Does warming up before ranked actually help?

Everyone has an opinion: "never queue cold", "deathmatch just tilts me", "one swiftplay and I'm good". We took the match histories of 175,000 players, found every first ranked game of the day, checked what they played right before it — and measured how that ranked game went.

Data: — players · — first ranked games of the day with a known result Regions: NA · EU · AP · KR · BR · LATAM Generated: —

TL;DR

Loading…

How we measured it

Riot's match history for a player lists every game they played in roughly the last 60 days, in every queue: competitive, unrated, swiftplay, deathmatch, team deathmatch, spike rush, custom games and the limited-time modes. That is enough to reconstruct a player's day.

A session starts when a player's previous game of any kind was more than 6 hours earlier — they slept, worked, or otherwise took a real break.
The target is the first competitive game of each session. If nothing was played before it in that session, the player queued ranked cold. If one or more non-competitive games came first, and the last of them started within 120 minutes of the ranked game, the player was warmed up.
We then fetched the result of every target game: win or loss, plus the player's kills, deaths and combat score in that game, and their rank at the time.
Warm-ups are grouped by what was played: aim modes (Deathmatch, Team Deathmatch), quick modes (Swiftplay, Spike Rush, Escalation, 2v2), Unrated, custom games, or a mix.

The headline: cold vs warmed up

Why compare the same player with themselves. People who warm up may simply be different kinds of players — more dedicated, more tilted, on a promo, whatever. The within-player estimate only uses players who have both kinds of first games and compares their cold days with their warmed-up days, which removes that difference between people.

Does more warm-up help more?

If warming up genuinely improved the first ranked game, more of it should help more, and a warm-up that ended minutes ago should help more than one that ended an hour ago.

Performance in that first game

Win rate is a ten-player outcome; a player's own kills, deaths and combat score in the target game are a more sensitive read on whether they were "warmed up".

Same player, warmed-up days vs cold days

The raw table above mixes different people. This compares each player's own first games: their combat score and kills per round on days they warmed up minus the same on days they queued cold.

Does the first game of the day go worse at all?

For a random subset of players we fetched every ranked game in their sessions, not just the first. This is the win rate by position in the session — the "rust" curve that warming up is supposed to fix.

How long since the last session

Cold and warmed-up first games, split by how long the break before the session was.

Across ranks and regions

Caveats

Players choose whether to warm up, and why they choose it matters. Someone who plays deathmatch because they feel off may be warming up on exactly the days they were going to lose. The within-player comparison removes differences between people, not differences between days.
Only games that appear in Riot's history count. Custom games and some limited-time modes are included; aim trainers, the Range, and other games (Aim Lab, Kovaak's) are invisible to us — a "cold" player may well have warmed up somewhere we can't see.
Session boundaries are a heuristic. A six-hour gap is our line between "same session" and "new day"; the results are not sensitive to that choice in the ranges we tested.
Ranked win rates hover near 50% by design. Matchmaking equalises teams, which is exactly why even a real warm-up effect on individual performance can show up only faintly in win rate.