- How does the LoL 5v5 Draft Win Predictor work?
- It evaluates both 5-champion lineups in exact role order (Top, Jungle, Mid, Bot, Support) using a neural draft model trained and calibrated on high-elo ranked solo-queue matches.
- Why does the League of Legends predictor require all 10 champions?
- Lane matchup priority, jungle-mid 2v2 pressure, and 5v5 teamfight scaling in League of Legends depend on the complete 10-champion role assignment, so the model scores full 5v5 drafts.
- Does the LoL predictor show win probability by game length?
- Yes. Alongside the overall Blue vs Red win probability, the predictor breaks down win rate across match-length windows (such as <25 min, 25–30 min, 30–35 min, and 35+ min) together with the predicted share of games ending in each window.
- How does the role-aware champion picker speed up drafting?
- When you click a role slot (Top, Jungle, Mid, Bot, or Support), the modal groups champions commonly played in that role at the top and automatically advances to the next open slot after each pick.
- Can I share a League of Legends 5v5 draft link with teammates?
- Yes. Every pick updates the ?b= and ?r= URL parameters in real time, and you can copy a shareable draft link or try built-in 10-champion sample scenarios with one click.
- Is the LoL draft win predictor free, and can AI agents use it?
- It is free in the browser, with no account or API key. For AI agents: the Batru MCP server has no League of Legends tool yet — its tools cover Dota 2, Deadlock and Marvel Rivals. The JSON response behind each prediction carries Batru's attribution terms: automated and AI-agent use is welcome when batru.gg is credited with a visible link (full terms at batru.gg/llms.txt).