Measured, not announced

Did the balance changes work?

Every brawler an update changed, and what its adjusted win rate did in the 14 days afterwards against the 14 before. From sampled Ranked battles.

This shows what each number did, not what the update caused. A brawler can be buffed and ignored, nerfed and still climb, or moved by the map rotation turning over in the same fortnight. Sample sizes are printed next to every row so you can weigh them yourself, and anything under 500 decided battles on either side is left out rather than guessed at.

Updates that cannot be measured

Listed rather than hidden — an absence with a reason is a finding too.

  • The two windows were sampled at very different rates — 72,027 decided Ranked battles before against 401,768 after, a factor of 6. Any difference would be partly the sampling and not the update, so nothing is scored.

    It changed 41 brawlers. 8-bit, Amber, Ash, Bea, Bo, Bolt, Brock, Buster and others.

  • Too little was sampled either side of this update to measure it: 35 decided Ranked battles before and 83 after.

    It changed 57 brawlers. 8-bit, Amber, Barley, Bibi, Bolt, Brock, Bull, Buster and others.