Is home-court really an advantage?

The well-known phenomenon of home-ground (or court) advantage superficially seems to be as alive and well for the men's game in Australian tennis, as shown by Lleyton Hewitt's 2014 win in Brisbane and Bernard Tomic in Sydney the year before.

Throw in the emergence of young stars Nick Kyrgios, Thanasi Kokkinakis and Luke Saville, not to mention promising form by journeymen Marinko Matosevic, Sam Groth and James Duckworth; and all indicators point towards 2015 potentially being a stellar year for Australian men on the month-long home leg of the ATP Tour.

However, after countless summers of disappointment watching highlyrated Aussies bombing out of the Open earlier than expected, I wondered whether it was because my expectations were irrationally inflated due to having factored in too much home-court advantage, or rather that the assumed advantage itself was simply well short of the general anecdotal mark.

Either way, one has learned to temper one's expectations this time around. After all, no Australian male (or female for that matter) has won the major singles title in Melbourne for nearly 40 years. The last was Mark Edmonson - the answer to a useful trivia question.

As econometricians tend to do, I believed it to be worth putting the home advantage concept to a rigorous scientific test The opportunity to do so was made possible by way of research projects currently in progress with a collaborator, Dr James Reade (University of Reading, UK), for which we required lots of tennis data anyhow.

We started with the full sample of 29,531 Grand Slam and ATP matches over the period 2003-2013, excluding the Davis Cup, which is not entirely comparable. We began by comparing the winning probabilities of a player beating an opponent when the match takes place in that player's home country, assuming the opponent was not a compatriot However, this probability is implausibly low, predominantly because of selection biases towards lesser-light domestic players finding themselves in the main draw of home tournaments (wild-cards, unlikely qualifiers for example).

Therefore, we wanted to produce estimates to account for a range of match-specific factors that might condition these probabilities, most obviously these player-ability differences, but also others such as prize money and ranking-point incentives, as well as age, height, weight rest days, even handedness.

For this, we used sophisticated regression modelling to estimate the magnitude of home-court advantage overall in men's tennis, and then break it down for each of several individual major tennis-playing nations.

Our results indicate that generally, home-court advantage is significant Taking the 'baseline' of equally-able players on neutral ground, where the victory probability for both players is 50 per cent playing on a home court then increases the home player's likelihood of winning to a conditionally estimated 56.3 per cent.

However, when isolating the 1946 matches over the same period on Australian soil, the home advantage by Australian male tennis players was substantially reduced, with the probability of victory rising to 53.7 per cent, which was not statistically significantly different to 50 per cent when accounting for estimation error.

Put simply, home-court advantage was nonexistent in that sample. The results do not explain why Australian players do not perform more strongly on home soil, but would appear to justify my reticence to get my green and gold hopes up.

Intriguingly, we find this result is not purely an Australian phenomenon, as this was found to also be the case for UK and German players. Nevertheless, for all the other major tennis-playing nations, such as the US, France, Spain and Italy, home-court advantage was significant and strong.

Also of note for the Australian case is that when the non-seeds (lower ranked) players were separated out, home advantage became significant, meaning that the more highly-fancied Australians carry much of the blame for the overall result One possible interpretation is psychological - maybe playing at home adds too much pressure and makes them tighten rather than firing them up. Indeed, we could be investigating this possibility further with cooperation from Tennis Australia.

Ultimately, I am not expecting much Oz joy watching this year's Open, starting Monday week, but the youngsters mentioned earlier, who on early signs seem to thrive on playing in front of home crowds, may prompt an upward tick in my expectations in the near future. Who knows - maybe they will also play a big role in reversing our main finding (with an updated sample) in the future.

This piece was originally published in The Australian Financial Review.

Liam Lenten is a senior lecturer in economics.

Image: Core Materials