Our stats analyst Omair is already famous for his statistical articles and now he’s inspired Ludolf, another reader of Women’s Tennis Blog and fan of tennis numbers, to contribute his take on the WTA rankings, i.e. he compared the actual rankings with the performance of players at the biggest events – Grand Slams and WTA Championships. Enjoy!
The WTA ranking does not rank properly the strength of players in the sense of the ability to win a match or a tournament. The main reason is that it is cumulative and therefore underestimates players without a sufficient number of tournaments. We need to compute an average to avoid this. Moreover, the WTA ranking considers tournaments with weights not depending on the quality of players on them.
To achive the above-mentioned goal it seems to be more appropriate to take the average result from majors and the WTA Championships, the tournaments where the best players are almost always present. By the result of a player we mean the number of her losses (including the rounds which she did not achieve), i.e. 0 for the winner, 1 for the defeated finalist, 2 for defeated semifinalists, and so on. The results for the last 52 weeks are presented in the table, completed by the WTA ranking (after the Australian Open) for comparison. (FO: French Open, W: Wimbledon, USO: US Open, WTA: WTA Championships, AO: Australian Open)
It is not a surprise that Serena Williams and Kim Clijsters reached much better position than their WTA ranking. It should be noted that the average results of these players (and some others, too) are not relevant in the same degree as of the other players due to a small number of tournaments they took part in. For instance, had Kim Clijsters reached the next round at the French Open, she would have taken the 6th position with the average of 3.50. Also, Samantha Stosur and Agnieszka Radwanska got significantly worse result than their WTA ranking.
We can conclude that the first three players are significantly better than others, except maybe Serena Williams and Kim Clijsters. This is in a good correlation with the WTA ranking, which need not be true in general.
There are some weak points of this average result. First, rather small number of tournaments is considered, hence the average has a rather big deviation. This has been already pointed out for Kim Clijsters. Let us also note that the first three players have the smallest possible differences. Also, it does not describe the actual level of players but an average in past 52 weeks.
What do you think of this perspective? It’s nice to have one more approach, but as Ludolf himself stated, it still has many flaws and the fact that only five tournaments are taken into account, it’s pretty limited. Plus, almost all the best players showed up at the tournaments analyzed, but still it’s “almost”. (photo: © Neal Trousdale)
Sorry to say, Marija, but, although some stats can be interesting, this bombardment of stats articles recently is spoiling your website in my opinion. Seems to be never ending and has gotten boring. No offence meant, just feedback.
Omair, this time I must disagree with your whole premise that Slams and the Championships are somehow more important indicators of a player’s ranking. I have thought for years about a better system to rank players. My premise is it doesn’t matter where player A beats player B, that win should be counted the same in any venue. The problem is trying to come up with a better ranking system instead of using where a player finishes in a tournament use the ranking of the players she beat in that tournament. Start with that premise and help me figure out how to do a ranking based on who you beat or lose to, not where you finish in a tournament.
I geek out over this number-crunching stuff so keep it coming!
I agree with Tony’s comment above, but would like to extend it beyond what’s happening on this blog. Statistics are the rage of the digital age – what with a simple algorithm (or, as far as tennis goes, the hawk-eye technology) making virtually (excuse the pun) any statistic possible; and whatever is possible shall be done… and, soon enough, – overdone. As if it makes us all much wiser. Now, I am not trying to say that we cannot learn anything from the statistics – of course we can -, what I am trying to say is that, fascinated by the toy, we may easily forget the game, and the game is, or at least should be, what it’s all about. Any game worth the name is intrinsically not analysable and, therefore, unpredictable.
Tony, I see what you mean and I agree that it could’ve been too much. I will definitely make a break with stats, at least until Roland Garros, but even then I will post less stats that during this Australian season.
John, this article was not written by Omair, it was written by Ludolf.
Erwin, there will definitely be more stats analysis, but more occasionaly than this January-February.
Tulp, good point, but as you know we are only viewing the statistics in order to have fun and gain another perspective. We all know the game is unpredictable, as it should be, and we are not putting the stats ahead of the beauty of the game.
I am also not a big believer in statistics especially nowadays when tennis has got so unpredictable.
So yes, let’s go back to tennis talk :-). How about a revision of latest tennis outfits. Caro’s Roland Garros dress is already available.
JohnyB: I agree that the best indicator are mutual matches whenever and it might be interesting to find a proper evaluation system. This is why I like the statistic of Top 10. It would be nice to have a tournament like the WTA Championship (with a plenty of matches of top players) several times during a year. However, in other tournaments the number of matches between top players is limited.
I disagree with your objections against this stat. I do not claim that it is more important if a player A beats a player B in these tournaments (nevertheless the players are more motivated to play their best) but the importance of these tournaments is that we have more matches between the top players and the results give more information to evaluate their strength. It surely gives a better evaluation than the WTA ranking and might be considered (by results) as something between the WTA ranking and the system you propose.
Marine, haha and some people would like a break from fashion 🙂 Where did you see Caroline’s new look?