The Favourite Longshot BiasWhen you take a longshot punt with plans for the winnings already laid out, alarm bells should sound off. Robert Blincoe says that getting ahead of yourself shows that it’s time to go back to basics.
If you’ve ever looked at a longshot, a 33 to 1 chance say, and thought “that must be worth a punt”, then you need a simple crash course in betting markets.
Bettors value longshots more than expected given how rarely they win, and they value favourites too little given how often they actually win. This is the favourite longshot bias and it’s true. It doesn’t matter what the sport—horse racing, football, tennis, golf, boxing, baseball—the same bias is there.
This year is the 60th anniversary of this betting anomaly. And since the beginning, study after study by economists and statisticians has confirmed the existence of the bias (with some very specific exceptions I’ll mention later) in the US and UK for bets placed with bookies and on the Tote. There’s even evidence of this bias on the betting exchanges, though much less pronounced. The concept is so important that its impact and effect is still studied by economists and forecasters.
Unfortunately the window when it was possible to make a profit by following a simple betting system, which required no knowledge of anything other than the available odds, is closing. Sophisticated bettors have acted to reduce the effect of the bias, but who wants to bet that blindly anyway? Knowledge and understanding of the bias is still a valuable part of a gamblers armoury.
Professor Leighton Vaughan Williams, director of the Political Forecasting Unit and Betting Research Unit of Nottingham Business School, Nottingham Trent University (and author of Betting to Win) has extensively studied and written about the bias. “It still exists, but it’s not big enough to make a profit,” he says. “You can make an above average return, but you can’t make an abnormal return. You lose less by backing at short odds, significantly less.”
But losing less is the foundation of a gambling plan.
“If you’re backing at short odds, you can lose very little. If you’re able to know a little more than that, and you’re not backing every favourite, you haven’t got far to go to turn it into a profitable strategy,” he continues.
At the longshot end you’re throwing your money away unless that 20-to-1 shot should really be priced at 10 to 1. “If you’re backing every 20-to-1 shot, you’re going to be losing more than half your money. The bias applies on average—it’s not a strategy which says you shouldn’t back 20-to-1 shots, but it makes you think hard about betting on longshots,” says Vaughan Williams.
Crunching the Numbers
A lot of longshot-favourite bias studies have been made on horse racing. Renowned UK economist Jack Dowie studied 2,777 flat-season races in 1973. A pre-tax profit would have been made on betting all horses starting at 4 to 6 or shorter. A study by Robert Henery, published in the Journal of the Royal Statistical Society, looked at 883 races run in 1979 and 1980. It showed a return of 97.9 percent to bets on horses starting at 2 to 5 or shorter, but only 10 percent to horses starting longer than 38 to 1.
A paper in 2000’s Scottish Journal of Political Economy provided evidence that the bias was there in football. Plus, work conducted by Michael Cain (not that one), David Law and David Peel published in the Bulletin of Economic Research identified the bias in boxing, cricket, snooker, tennis and baseball as well. A nice result from this was that backing favourites (1 to 2 or shorter) in the boxing and cricket would have generated pre-tax profits ranging from 12 percent to 16 percent.
All this work stems from the efforts of psychologist Richard Griffith, stationed at the Veterans Administration Hospital in Lexington, Kentucky, who identified the favourite-longshot bias, and published his findings in the American Journal of Psychology in 1949.
However, where it doesn’t stand up is in Hong Kong and Japan. The reasons aren’t known for sure, but a popular theory for the Far East is based around how their large Tote pools work. With the betting exchanges, the longshot odds given are more reflective of reality and that’s why the bias isn’t so pronounced with them. A 33-to-1 shot on Betfair would really be something like a 33-to-1 shot, whereas a 33-to-1 shot with a bookmaker would really be something like 100-to-1. But at the lower odds, an even-money shot with a bookie is even money.
Vaughan Williams does warn, however, that at very shot odds on the betting exchanges, the bias works the other way. “People tend to overestimate at extremes,” he says. “If someone is 100-to-1 on in golf match, people think it’s a done deal, but it’s not. On the exchanges, they think the real odds must be 1,000-to-1 on and can easily get their fingers burned. When the odds are 1.01 or 1.02 on the exchanges, there’s value to be had there.”
But it’s not just in sports. The favourite-longshot bias was seen in action in the US presidential election. In fact there is evidence there was a super-bias. Vaughan Williams saw this highlighted in the 2004 US Presidential election when the favourite to win each of the states of the Union and the eventual winner was in every single case the same person. “This is the equivalent of the favourite in 50 successive two-horse races winning. Favourite-longshot bias or not, it just won’t happen at the racetrack. Move forward to the 2006 Senate elections and the feat was replicated, as every single contested seat fell to the polling-day favourite. If this is not just an amazing set of coincidences, we are witnessing here a bias so extreme that a strategy of backing the favourite in every general election race is a sure-fire strategy for winning every single bet placed.”
And there was value to be had by betting on the favourite in the 2008 presidential elections, with Obama a clear favourite, and his lead very accurately forecasted by the polls. This led Vaughan Williams to put his money where his mouth is. “I made a respectable profit,” he says.