Couldn't load analytics
Pending follows
โ
โ
โ
โ
โ
Cumulative P&L
Returns by sport
| Sport | Reads | Hit % | ROI | P&L | Brier | Log loss |
|---|
Where the model is correct
bar = miss vs claim, in points
Calibration by sport
| Sport | Reads | Model | Actual | Gap |
|---|
Couldn't load analytics
Pending follows
โ
โ
โ
โ
โ
Cumulative P&L
Returns by sport
| Sport | Reads | Hit % | ROI | P&L | Brier | Log loss |
|---|
Where the model is correct
bar = miss vs claim, in points
Calibration by sport
| Sport | Reads | Model | Actual | Gap |
|---|
Read this once to get the most out of the data.
Variance is real.
Not every read will land. Each one shows where the model's probability differs from the market price. That difference is the model's view, not a guaranteed advantage. Judge the record over hundreds of reads, never over one weekend.
Stake responsibly.
If you act on a read, keep it to 1 or 2% of your bankroll. Staking discipline matters as much as the model does.
Market odds adjust quickly.
Bookmakers reprice fast. The widest gaps between the model and the market tend to close soon after they appear, so a read is freshest the moment it surfaces.