When Philipp Kohlschreiber defeated the French Open Champion Rafael Nadal at the Gerry Webber Open at big odds of almost 7.00, it was one of the biggest upsets over the past month. Nadal, was clearly tired after winning the French Open but the punters still backed him to win.
Of course, it was a massive win for Kohlschreiber, who had never defeated a number one seeded play since he defeated Marin Cilic back in 2010. So perhaps he was in a rare stint of form. It was his third win on the trot, and faced fellow compatriot, ageing, yet in form Tommy Haas.
Kohlschreiber was listed as favourite at 1.66, whilst Haas had to settle for the slight outsider odds of 2.40. Least to say that Sportpunter rated Haas at 47% chance to win, so there was some value there on Haas.
If you don’t already know, Haas defeated Kohlschreiber in straight sets. This got me thinking. How often does the market over compensate for a player who had a massive win the week after? It is possible that the chances of a player winning the week after are less than what the public perceive?
Perhaps Kohlschreiber had already won his grand final after defeating Nadal, or was too happy and excited to hold his nerve for the next match. Perhaps Haas had to focus more knowing that his opponent had just taken out one of the world’s best.
Well I decided to test this theory, and using data going back to the start of 2005, looked at all matches where a player had won at odds of greater than 5.00 (not including retirements) and how they went in their next match, so long as it was in the same tournament.
Since 2005, this has occurred 328 times. In 117 (36%) of them, the player who won previous at odds of 5.00 or more won their next match, whilst 64% of the time, or almost 2/3rds of the time – they lost.
Had we bet to win $1000 on a player who previously won at 5.00 the game before, we would have bet $254,305 (average of $775) and lost a total of $30,130. Our precent return on investment would therefore be -11.8%.
Clearly, without doing any form whatsoever, we have shown that there is no value in betting on a player who had previously won at odds of 5.00. However interestingly, is how we would have done betting against a player who had won at 5.00 or greater the match before.
Since 2005, we would have bet $992,724 and made a profit of $33,591. This represents a return of investment of 3.6%
So just by this analysis alone and not even performing any form analysis, one would have made 3.6% ROI and $33,000 by backing against a player who previously just had a big win.
So clearly, the market overreacts based on a big victory the previous game. This, I can assure you, happens in every sport.
Sportpunter’s tennis model is available to subscribe to. Click here for more information.
![Related Posts with Thumbnails](http://www.linkwithin.com/pixel.png)
Very nice, Jonathan.
I’m always reading your such posts with great pleasure.
Does this factor include in model i.e. change probabilities in general about 5-6 percent in all theese situations?
HI Kirill, no this isn’t factored into the model, as the model simply tries to predict the probability of winning and doesn’t take into consideration biases in the odds.
However, because of this, the model doesn’t bias like the odds do and mroe often then not, from my experience, you will find a bet against the player who just had the big win.
Winners don’t come much bigger than Rosol, so that would make Kohlschrieber a near certainty in R3?
ha you would think so. Massive massive win. Although I just checked if the odds were greater than 5.00. Didn’t check if the odds were greater than 10 or 20!