I had a crazy friend tell me about this the other day. Claimed that Valorant wasn't actually sbmm and then sent me this white paper from UCLA to tell me it wasn't bs.
https://web.cs.ucla.edu/~yzsun/papers/WWW17Chen_EOMM
Thoughts?
I had a crazy friend tell me about this the other day. Claimed that Valorant wasn't actually sbmm and then sent me this white paper from UCLA to tell me it wasn't bs.
https://web.cs.ucla.edu/~yzsun/papers/WWW17Chen_EOMM
Thoughts?
I'm confused about what engagement-optimized matchmaking looks like in-game as opposed to skill-based matchmaking.
I understand it's the set of games for each player that makes them least likely to not play the game for a while afterward, but what do those games look like? Is it favoring matchups that are likely to result in a win for a player that has lost multiple games already?
Also, these findings were specifically for 1v1 matchmaking, and the authors wrote that they need to test if their findings hold up in team-based games.