Over/under goals in betting

A bet on whether total goals in a match finish above or below a set line, most commonly 2.5. Team identity is irrelevant — only the goal count matters.

Goals arrive roughly according to a Poisson process, so once you estimate expected goals for the fixture you can compute the whole distribution. An expected total of 2.7 goals gives Over 2.5 around 54%; an expected total of 2.2 gives around 43%. Half-goal lines cannot be pushed, whole-goal lines can, and quarter-goal Asian totals split the stake in the same way as Asian handicaps.

P(k goals) = (λ^k x e^-λ) / k!, where λ is expected total goals.

Expected goals for a match is 3.1. Poisson gives P(0 goals) 4.5%, P(1) 14%, P(2) 21.6%, so P(under 2.5) ≈ 40.1% and P(over 2.5) ≈ 59.9%, a fair price of 1.67. If the book is 1.80 on Over, that is roughly a 7.5% edge — assuming the 3.1 estimate is honest.

Totals are the cleanest bridge between statistical modelling and betting. You can build a credible estimate from xG data alone, without predicting who wins.

NPFL matches average lower totals than most European leagues, so Under lines carry more value than the acca-culture default of stacking Over 1.5 and Over 2.5 legs would suggest. Check league base rates before importing habits.

• Applying a Premier League goal average to a defensive or low-scoring domestic league.
• Ignoring weather, pitch condition and dead-rubber motivation in end-of-season fixtures.
• Confusing Over 2.5 with Over 2.0, which refunds on exactly two goals.