Football Intel

How It Works

How we find statistical edges against bookmaker odds.

The Model

We use a Dixon-Coles Poisson model to predict match outcomes. The model analyses historical team performance, home/away splits, goal-scoring patterns, and recent form to estimate the true probability of each market outcome.

1

Team Stats

Historical performance data

2

Model

Poisson probability estimation

3

Compare

Model vs bookmaker prices

4

Edge

Publish when edge is 8%+

The Edge

When our model disagrees with the bookmaker by 8% or more, that's a signal. We only publish picks where we have genuine statistical edge — not opinions, not hunches, not “value” based on gut feel.

Model says

67%

Bookmaker implies

51%

=

Edge

+16%

The Markets

We focus on markets where statistical models have the biggest advantage over bookmakers.

Goals Over/Under

Active

O1.5, O2.5, O3.5, U1.5, U2.5, U3.5

Both Teams to Score

Active

BTTS Yes / No

Shots on Target

Active

Team shot totals O/U

Corners

Active

Team corner totals O/U

Cards

Active

Team and match card totals O/U

The Data

18

Leagues

42

Odds Markets

2hr

Update Cycle

Auto

Settlement

Real bookmaker prices captured every 2 hours. Picks settled automatically when matches finish.

Why Not Just Bet on Everything?

Our backtest shows blind betting on Over/Under 2.5 goals loses 4-7% to the bookmaker margin. The bookmaker is right most of the time. But in 15-20% of matches, the data says the bookmaker has mispriced the market. That's where the value is — and that's what we find.

Ready to see today's picks?