Every figure shows its source. Where a source is missing, the number is a placeholder — and it is never written as fact.

By GamGem ResearchLast updated 20 Aug 2026 · 14:32

Patch 26.16 — what moved in Mid

20 moved up, 15 moved down, and 29 did not move by half a point either way.

Moved up

  • Kai'SaC · 49.0%+1.00 pt
  • Miss FortuneA · 47.0%+0.90 pt
  • MalzaharA · 49.4%+0.90 pt
  • EzrealB · 46.6%+0.90 pt
  • IvernB · 51.8%+0.90 pt
  • OriannaB · 54.5%+0.90 pt
  • ZoeS · 53.0%+0.80 pt
  • HeimerdingerB · 48.2%+0.80 pt

Moved down

  • SorakaB · 48.7%-0.80 pt
  • ViktorA · 54.3%-0.80 pt
  • MilioB · 53.1%-0.80 pt
  • Renata GlascA · 54.1%-0.70 pt
  • EkkoB · 51.8%-0.70 pt
  • KassadinC · 47.3%-0.70 pt
  • TaliyahS · 49.3%-0.60 pt
  • ZyraB · 48.5%-0.60 pt

Where it landed — by class

  • Fighter+0.245
  • Tank+0.152
  • Marksman+0.148
  • Support+0.0922
  • Mage+0.0856
  • Assassin-0.0517

Where it landed — by lane

  • Top+0.1371
  • Bot+0.1344
  • Mid+0.0795
  • Support+0.0664
  • Jungle+0.0570

Patch timeline

Riot’s notes

We hold the rates, not the balance log. What actually changed each patch is in Riot’s own notes.

  1. 26.16Live now-0.3
  2. 26.14Sample win rate across this patch+0.5
  3. 26.11Sample win rate across this patch-0.7
  4. 26.08Sample win rate across this patch-0.8
  5. 26.05Sample win rate across this patch-0.4

What this page measures

Every figure here is how far a champion’s win rate moved in our sample across patch 26.16 — not what the publisher changed. A champion can move because its own numbers changed, because an item it buys did, or because the champions around it did. This page cannot tell those apart, and does not pretend to.

The balance notes themselves are the publisher’s and stay on the publisher’s own site, which is also the only place they are current.

Methodology

Data window
512,340 games on patch 26.16, Ranked Solo/Duo, NA. Last refreshed 20 Aug 2026.
Rank aggregation
Emerald+ and above, pooled. Splitting further leaves samples too small to separate a real effect from noise.
Role assignment
A champion is filed under the lanes it actually plays, not the one class it belongs to — which is why several appear in two.
Counters
A matchup, not a tier gap: how this champion does into that one specifically, over that pairing’s own games.
Minimum sample
The confidence indicator beside every rate is driven by how much of the window this champion appears in. A low reading means read the figure loosely, not that it is wrong.
Why a build is not a cause
The best-scoring build is the one the best-placed players happened to take. Correlation is what a rate can show; it is not evidence that buying those items caused the wins.
Where the numbers come from
Names, classes, lanes, abilities, items and runes are Riot Data Dragon. Win, pick and ban rates are derived samples — Riot publishes none of them, and nothing on this page presents one as measured.