Predictions

LoL match predictions, with receipts.

We do not chase clicks by predicting the popular team. Each prediction exposes its inputs: draft priority in the current patch, gold diff at 15 across the last 20 games, average game length, blue-side advantage for the specific matchup, and schedule fatigue. If the model disagrees with the market by more than 5 percent, we say so.

Predictions are a forecast, not financial advice. The historical hit rate is tracked and published at the end of each split.

How our predictions work

  1. Gold diff at 15. Net gold at the 15-minute mark over the last 20 maps, patch-weighted. Correlates with win rate better than kills or tower diff.
  2. Draft priority. We simulate 10,000 possible drafts from the current patch meta and each team's champion pools. Resolves to a per-matchup win probability.
  3. Side advantage. Blue-side win rate in 2026 varies by patch between 51 and 54 percent. We apply the current patch number, not the split average.
  4. Roster continuity. New signings, academy callups, and role swaps cost an adjustment in the first three weeks.
  5. Model vs market. We compare our number to a no-vig estimate of the closing line. Edges under 3 percent get ignored because vig eats that.

Latest predictions

What you will not find here