Bookmakers and Bettors Both Misprice Longshots — Research Proves It
For decades, academic economists have studied horse racing markets looking for systematic pricing errors — and they keep finding the same one. Longshots are consistently overbet relative to their actual chance of winning, while favourites are consistently underbet. The effect is persistent, documented across multiple countries and decades of data, and it has direct implications for how you construct a betting strategy.
UK favourites win approximately 34.4% of all races. That figure, taken at face value, makes favourites look like poor value — they lose two thirds of the time. But the favourite-longshot bias tells a subtler story. When you compare the implied probability of each horse’s odds against its actual win rate, favourites consistently outperform their odds while longshots consistently underperform theirs. A horse at 2/1 is “supposed” to win 33% of the time — and favourites in that price range tend to win slightly more often. A horse at 33/1 is “supposed” to win about 3% of the time — and longshots at that price tend to win significantly less often.
Understanding why this bias exists and how to exploit it changes the way you approach favourite statistics and the odds you accept.
What Academic Research Tells Us About the Bias
The favourite-longshot bias was first formally documented in American pari-mutuel markets in the 1940s, and subsequent studies have confirmed it in UK bookmaker markets, exchange markets, and international racing jurisdictions. The core finding is remarkably stable: the rate of return on bets placed at shorter odds is higher (less negative) than the rate of return on bets placed at longer odds. In other words, if you must bet, favourites lose you money more slowly than longshots — and in some subsets of races, favourites can actually produce a small positive return before commission.
One of the most illuminating studies, conducted by Paton and Vaughan Williams and published on ResearchGate, examined the performance of debutant horses and found that previously unraced two-year-olds won at a rate approximately 16% higher than similarly priced horses that had already raced. This effect was attributed to insider information flowing into the market — connections of unraced horses know things the public doesn’t — and it represents a specific, measurable version of the broader favourite-longshot bias: the market misprices horses where information asymmetry is greatest.
Horses at odds of 2/1 or shorter win approximately 50% of UK flat races. That’s a better hit rate than the implied probability at those prices would suggest, which means the market is slightly too generous to anyone betting at the shorter end. At the other extreme, horses at 33/1 or longer win considerably less than the 3% their odds imply. The gap between implied probability and actual win rate widens as the odds increase — a 100/1 shot isn’t just slightly overpriced, it’s dramatically overpriced relative to its actual chance.
Why the Bias Persists: Psychology, Pools, and Bookmaker Behaviour
The favourite-longshot bias should, in theory, be corrected by the market. If longshots are systematically overpriced, smart money should pile onto favourites and drive down their odds until the pricing error disappears. Three forces prevent that correction.
The first is psychology. Recreational bettors are drawn to longshots because of the appeal of a large return from a small stake. The dream of turning one pound into fifty is more exciting than the prospect of turning five pounds into twelve — even though the latter is the better bet. This psychological preference for high-payoff, low-probability outcomes has been documented extensively in behavioural economics and is sometimes called the “hope” factor. Alan Delmonte, then Chief Executive of the Horserace Betting Levy Board, observed that bookmakers’ operating margins had exceeded recent comparative levels in two consecutive financial years — margins that are partly sustained by recreational bettors’ persistent overinvestment in longer-priced horses.
The second is bookmaker pricing. Bookmakers know that the public wants to back longshots, and they shade their prices accordingly. The overround — the built-in margin on any market — is not applied evenly across all runners. Bookmakers typically extract a larger proportional margin from the longshots than from the favourites. A 10-runner race with a 120% book might have the favourite priced fairly close to its true odds, while the 33/1 shots carry a much higher percentage of the total overround. This makes longshots even worse value than the raw bias would suggest.
The third is structural. In pool (pari-mutuel) markets, the collective behaviour of all bettors determines the odds. When a disproportionate volume of money flows onto longshots, it drives their pool odds lower than they would be in an efficiently priced market — which means the payout on a longshot winner is smaller than it should be, while the payout on a favourite winner is larger. The bias feeds itself.
How to Use the Favourite-Longshot Bias in Your Selections
The bias doesn’t mean you should blindly back every favourite. Favourites still lose the majority of their races, and the bookmaker’s overround ensures that backing all favourites is still a losing strategy over a large sample. What the bias tells you is that value in horse racing skews towards the shorter-priced end of the market, and the longer the odds, the more sceptical you should be about whether the price reflects a genuine chance or just public hope.
I use the bias in two ways. First, when choosing between two selections at different prices, I give a mild preference to the shorter-priced option — all else being equal — because the bias tells me the market is more likely to be right about that horse’s chance than about the longshot’s. This isn’t a rigid rule; strong form analysis can override it. But it functions as a tiebreaker when two horses look equally appealing on form.
Second, I apply heightened scepticism to any horse priced at 20/1 or longer. At those prices, my form analysis needs to give me a specific, evidence-based reason to believe the horse has a better chance than its odds imply. “It could run into a place” or “you never know with these handicaps” are the kinds of reasoning that feed the bias — they’re hope dressed as analysis. A genuine value longshot exists when the form, the going, the draw, and the trainer pattern all align to suggest the market has genuinely underestimated a runner — not merely when the potential payout is exciting.
Exchange markets partially correct the bias because the odds are set by peer-to-peer trading rather than bookmaker pricing decisions. On Betfair, longshots tend to be priced slightly more accurately than at traditional bookmakers because the market participants are, on average, more informed. If you do bet longshots, exchange prices usually offer better value than fixed-odds bookmaker prices for that end of the market.
