Finding the optimal solution to draws in Overwatch is no easy task and Blizzard has tried many different things, but hasn’t found the answer yet.
Subsequently, every time in a tournament two teams tie on a map, everyone is waiting to see what kind of solution will be provided that time and how it will affect the overall map score. Let’s go over some of the solutions so far, especially in the very recent months, and see their pros and cons.
During Contenders S0 group stage Blizzard implemented a system that awarded points for map wins and draws, 3 for a win and 1 for a draw. That way there was no need for a tiebreaker and match continued with the next map. However, 4 Game Sets work only in group stages and when you go into playoffs, you can’t award points. Another solution is needed then and, everyone will agree with me on this, that it’s bad to have two different solutions for the very same thing within a tournament format. There were also other issues with the point awarding system, that lowered the overall impact of map wins it felt unfair.
Moving on to Contenders S1, Blizzard decided to drop the point awarding system and judge a teams’ performances on the basis of match wins, with map wins to be used as a standings tiebreaker. When two teams were to draw on a map, they would both get a point, like if they had both won the map; the team that had chosen the tied map would get to pick the next map and if the overall map score was tied in the end, a tiebreaker Bo3 Control map (Ilios, in that scenario) was to be played to decide who gets the match win. Again, this initially made sense in the Round Robin group stage with match format being “Best of 4 Game Set” (weird, I know, but that’s that what it was). It is still not perfect, though. If a team was already down 0-2 in map score, a tie would end them and they would lose the match.
The arrangement of the game modes suddenly becomes very crucial, since most draws occur on 2CP maps, even now that Blizzard has somehow limited them. Do we want players and viewers to know that they have lost the match while in the 3rd of the 4 maps to be played, because of they way draws are handled? Some would argue that it is legitimate, because a team needs to win more maps than their opponents. For some it still doesn’t feel right. And the biggest issue is raised when we go into the playoffs and the Bo5 format. If two teams get 2 wins each and have a tie on Assault, they will have to play another map to solve the tie, negating the very essence of the Bo5 format, which is to decide the winner by the end of the 5th map. It didn’t happen but the possibility is there and it is ridiculous to even think about it.
Other tournament organizers like APEX and even the APAC ones have tried other solutions. OGN deals with the issue by enforcing a Sudden Death approach. To solve the tie, the two teams will play a Bo1 Control submap (randomly selected or know beforehand, doesn’t matter) and the best team on that map will take the win. This approach obviously favors the better team on the specific submap and on Control, in general. APAC organizers, on the other hand, have come up with some pretty interesting ideas: replaying the whole map or playing another map of the same game mode to solve the tie. The downside to this approach is that it takes more time for the match to conclude, affecting the broadcast overall time, and there is no guarantee that the two teams will not end up playing 2-3 maps more. Suddenly the Bo5 becomes a Bo8 or something, which is very funny as a concept. Nonetheless, with the latter’s solutions competitive integrity is not compromised at all and the best team on the most maps will get the match win.
Competitive integrity should be at the center of the optimal solution. Second in line thing to consider is the match format. If teams are to play 4 Game Sets or Bo5s, or even Bo7s, tied maps should not exist in any capacity. There should be a clear winner of each and every map, without the teams having to play a different game mode to solve the tie in any way, unless this is the last possible solution.
My proposal is pretty simple and easily applicable. If teams draw on a map, you give them extra rounds with a time bank of around 3 minutes. Tests must be carried out to find the sweet spot, but overall you give teams enough time to prove who is the better of the two. In other words, if two teams reach a draw on Hanamura, then they will play one extra round each with 3 minutes in the time bank. Whoever makes the most progress on the map, wins the map. For these extra rounds, the true overtime rounds, some rules of the game modes will not apply. For example, teams will not be given extra 30 seconds if they cap point A of Assault maps, neither they will need to cap 1/3 of the circle for their progress to matter; every single percent will be crucial. If the teams tie again, then they are given another 3 minutes. If somehow they manage to reach a draw three times in a row, then they will play a Bo1 on a Control submap that both agree on, selected beforehand.
Only drawback of the solution I suggest is that if two teams go the full distance, all the way to the Control submap, a lot more time has been wasted on the broadcast. Not more than replaying the full map and tying on it, but it’s still at least 18 mins of pure gameplay + the duration of the Control map. Including the canon duration of the map, it will take two teams around 40 mins to decide the winner. But this will be a rare sight. I can hardly imagine two teams requiring more than 2 overtime sets to decide who is the best on an Assault or Hybrid map.