TEAM BUILDER

How the recommendation is worked out

Each suggestion combines three things. The lane matchup is how the candidate has actually performed into the champion opposite them, rather than how it is reputed to go. The composition gap looks at what your four other picks already give you — physical and magic damage, engage, front line — and prefers a champion that supplies what is missing. The sample weight pushes down a champion whose record rests on very few games, so a 100% win rate over three matches does not outrank a 53% over nine hundred.

Where the numbers come from

Our own aggregation of ranked solo-queue matches, not a scrape of another site. Every win rate behind a suggestion is counted from games played, and the figure shown alongside a champion is the rate for that champion in that lane — which is frequently nothing like its overall rate. Aatrox is a useful example: 47.1% top and 55.1% jungle in the same sample.

What it will not tell you

It does not know your champion pool, your team’s communication, or whether you can actually play the pick it suggests — the best champion on paper is a bad pick if it is your first game on it. It has no view of professional play or of the patch that shipped this morning: an aggregation describes games already played, so a champion reworked yesterday will read as whatever it was before. And it ranks the pick, not the player; a counter advantage is a few percentage points, not a decided game.

The draft you build lives in this page’s address, so sending somebody the link sends them the draft.