be much easier to double 2 of them and show what the character of the game is while leaving the third at one map to add variety and change the pace.

In Overwatch, the problem is further enhanced because Control, despite being the most unique game mode, is probably the one the community loves the most since it requires the least synergy and individual actions are rewarded more. On the other hand, Assault, which requires the most synergy and has the most strategical depth (despite the stalling flaws, admittedly), isn’t as enjoyable for the average viewer/player.

If we were to remove Control and double Escort and Hybrid maps while keeping Assault at one map, the game would revolve too much around payloads and I am not entirely convinced whether it should at this point. I personally enjoy Assault maps, despite the stalling problems, because this game mode greatly differentiates between the teams with top synergy and teams with good to great synergy, so I would be in favor of doubling Assault maps and Escort while keeping Hybrid, the fusion of the two, at one map for the Bo5.

Because Blizzard, however, has no intention of removing a game mode, then, it’s only reasonable that the game mode doubled should be the one that requires the most practice and I believe that this is Control due to the amount of submaps a teams has to practice. Also, it makes sure that any kind of score tie is solved in the last map of the canonical duration of the match without the need for a map tiebreaker.


The Final(s) Decision

In the end, a decision must be made as to whether Overwatch needs a Bo7 in the match format rotation for Grand Finals or finals, as a whole, because the leap from Bo5 to Bo7 is quite a big one. It’s definitely not a mere addition of 2 more maps.


Let’s examine again the Bo7 formats from APEX and Contenders. OGN keeps Control in one map per series and doubles all other gamemodes with the following arrangement: Control – Hybrid – Assault – Escort – Assault – Escort – Hybrid. They could have repeated the sequence (Hybrid – Assault – Escort) but for some reason they decided that Hybrid has to be at the end. Blizzard, on the other hand, is just adding another two maps to their Bo5 format, which are Escort and Hybrid. So, the final arrangement is: Control – Escort – Assault – Hybrid – Control – Escort – Hybrid. Blizzard decided to avoid doubling Assault and they doubled everything else.


So, in APEX’s case, a team has to go through 2 Assault maps (with all their problems) to succeed, whereas in Blizzard’s case, they have to play a second Control map. Again, this is a problem as to how one choice was made over the other. Was it due to the general sentiment in the West that Assault is problematic and Blizzard decided to not double it? Do they think Control is a better game mode overall? It’s still unclear. The only thing certain is that Bo7 creates another problem because of the need to double even more gamemodes, or to triple if one wants to go that route, which is a possibility.


My optimal format

In conclusion, the format I have in mind as the best one -and probably one that many will dislike- excludes Control maps, which will be used only as tiebreakers. I prefer to see the focus revolving around Assault and Escort maps, which I would have doubled for a Bo5, and Hybrid being the connection point. For a BO5, I would have: Escort – Assault – Hybrid – Assault – Escort. With this format, a team must play all three game modes once before they can secure the win and since it’s almost impossible to tie on Escorts (some really weird things must happen to reach a tie there), there will be a clear victor in the end. Ties on Assault can be solved either by playing Bo1 Control, which teams will know beforehand and they will practice this one and only specific submap for all cases, or I would also experiment with giving teams extra rounds with 4 minutes so as to solve ties inside the game mode and not on a completely different one.

In the case where someone would prefer double Hybrid over Escort, keep in mind that it is possible to tie on Hybrid and more tiebreakers are needed then. That’s the reason why Blizzard prefers to double Control than Hybrid, I guess. When there is the possibility for more tiebreakers, it makes the match duration a bit more inconsistent and it affects broadcasting.

All in all, I think Blizzard is doing some good work with the tournament formats, taking steps to solve tournament issues that have previously arisen and making sure that there will be a solid basis to build upon the competition for the Overwatch League. I, for one, welcome these formats they are promoting, despite my objections for Control.

– Karahol