Fight Winrate in OWL Will be a Metric for Both Teams and Players

November 28th
written by Karahol

OW_Doomfist_004 copy copy

New teams in Overwatch League, larger rosters and a different format will create a new environment of competition and we will need to rethink stats too. 

One such stat is the Fight Winrate, which is a stat that only Winston’s Lab manages to track as of now, due to the site having introduced a way to break down maps played in matches by fights. The stat is pretty straightforward, is calculated as a percentage and describes how many fights in total a team has won over a period of time, be it a round, a map, a match or multiple tournaments. The stat scales accordingly.

Up until this point, due to the rosters not fielding usually more than 6 players, all players on a 6-man team had the same fight winrate during a tournament, because they all played the entire time. Instances where this wasn’t the case include the APEX tournaments, where teams could field rosters with more than 6 players; Contenders S1, mostly because teams could make roster changes mid-tournament, although they were given from the start the right to have larger rosters and most recently in APAC Premier 2017 (see image below for Runaway players). In the Overwatch League this will be the trend, however, for all teams. It’s still unclear how substitutions will be utilized and how many players Blizzard will allow teams to swap between maps, but whatever the case might be, the fragmentation of players’ stats will increase, compared to before.

apac 2017 runaway

So, here are my thoughts on how this metric could help us identify patterns in terms of how teams will arrange for their player swaps. Before I begin, though, let me make clear that a good amount of games must be played first to provide us enough data to look at.

Judging by how Team Envyus swapped their players in Contenders S1 NA playoffs and with the knowledge obtained by the coach of FNRGFE from his appearance on the Overrated podcast, that Blizzard wanted all player swaps known beforehand, one can tell that most teams will rotate players on a map basis. I imagine that each team will have some specialists, depending on their roster of course, that they will field on certain maps for certain tactics against certain opponents. Looking at players’ fight winrate and making comparisons for players of the same role (Main Tank, for example) on that map over the course of games will give us an idea who is better than the other and who should probably be preferred, especially for the tough games.

The tricky part for this metric will be when more than one players are swapped for a map. Ideally, one would want to see how well the team does when they have player A on for role A and when they have player B on for the same role. But we won’t get ideal situations, unless somehow we get our hands on scrim data, and thus one will have to factor in the level of the opponents the two players will have faced, as well, before drawing conclusions.

One more interesting aspect of this metric will be revealed when combined with the overall rating of the players’ other stats (like deaths, kills, ultimates charged etc). Having better fight winrate but worse individual stats could also indicate that the player is not as clutch as others. I think it could really help paint the narratives of underdogs, or of players who really elevate their teams’ performances once on the battlefield. And I deem that those narratives will be as important as they are in traditional sports, in case of illnesses, injuries or unforeseen causes that could bench a player. In the past, due to teams not being built with this mindset and just fielding the bare minimum of players required, 6, it was about scrapping the bad players and getting new. Now, if a player has stunning performances and he’s not available for a match, his absence will be felt a lot more. It kinda reminds me of the time when WhoRU was benched due to misbehaving, Lunatic-Hai had their traditional loss in the quarterfinals of the tournament and fans were panicking.