Forebet (forebet.com) is one of the most visited football prediction sites, and it works differently from a tipster site. Nobody at Forebet sits down and picks a winner. Every match goes through the same statistical model, and what comes out is a set of probabilities. That makes Forebet useful in a specific way and useless in another, and the difference is worth understanding before you put money on one of its predictions.
What Forebet is
Forebet publishes free football predictions for a very large number of leagues. Its own app listing (published under the developer name Fore Creations) describes it as a system for predicting football matches ”on the basis of statistics using complex mathematical algorithms”, says the first statistical model went online in 2009, and says the model works on a database of football data going back to the year 2000. According to the same listing, it covers most football leagues around the world and calculates probabilities ”for the final outcome of the match, goals scored, etc.” The service is available both on the website and as a mobile app.
Forebet does not publish the formulas behind the model. What we can say is what kind of model it describes itself as: historical match data in, probabilities out, the same treatment for every fixture.
What a Forebet prediction looks like
The key output is a probability for each possible result, not a single tip. For a given match you get a percentage for a home win, a draw and an away win, and figures related to goals. Imagine a mid-table home side against a weaker visitor: the model might give something like 50 % home, 27 % draw and 23 % away. Forebet’s headline ”prediction” is simply the outcome with the highest number. But 50 % also means the model expects the home side to fail to win half the time.
That is why the probabilities are more useful than the pick. A pick tells you what is most likely. A probability, set against the odds, tells you whether it is worth betting on.
How to use the numbers: compare them with the odds
Every decimal odd has an implied probability: 1 divided by the odds. If the bookmaker offers 2.30 on the home win, the implied chance is 43.5 %. If Forebet says 50 %, the model thinks the home side is underpriced. If the bookmaker offers 1.80 (55.6 %), the model thinks the opposite, even though both cases show the home side as Forebet’s ”prediction”.
A gap between model and market is a question, not an answer. Bookmakers price in team news, injuries and heavy money from sharp bettors within hours. A model built on historical results reacts more slowly. So before you trust a gap, ask what the market might know that the model does not: a key striker suspended, a coach sacked this week, a side resting players before a cup tie. If you find a reason, the model is probably just behind. If you do not, the gap may be genuine value.
If converting odds to probabilities is new to you, our beginner’s guide to odds and betting (in Swedish) goes through it step by step.
What to check yourself before a bet
A model is only as good as what it measures. These are the football numbers worth checking alongside a Forebet probability, and what they actually tell you:
- Points per game (PPG): points won divided by matches played, ideally split into home and away. A side with 2.1 PPG at home and 0.9 away is two different teams depending on venue.
- Goal difference over recent matches: a team winning narrowly every week is more fragile than its points suggest.
- Expected goals (xG): a measure of the quality of chances created and allowed, available on many stats sites. A team that wins while giving up more xG than it creates is often due a correction.
- Team news and schedule: suspensions, injuries to key players and fixture congestion. This is where models are weakest and where your own reading adds the most.
Many punters who use Forebet’s numbers combine them with human tipsters who follow specific leagues closely. If that is what you are after, the important thing is that the tipster’s record is logged with odds before kick-off.
Looking for the best tipsters?
Strengths and limits
Strengths. It is free, it covers far more leagues than any human team could, and the method is consistent: the same model for every match, with no favourite teams and no bad days. The probability format is honest about uncertainty in a way a single ”tip” is not.
Limits. The method is not published, so you cannot see what the model weighs. It relies on historical data, so late news reaches it slowly. Coverage of lower leagues is a strength for breadth but also where data is thinnest. And we found no full, odds-linked public record of Forebet’s predictions, so claims about its accuracy, whether from Forebet fans or critics, are hard to check. Any ”Forebet is X % accurate” figure you see quoted elsewhere is worth treating with caution unless it says how many matches, which markets and at what odds.
How accurate is Forebet, honestly?
Accuracy in the sense of ”how often is the top pick right” is the wrong question. The model’s most likely outcome will often be a favourite, and favourites win often, so the hit rate can look good while betting every pick still loses money once the odds are counted. The questions that matter are whether its probabilities are calibrated (do outcomes it gives 60 % actually happen about 60 % of the time?) and whether betting the gaps between model and odds produces a positive return. The same trap applies to correct-score sites such as PredictZ.
You can test the second one yourself. Choose a league, note every match where Forebet’s probability is clearly above the implied probability of the best available odds, record the odds and a flat stake, and do this for a few hundred bets. Also note the closing odds just before kick-off. If your prices tend to beat the closing line, the gaps are finding something real. If they do not, the model is mostly lagging the market.
Betting from Sweden?
To act on a prediction from Sweden you need a Swedish-licensed sportsbook. Several of them use BankID as the whole login, with no separate account to create; our list of sportsbooks without registration shows which. If you want to back a model’s view without the draw in play, an Asian handicap line is often the cleaner bet; our guide to Asian handicap (in Swedish) explains the lines.
Frequently asked questions
How does Forebet make its predictions?
According to its own app listing, with mathematical algorithms working on football data collected since 2000. The exact model is not published.
Is Forebet free?
The predictions on the website are free to view, and there is also a mobile app.
Can you make money betting Forebet’s predictions?
Not by backing every top pick blindly; that ignores the odds. The only way to find out whether the probabilities give you an edge is to bet where they differ from the market and track the results over a large sample.
What does the Forebet percentage mean?
It is the model’s estimated chance of that outcome. Compare it with 1 divided by the odds: if the model’s number is higher, the model thinks the odds are too generous.
18+ · Spela ansvarsfullt · Stödlinjen · Spelpaus