How we score
Two grades. Two different questions.
Your in-game letter grade and your Vetta Score can disagree on the very same match — and both can be right. They’re measuring different things.
Riot’s letter grade
How this game ranks against everyone’s.
Riot grades your performance as a percentile of every game ever played on your champion and role — mostly from raw totals like CS, damage and KDA. An S means “statistically, a great game for this champ.” It can’t see that your team fought without you.
It answers: “Was this a strong game for this champion?”
The Vetta Score
Your weight in the game you actually played.
The Vetta Score measures your share of this match — your part of the team’s fights, damage, vision and objectives — against the ten players who were there. No percentiles, no rank, no strangers. The letter grade next to it is the piece that looks outward, weighing you against real peers when the data’s there to do it.
It answers: “What was your weight here — and what should you fix?”
What we evaluate
Four rules behind every score.
Your share of THIS game, not a leaderboard
The Vetta Score weighs you against the other nine players in your match — your slice of the team’s fights, damage, vision and objectives. In that number there’s no percentile, no rank, no comparison to strangers; only the letter grade ever looks past your lobby.
Role-aware by design
A support is never judged on damage; a tank gets credit for soaking it. We pay each role on the things it can actually move, so the verdict fits how you actually played.
Every claim cites evidence
Findings carry a real number and a confidence level. We never invent a percentile or a stat — when the data isn’t there, we stay quiet rather than guess.
The letter is a peer percentile — or nothing at all
When enough ranked players share your rank and role, the letter grades you against them — an A means the same thing for a support and a mid. What the numbers say leads, with Vetta’s findings only allowed to tilt it, never flip it, and a loss can never grade S. It’s anchored to the current patch, or the one just before it while the current patch is still filling — and we tell you which. When the sample is still too thin to anchor, there is no letter: “not enough data yet” is the honest answer, never a number we invented.
A worked example
When an S meets a B.
A common split: a top-laner with a beautiful stat line on the losing team. Both grades are telling the truth.
Riot
S
“A great stat line for the champion — but the fights happened without you, and you owned no vision. Riot rewards the farm and the damage; Vetta weighs your absence from the game your team lost.”
A low score next to a high rank isn’t a knock.
It’s the gap between “good for the champion” and “decisive in this game.” That gap is exactly where climbing happens — and it’s what we coach.