US Open 2014: Avoid Murray and Wawrinka

The statistics suggest it will be a stretch for Andy Murray to progress far at the US Open

Taking his usual sideways look at the data behind the men's game, Jack Houghton presents a statistical smorgasbord to help you on your way at the US Open...

Tasked with coming up with tips for the men's Grand Slams that are based on (often quirky) statistical approaches can be laborious. Anyone who enjoys playing around with data knows that, for all that spreadsheets and databases can do marvelous things very quickly, they require information to be in a particular form in order to unleash their computing power.

When it comes to sports - and especially tennis - the chances that you can find the data you want, in the way that you want it, are slim. More often, it's a case of spending hours manipulating the information that is out there, until it is at last in a state that you can do something with it. Sometimes, midway through your labours, the data dries up. Sometimes, you get to the end and the data tells you nothing.

In the last few weeks, the data gods have not been kind to me, and a lot of potentially fruitful work has been halted, or has come to nothing. So rather than taking my usual approach of presenting a single piece of analysis, and showing how it might prove profitable in the upcoming Grand Slam, I thought I'd present you with a smorgasbord of statistical titbits, many not fully realised as of yet, and see which ones might prove useful at Flushing Meadows.

Second Serve and Second Serve Return

My big tennis data project this year has been analysis of which key player metrics are the best predictors of future performance. It led to recommendations of Stan Wawrinka at 48.047/1 and Rafa Nadal at 4.67/2 in Australia, and Roger Federer at 7.06/1 at Wimbledon.

The analysis - an ongoing project - is showing that two metrics that are head-and-shoulders above everything else in terms of predicting winners: a player's second-serve points won percentage; and a player's second-serve return-points won percentage.

Applying the analysis to the US Open highlights three players who seem to have their games in the best shape on the hard courts: Berdych (75.074/1), Federer (4.67/2) and Djokovic (2.3611/8). The last two in that list might not offer any great insight, but Berdych looks a big price in a year where the top seeds are proving less reliable than in recent times.

Perhaps most usefully, the analysis also suggests that Andy Murray (7.26/1) and Stan Wawrinka (24.023/1) are players to be avoided. Murray's second-serve percentages continue to be woeful for a player of his standing, perhaps demonstrating that it is this area of his game that has been most affected by back surgery. And Wawrinka, topping the polls in the key metrics at the start of the season, has been unable to get either his second serve, or return of second serve, up to the levels he was reaching in January during this hard-court season.

Australian Open versus Wimbledon as Predictor of US Open Success

Despite the greater similarities between playing conditions in Australia and the US, early work is showing that those players successful at Wimbledon are more likely to go on to US Open success than those who were playing well at the Australian Open. It seems that recent form is more significant than a preference for a particular surface - in this instance anyway. More work is needed here, but it adds weight to the idea that Wawrinka is a player to be avoided.

Losing Service Games

Early days for this one, but it seems as if the market is under-reacting when a player loses a service game. The analysis is complex, and my data set is currently too small, but laying players immediately after they lose a service game at the US Open is likely to be a profitable strategy. Hopefully more on this at a later date.

The Effect of Long Matches

Again, more work is needed here, but one branch of analysis I have done, which looks at how players performed against their ranking depending on how many sets they had played in their previous match, shows nothing of any significance. So if you tend to factor in to your betting what the television pundits tell us - "Djokovic had a hard semi-final which has to be an advantage for Federer..." - then you are likely getting it wrong. As the US Open reaches the closing stages, do not factor in the relative fatigue a player might be feeling.

Taller Players are Better

But this is not as significant as some people think. Taller players tend to win more matches against shorter players, but it's not hugely significant. And the advantage has a definite ceiling. Again, the analysis is very much at the crude stage, but it seems as if a few centimetres of height is advantageous, but any more that 5cm, and the advantage tails off. Haven't worked out how this information might be profitable as of yet, but anyone who is planning to lump on Isner at 160159/1, or Karlovic at 1000n/a, on the basis of their loftiness, you might want to duck out. Although, having said that, Karlovic's second-serve percentage has been a stunning 59 on the hard courts so far this season!

Recommendations
Back Djokovic at 2.747/4
Back Federer at 17.016/1
Back Berdych at 75.074/1
Oppose Wawrinka and Murray in their early matches.