Let me bring you up to speed… I tipped Djokovic to win the French Open a couple of weeks ago. He finally beats Nadal at Rolland Garros at the 7th time of asking, he then beats Murray to get into the final. His odds of winning this final were 1.15 (-666 US).
Things are looking rosy. Novak is set to complete his career grand slam, joining a very select group of 7 tennis legends. A group that does NOT include Sampras, Becker, McEnroe, Bjorn, Lendl or Tim Henman. Djokovic looks set to win his 9th grand slam placing him solely 5th in the all-time list. At the age of 28 there’s time to win several more slams cementing his place as one of the greatest players of all-time… More importantly that all that I’m going to win some money!
He loses the final in four sets to a fat lad wearing a pair of shorts from the ‘lost property’ box:
…that wasn’t in the script.
In all seriousness Djokovic didn’t play well on the day, Wawrinka did and was the worthy winner. But this does bring about some questions about Djokovic’s Grand slam record: he’s been on top of his game since 2011 and he’s “only” got 8 slams. Is that figure too low, too high or about right?
Ok let’s compare this dude to all former no.1 ranked players:
On first glance it doesn’t look too bad considering Lendl and Connors finished with 8 slams despite (currently) being ranked no.1 for over 2 years longer than Djokovic. A 50% win rate in Slam finals just below the average rate of all former no.1s (56%). Note Lendl’s 11 loses in Finals.
His Slams/Weeks No.1 ratio is also ok compared to the chaps above him in the list. Although I’m the 1st to admit this is a bollocks metric. There are only 4 slams a year so the guys who are ranked no.1 for years are going to appear worse compared to guys who are no.1 for less time. Extreme case: Pat Rafter won 2 Slams but was only ranked no.1 for a week.
The table above provides a little bit of interest, but nothing deeper. The question I want to answer is if Djokovic’s Grand Slam record is worse than what was expected.
So what better way than to have a look at his performance compared to the odds. There’s been plenty of research done that suggests than betting markets are a very good (but not perfect) predictor for what’s going to happen.
Such data can be found at the excellent Tennis-data.co.uk. They have the odds going back about 12 years for ATP, back 7 years for WTA. Using games where they have odds for, and only using games that were completed or someone retired after the first set, we get this:
I’m only showing players that have completed more than 90 games at Grand Slam events, and I’m using the using the Average odds that Tennis-data provides. The important field is the ROI column, what % return you would have by backing a player in every Grand Slam match in the dataset. Basically if a player has done worse than the betting market expects than they will have a negative ROI like James Blake. If they have done better than expected than will have a positive ROI…and won’t you look at fashion icon Stan Wawrinka at the top of this particular list. A surprise winner of 2 slams now!
If you backed Novak using the average odds at every slam match you would have a ROI of 3%. That’s pretty good considering. Maybe this dude’s record isn’t bad after all.
3 thoughts on “French Open Fallout”
Loving the website and this post but what I particularly love is a man who uses average odds for calculations as opposed to top bullshit price at nonsense book. Keep up the good work sir.
Thanks for the kind words Matthew!
Yep, best prices are load of rubbish, but loved by tipsters to inflate their stats.