Every match page on nova11 shows four scores and, sometimes, a recommendation. Most of the time there is no recommendation at all — and that is the part worth explaining, because it is deliberate.
Each fixture is scored from 0 to 100 on four goal markets: Over 0.5 goals in the first half, Over 1.5, Over 2.5, and both teams to score. The scores come from recent form — each team's last matches overall and at the relevant venue, friendlies excluded, 90-minute results only.
A score is not a probability. It is an internal index. An 80 on one market and an 80 on another do not mean the same thing, which is exactly why we do not simply recommend whichever number is highest.
Some markets land often no matter what anyone predicts. Over 1.5 goals happens in roughly 78% of matches; Over 2.5 in roughly 54%. Being right about Over 1.5 most of the time is therefore worth very little — you would be right most of the time by guessing.
So a market is only recommended when its score points to an outcome meaningfully more often than that market's own baseline, by enough of a margin to be worth anything. On most fixtures, nothing clears that bar, and the honest answer is to say nothing. We would rather publish fewer calls than pad the record with ones we do not believe in.
Because the baselines differ. A 74 on Over 2.5 can carry more information than an 88 on Over 1.5, since Over 2.5 starts from a far lower baseline and the score has more room to say something. Comparing the raw numbers across markets would be comparing different units.
We measured it. Across several hundred settled matches, our Over 1.5 score showed effectively no relationship to whether the market landed — matches scoring high did not land meaningfully more often than matches scoring low. The score is still displayed, because it is information; it is simply not something we will call.
A competition's track record counts only the markets we actually called. A market we declined to call is not graded, and never appears as a hit or a miss — grading a call we never made would flatter or damage the record for no reason. Where fewer than ten matches have settled, we show "early data" instead of a win rate, because a percentage from a handful of matches is noise.
Nothing on these pages is betting advice, and we deliberately do not publish target prices or stake sizes. But a few general principles matter far more than any individual prediction:
Stake the same amount every time — a small, fixed percentage of a bankroll you have set aside for this and nothing else. Varying your stake by how strong a match feels requires a confidence estimate; we removed ours from this site precisely because we could not support it.
Singles, not accumulators. A bookmaker takes its margin on every selection, so combining picks pays that cost repeatedly while making the outcome far more volatile. Multiples lower your expected return and raise your risk at the same time.
The price matters as much as the pick. A well-judged call at a poor price is still a losing bet. If the odds available do not compensate you for the risk, the correct action is to not bet at all.
Judge over months, not weekends. Losing runs of fifteen or twenty are ordinary at any realistic strike rate, and a good week proves nothing. Expect that variance, and never increase your stake to recover a loss.
These are statistical estimates, not predictions of what will happen, and not betting advice. Our own measurements carry wide error margins — the sample is still small, and we say so on the pages rather than hiding it. Nothing here is a guarantee, and no result is certain.
18+. If you choose to gamble, only ever risk what you can afford to lose.