Which market actually profits: 271 bets broken down by type
What we are comparing
We have 271 settled bets across six European leagues. Below is the split by market: how many bets, the hit rate, the average price, the break-even rate and the resulting ROI.
One caveat first, without which the table misleads: the samples are wildly uneven. One market has 226 bets, another has six. That changes how much any of these numbers can be trusted.
The table
| Market | Bets | Hit rate | Avg odds | Break-even | ROI |
|---|---|---|---|---|---|
| Over goals | 19 | 84.2% | 1.71 | 58.5% | +45.4% |
| Double chance | 10 | 80.0% | 1.58 | 63.5% | +30.6% |
| Match winner | 8 | 62.5% | 1.86 | 53.9% | +13.8% |
| Under goals | 226 | 68.6% | 1.62 | 61.6% | +11.0% |
| Both teams to score | 6 | 50.0% | 1.94 | 51.5% | −16.7% |
How to read it
The first thing you notice is "over goals" at +45% ROI. That is exactly the number not to take seriously: 19 bets. One good run produces that by chance. Same for double chance (10 bets) and both teams to score (6 bets).
The only row with a meaningful sample is under goals: 226 bets, ROI +11.0%. That looks like a durable result rather than luck.
The break-even column shows why it matters. Under goals hits 68.6% against a 61.6% break-even — a seven point margin. Both teams to score hits 50% against 51.5%: the model did not even reach break-even, hence the loss.
Why so much "under"
226 of 271 bets — 83% of the portfolio in a single market. We did not pick it deliberately: the model is built on the Poisson distribution and works best with goal counts, and bookmakers misprice low totals more often.
That concentration has a downside worth stating plainly: the bets stop being independent. A high-scoring round hits the whole portfolio at once. An overall hit rate hides this — losses are not spread out, they arrive in clusters.
We are currently testing markets unrelated to goals — corners and cards. They run separately and stay out of the published record until the sample is large enough to mean something.
What follows from this
When choosing a source of predictions, ask not only about overall ROI but about the spread across markets. A one-market portfolio and a five-market portfolio carry very different risk at the same return.
Live figures by market and league are on the statistics page. Why a hit rate without the price means nothing: see the break-even rate.