Why football prediction sites are free
Published
Ask why a site gives away football predictions and the answers you find are written by people with a stake in the answer — mostly affiliate pages ranking each other. The question deserves a straight response, because the funding model behind a set of predictions tells you more about what they are optimised for than any accuracy claim on the homepage.
There are four ways this industry pays for itself. Each one rewards a different behaviour, and none of them is inherently dishonest.
1. Display advertising
The oldest model, and the one behind most of the very large free prediction sites. Predictions are the product that brings traffic; the traffic is sold to advertisers. Gambling is a high-CPM category, so a site with millions of monthly visits can run a real business on ad revenue alone.
What it rewards: page views and return visits. Coverage of hundreds of leagues, a prediction for every fixture, and a reason to come back tomorrow. Depth per match matters less than breadth across matches.
What to watch for: the incentive is to publish a confident-looking call on everything, including fixtures where the honest answer is that the model has very little idea. It also rewards ad density, which is why these sites tend to be the least pleasant to use.
2. Bookmaker affiliate commission
The site links you to a bookmaker and earns a commission when you sign up and deposit — often a share of your net losses for as long as you keep an account, sometimes a one-off fee.
What it rewards: clicks that convert into new betting accounts. Predictions become the reason to be on the page; the sign-up offer is the actual product.
What to watch for: the interests here are worth stating plainly. Under a revenue-share arrangement, the site earns more when the people it referred lose more. That does not mean the predictions are bad — plenty of affiliate sites publish reasonable analysis — but it does mean nobody in the chain is paid for your bets going well. Look for whether the commercial relationship is disclosed at all.
3. Subscriptions
Free tips on the surface, the “real” picks behind a paywall — or, in the better version of this model, free predictions with paid depth: more leagues, more markets, historical data, an API.
What it rewards: a visible gap between the free and paid product. Something has to be held back for the paid tier to make sense.
What to watch for: how the gap is constructed. Charging for depth and convenience — more data, better tools, more coverage — is a straightforward trade. Charging for “access to the predictions that actually win” is a different proposition entirely, because it implies the free ones are knowingly worse, and invites the question of how the split was decided.
4. Nothing yet
Some sites are not monetised at all, either because they are early, or because they are somebody’s project. This is our position: FootInsights runs no advertising, has no bookmaker affiliate links, and sells nothing. Not out of virtue — it is a deliberate constraint of the current phase, and it will change eventually.
What it rewards: nothing, which is its own kind of information. When it changes, the honest thing is to say so on the page rather than let you find out from a banner.
What this means when you read a prediction
The useful question is not “is this site free?” but “what does this site get if I follow this tip?” Three practical consequences:
A prediction on every fixture is a business decision, not a modelling one. Any honest model has fixtures it barely understands — newly promoted teams with no top-flight record, early-season matches, competitions with thin data. A site that never says “we don’t have a strong view here” is telling you something about its funding, not its confidence.
Free is not the opposite of good. Some of the most rigorous football modelling in public is free, and some of the most expensive tips are worthless. Price signals nothing about accuracy in this market, in either direction.
The record is the only thing that survives the incentive question. A published, complete, pre-committed record is hard to fake regardless of who is paying for the site, which is why it is the thing worth demanding — and why we wrote a checklist for verifying one and a 30-day audit you can run yourself on any site including this one.
The claims that should stop you
Independent of funding model, some things are simply not true of football, and seeing them is enough on its own:
- Guaranteed or “100% sure” wins. Football’s outcome distribution does not permit this. The honest ceiling is far lower than the marketing in this industry suggests.
- Fixed-match access. Either a lie, or a criminal offence being advertised to you.
- A win rate with no denominator. “87% accurate” across what sample, which market, which leagues, and measured when? Without those, the number is a shape rather than a claim.
- A results page that only grows on good weeks. Records that can be edited after the fact are marketing, not evidence.
Where we stand
We publish probabilities rather than picks, freeze every prediction before kick-off with its model version attached, and grade all of them — hits and misses — on a public track record scored with the Brier score rather than a headline win rate. The methodology page sets out what the model does, what data it uses and where it fails, including its own backtest against the closing-odds market.
That is what our funding model currently allows us to optimise for. Judge it on the record, not on the price — and whatever you read, anywhere, stake only what you can afford to lose.