Free bets promotions are everywhere in sports betting. But how much are they actually worth, and what is the best way to use them?
Free bets typically have a “stake not returned” format. If you have a €10 free bet and place it at decimal odds of 2.00, you don’t get €20 back if it wins. You just get the €10 profit. That detail turns out to matter a lot for how you should use them.
If you only care about expected profit, the answer is simple: use your free bet on the biggest longshot you can find. The lower the probability of winning, the bigger the potential profit, and because you haven’t put up the stake yourself, the expected value of the free bet rises as the win probability falls.
But those bets come with a lot of risk and you only get so many free bets. The most likely outcome from the longshot strategy is that you win nothing. Many people would rather turn a free bet into something close to guaranteed money. You can do this by placing the free bet on one outcome and then using your own money to bet on the opposite outcome with another bookmaker. This is often called matched betting.
Here things get more interesting. Very short odds are bad because the free bet does not return the stake, so there is little profit to hedge. But very long odds are also bad because you need to place a large real-money hedge on the other side, and the bookmaker’s margin on that hedge starts eating heavily into your guaranteed return.
Using a simple model in which bookmakers earn the same expected margin on each side of a market, I show that there is a precise sweet spot. If the bookmaker’s margin is around 4.5%, the guaranteed profit is maximised when the free bet has about a 21% chance of winning. A free bet of €10 can then be converted into about €6 of guaranteed profit.
I then test this using more than 25,000 betting quotes from FanDuel and DraftKings on over 6,000 ATP and WTA tennis matches. Tennis is particularly useful because there are only two possible outcomes (which makes hedging easy) and matches with a roughly one-in-five underdog are very common. The data fit the theory remarkably well: guaranteed returns peak at around the predicted probability, and about one-fifth of the bets in the sample produced guaranteed returns above 58 cents per dollar of free bet.
There is one wrinkle. Placing the free bet on extreme longshots does somewhat better in generating guaranteed profits than the simplest model predicts because bookmakers tend to apply much higher effective margins to longshots than to favourites. But this does not change the main result: if your objective is guaranteed profit, the best free bets are on outcomes with roughly a one-in-five chance of winning.
Key research finding: The best free bet depends on what you are trying to maximise. For expected value, use the biggest longshot available. For guaranteed profit, the sweet spot is much less extreme — roughly a 20% win probability at typical sportsbook margins.
Practical advice: If you want to turn a stake-not-returned free bet into guaranteed money, look for bets with approximately a 20% chance of winning (typically decimal odds of about 4.5) and hedge the other side with a different bookmaker.