The Value Predictor track record shows how the rules behind our free Value Predictor would have done on matches that have already been played. It’s a backtest, so it uses only the results each team had before each match. The table updates itself several times a day as new results come in.
We publish it because a filter is only useful if you know how often it lands. That number also tells you the fair price for a typical pick: 1 divided by the hit rate.
Read the transcript
Here's how to use our football goals predictor with the odds.
The predictor checks each team's recent games against a simple rule. For over two and a half goals, both teams need seventy per cent of their recent games over two and a half, and an average of three goals or more.
A match that qualifies isn't a tip. It tells you recent form supports goals.
Next, think about the real chance. Say you judge over two and a half goals at sixty per cent. The fair price is one divided by zero point six. That's one point six seven.
Now check the best price. At one point eight, it beats your fair price. At one point five five, it doesn't, however good the form looks.
Form is a starting point. Price decides whether it's value. Please bet responsibly.
The short version
- The table shows how often matches that passed each rule actually landed.
- The fair price is 1 divided by the hit rate. A pick is value when the best odds are bigger than that.
- Compare the hit rate with “All matches”. If the two are close, the rule isn’t adding much on its own.
The Value Predictor track record so far
| Market | Rule | Picks | Hit rate | All matches |
|---|---|---|---|---|
| Over 2.5 goals | Both teams: 70%+ of last 5 over 2.5, avg 3.0+ goals | 132 (84 landed) | 63.6% | 57.1% of 1,824 |
| Under 2.5 goals | Both teams: 70%+ of last 5 under 2.5, avg 2.0 goals or fewer | 34 (16 landed) | 47.1% | 42.9% of 1,824 |
| Both teams to score | Both teams: 70%+ of last 5 with both sides scoring | 154 (96 landed) | 62.3% | 57.3% of 1,824 |
| Match result | Form side won 60%+ of last 5, 2.0+ points per game; picked to win | 800 (426 landed) | 53.3% | 47.6% of 1,824 |
Picks are matches that met the rule using form from before kick-off. "All matches" is the hit rate across every match with enough form to judge. Match result picks are the side with the stronger recent form, since past prices aren't stored.
Matches from 21 Oct 2025 to 19 Oct 2026 in the competitions the Value Predictor covers. Updated 1 second ago. Past results don't predict future ones.
Each row uses the same default settings as the Value Predictor page. For each match, we take the home team’s last 5 home games and the away team’s last 5 away games in the same competition. Next, we check them against the rule for that market. Finally, we record whether the pick then landed: over or under 2.5 goals, both teams scoring, or the stronger-form side winning.
How to read the track record
First, look at the hit rate. It’s the share of qualifying matches that landed, for example 3 or more goals for over 2.5. A 75% hit rate means about 3 in 4 qualifiers landed.
Next, turn the hit rate into a fair price. If the hit rate is 75%, the fair price is 1 ÷ 0.75 = 1.33. So you’d want odds bigger than 1.33 for a typical pick to be value.
Then compare the hit rate with the “All matches” column. That’s how often the market landed in every match we could test. If qualifiers land much more often than all matches, the rule is picking out something real. However, if the numbers are close, form alone isn’t telling you much, and the price matters even more.
Worked example
Say over 2.5 goals qualifiers landed 64% of the time. The fair price is 1 ÷ 0.64 = 1.56. If the best price on a qualifier is 1.70, that’s above the fair price. In contrast, a price of 1.45 is below it, so on this record it isn’t value, however good the form looks.
Why a qualifier still needs a good price
Bookmakers see the same form we do. As a result, matches that pass the rule often come with short prices already. A high hit rate isn’t enough on its own; the odds have to be bigger than the fair price. The Value Predictor now shows a fair price from the betting exchanges for every pick, and a green Value tag when the best UK odds beat it. Our guide to expected value betting shows how to check this, and the over 2.5 goals guide explains how to price the market yourself.
Limits of this Value Predictor track record
This is a backtest, not a record of bets placed. It’s useful, but it has limits you should know about:
- Sample size. A few hundred matches can still be swayed by luck. See our guide to betting variance.
- No prices. We don’t have the historical odds for every match, so the table shows hit rates rather than returns.
- Default settings only. If you change the number of games or the filters on the Value Predictor, results will differ.
- Football changes. Team news, managers and tactics move on, so past results don’t predict future ones.
In short, treat the track record as a guide to how reliable each rule has been, and always check the price. The Value Predictor gives statistics, not tips.
Related guides
- Value betting explained: the UK guide
- Betting markets: what the Value Predictor covers and why
- Implied probability: what betting odds really say
- Closing line value: the best test of your bets
For more on how results are judged over time, see the Wikipedia entry on backtesting.
18+ only. Betting should be fun, and every bet can lose. Set a budget you can afford to lose and stick to it. For free, confidential support, visit BeGambleAware.org or call the National Gambling Helpline on 0808 8020 133.