← Premier League 2016-17 · Sat, Aug 20, 02:00 PM UTC · ref: Mike Dean, England
The betting picturemarket vs our model, market by market
“Market” = what bookmaker prices imply after removing their margin — the world's best guess. “Model” = what our machine-learning model said BEFORE this game, replayed honestly (it never saw the result). A bet is only interesting when the model sees something meaningfully more likely than the market does (the edge) — it's never about picking the winner.
Who wins? (90 minutes)
| Outcome | Market says | Model says | Best price | Verdict |
|---|
| Man City win | 62% | 62% | 1.58 | fair |
| Draw | 23% | 19% | 4.34 | fair |
| Stoke win | 16% | 19% | 6.29 | value: +18% |
across 1 bookmaker · probabilities de-vigged (bookmaker margin removed)
Timeline
from ESPN event data · goals, cards and shootout kicks
The match in numberswhat happened + what each side looked like walking in
Match stats
| 0 | Half-time score | 2 |
| 8 | Shots | 12 |
| 3 | On target | 6 |
| 7 | Corners | 7 |
| 12 | Fouls | 13 |
| 3 | Yellow cards | 4 |
| 0 | Red cards | 0 |
Pre-match form
What each side looked like walking in — computed only from games before this one.
Everything else we hold on this match · 1 more recorded fact — tap to expand