in

Rioter on surrendering in LoL. Why does fighting in a losing game usually not make sense?

Surrendering matches in LoL and the chances of winning after a failed vote.


Surrendering matches in League of Legends is one of the more controversial topics raised by fans of Riot’s production. Some people strongly prefer to surrender a game that promises to be really bad and just start another match where the chances of victory will be better, while others adhere to the principle of never surrender and playing every match until the Nexus is destroyed by an enemy team.

Riot has made changes to the surrender vote more than once, and update 13.7 includes another fix. Now, in the voting that could start in the 15th minute of the game, it will not take 5 yes votes to surrender a match, but only 4, as before in the voting that was available from the 20th minute of the match. The Rioter explains the context of the changes and gives the data that guided the team.

Why is it easier to surrender now?

Riot Phreak, who details the updates released to League of Legends, shared his thoughts on the upcoming surrender changes. He said that for him, as a person who usually votes no, it will reduce his win rate (assuming he will be on a losing team), because the outcome of a game is not always a foregone conclusion when it may seem that way.

He mentioned, however, that the data collected by the team shows that the overwhelming number of games in which the vote in the 15th minute was unsuccessful, because only 4 players voted yes, ultimately ended in a losing game anyway.

I don’t remember the exact numbers, but 4/1 voting in ranked matches had about 97% for losing a match. This was a surprisingly high number, and even as someone who would normally be against this change, I am able to accept it.

His additional commentary appeared online, in which he elaborated on the topic he raised in the video. He mentioned that the exact number of games won after a failed vote for surrender is 3.3%.

In the case of unsuccessful votes:

  •  If a game ends before 21 minutes, ~90% are held hostage (45% get FF 20’d, 50% end naturally by 20 [open mid or just lose])

  • 80% of the games resolve by 25 minutes

  • Games that make it to 30 minutes have a 10% winrate

  • Games that made it to 40 minutes had a 45% winrate (but the incidence was very low)

While the “come back from behind, win at 45 minutes” moments are some of the highest moments in League, the vast vast majority of the cases are experiences where the majority of the team doesn’t want to play the game out and are trapped in a suboptimal experience.

Add on to that that a significant portion of our playerbase is paying by the hour to play. Having said that, there’s clearly a player behavior portion of this and we’re still coming at this from multiple angles (also addressing afk, etc.)

The Rioter also mentioned that many games that end in early victory after a failed vote for surrender are teams where someone started the vote for a joke. The fraction of victories after a failed vote may therefore be somewhat misrepresented because of this type of action.