Any measure of playing style is only as good as what it can explain beyond player overall ability. If we apply this standard to playing style categories derived from basic match stats, how do they hold up?
Last week, I looked at whether basic match stats, like aces and minutes played per point, could help describe a player’s playing style. In this post, I expand on the set of style features and delve into the clusters of playing styles they reveal.
Recently, I’ve had a few posts on head-to-head effects. The biggest takeaway was one we probably all knew going in: Head-to-head effects may exist, but good luck finding them. With so many small sample sizes for most head-to-heads, we need a way to group ‘similar’ players. In this post, I look at whether categorizing players by playing style might be possible using basic match stats.
Ahead of the final matches in Miami this weekend, we look back on the players who triumphed in the face of the most pressure-filled matches so far this month.
If you thought Serena vs Maria was a lopsided head-to-head, you are right. But it isn’t the biggest head-to-head effect in recent WTA history.
Matchup effects are a common idea in tennis commentary. It is the thing at the heart of comments like ‘her game matches up well’ against her opponent. One way to think of a matchup effect is as a surprising head-to-head, when results go against what the overall ability of both players would have us expect. Do such effects exist? And are they substantial enough that they matter when it comes to making better predictions about tennis results?
Novak Djokovic looked unstoppable at the Australian Open and he heads into Indian Wells as the hands down favourite for the title. With GIG predictions giving him a 1 and 2 chance to win the whole thing, the question of the event will be whether anyone can stop Djokovic?
The women’s draw for the first of the Sunshine Double is out. In this post, I break down the prospects for the 10 women most likely to impress at Indian Wells this year.
With the close of the second month of the 2019 season, we look back at the biggest pressure performances of the men’s tour. Reilley Opelka takes top honours for having faced the most service pressure in a match in February, racking up 15.5 break point equivalents (BPEs) against John Isner in New York. Opelka also saved the most of the BPEs faces winning a whopping 92% of high-pressure service points.
A model that could predict a player’s performance on serve would have a number of interesting uses like forecasting the outcome of matches or identifying surprising performances. But for any of these uses the model would need to be accurate and reliable. How should we evaluate a model’s performance? And how do we know when a model is good enough?