Ranking analysis: a comparison with WTA-rankings

In order to give a better impression of the differences between the WTA-ranking and my ranking for women's tennis, I will present an analysis of the most notable differences between these rankings. I will examine three players that have been relatively overrated by WTA, and three that have been overrated by me. I will assess the causes of the difference, and explain what this means for the interpretation and use of each ranking system.

I have compared the WTA-rankings of May 2 of this year [2016] with my ranking of April 30. By comparing rankings and ratings/scores of both lists, I have selected a few anomalies either way which I will use for my discussion of differences. I have selected three players that are ranked significantly higher in the WTA-ranking: Roberta Vinci, Anna-Karolina Schmiedlova and Nao Hibino. I have also selected three players that are ranked significantly higher in my ranking: Kiki Bertens, Peng Shuai and Arantxa Parra-Santonja. I will start with the first list.

Before that, I think I will have to give my ranking a name, with an acronym I can refer to (which I haven't done before!). I will call it my Expected Performance Level rating, or EPL-rating. In this article, and from now on, I will refer to it with that abbreviation.

Roberta Vinci: WTA 3615 (8th), EPL 5187 (24th)

Roberta Vinci is currently ranked solidly inside the Top 10 in the WTA-ranking. Of the WTA-points, a significant number (1300) came from reaching the US Open-final, and a lot of other points came from St.Petersburg, which Vinci won (470), and a semifinal in Wuhan (350). These three tournaments account for over half of the points. In my ranking, Vinci has been ranked at #12 on February 28 of this year, after a good performance in Doha, and just a few weeks after St. Petersburg. However, losses against lower-ranked players such as Chang Kai-Chen, Magdalena Rybarikova and Laura Siegemund have thrown her back. In other words, Vinci's performance is far from consistent. In my ranking, all matches are treated equally and highs and lows will cancel each other out. In the WTA-ranking, the highs outweigh the lows, leading to higher rankings for players with high peaks (even if they also have low lows).

Anna-Karolina Schmiedlova: WTA 1360 (35th), EPL 4879 (96th)

Anna-Karolina Schmiedlova has had a terrible 2016 season so far, only winning two matches, and losing eleven (up to April 30). Moreover, both wins went in three sets and all losses came in two sets. And one of her wins was against Arina Rodionova, who is ranked so much lower that her win didn't gain her any rating points. Over the entire season, Schmiedlova has lost 282 EPL-points, which is by far the most of all players. So where does the high WTA-ranking come from? Well, Schmiedlova played a very good summer last year, starting with a tournament victory in Bucharest (280 points), followed by a.o. a quarterfinal in Cincinnati (220), a third round in the US Open (130) and a quarterfinal in Wuhan (190). In my ranking, she ended the year as #32, comparable to her WTA-ranking at the time. However, her level has dropped considerably since then, and my EPL-ranking has picked this up immediately, whereas the WTA-ranking is based on the best 16 results in a one year time-slot, which makes it much less sensitive to sudden drops in performance level.

Nao Hibino: WTA 1010 (61st), EPL 4633 (197th)

Basically, Nao Hibino's claim to fame is winning one WTA-tournament last year, in Tashkent. This won her 280 WTA-points. Apart from that, she won three 50K ITF-tournaments and reached the final of a 100K-tournament, for 80 points each. In the WTA-ranking, this is enough to reach the Top 100. In my ranking, Hibino was slightly unlucky that some of the 50K-tournaments she won were "below the radar", i.e. they were not processed and thus these victories are not reflected in her ranking. That also means, however, that the opposition there was not strong enough to gain many points off of anyway. So, on top of the WTA valuing the high peaks much stronger than low lows, not taking into account the strength of the opposition explains a significant part of the difference in ranking positions for Hibino. All international WTA-tournaments will reward 280 points if you win them, regardless of the level of the participants. Tashkent was not a particularly strongly occupied tournament, to put it mildly. Hibino did win 113 EPL-rating points there, but she started at quite a lowly EPL-rating - she went from 4599 to 4712. She ended the season at 4722, rose some more in January, to 4744, and from there she started to slide back down the rankings. Responsible for this were heavy losses against Jelena Ostapenko, Duan Ying-Ying, Lauren Davis, Catherine Bellis and Cagla Buyukakcay. And, in the EPL-rating, the scorelines of matches count, unlike in the WTA-scoring, which is solely determined by the round you reach in a tournament. Hence, I can conclude that Hibino's WTA-ranking is an overestimation of her actual performance level, if you look at her opponents and the scorelines she has produced against them.

Kiki Bertens: WTA 803 (82nd), EPL 5141 (32nd)

Kiki Bertens is someone whose WTA-ranking is not a very good indication of how good a player she actually is. After losing her points from the 2014 French Open at the 2015 edition, due to a slightly unlucky draw (Svetlana Kuznetsova in the first round), she dropped outside of the WTA Top 100. As a result, she had to play qualifiers at most tournaments. At the time, in my EPL-ranking she was ranked #71. That difference was, and still is, for a large part due to the fact that the WTA-ranking does not include Fed Cup matches. Bertens has been impeccable in Fed Cup matches over the past few years, and has collected many EPL-points there. Another reason for the difference is that Bertens tends to win her matches in a dominant fashion - especially her qualifiers - and tends to lose matches in a much closer way that doesn't cost her too many EPL-points. The 2015 US Open is a case in point. There, she came through qualifying easily, won her first round in three sets and then lost to Serena Williams, playing well: 7-6 6-3. During the 2016 season, this pattern has continued, and Bertens has collected 198 ETL-points over the season, which ranks her in the top 10 of season's progress. In Fed Cup, she has again been perfect and collected 98 points there (almost half her total). Overall, what we see here is that Bertens tends to do well in matches that don't earn her many WTA-points, and she tends to falter when the stakes get higher. This could be a mental thing. If you've seen her play in Fed Cup recently, you could agree with me she is capable of playing at a level worthy of Top 20 or higher. What she needs for that is some sort of breakthrough performance in the near future.

Peng Shuai: WTA 74 (436th), EPL 5001 (65th)

For Peng Shuai, the difference in rankings is enormous. In this case, this is mostly due to her injury and the way injuries affect rankings in both systems. Peng Shuai has been out of singles competition for ten months due to a back injury. As a result, she has lost the majority of her WTA-points. When she got injured, at the French Open of 2015, she was ranked 26th (WTA) and 30th (EPL). Peng played her first singles matches after her injury at Indian Wells this year. Before that tournament, Peng had only 20 WTA-points left and was ranked #768. From then on, she's gradually trying to regain her ranking. Her lowly ranking of #436 reflects the early stage of this process. In January, she had already started playing doubles again, which was enough to get her EPL-rating back. She returned to #42 in my ranking after the Australian Open, because in my system, if you're injured longer than half a year, you start losing just 1 rating point per day. Overall, this injury cost Peng 43 EPL-points. After returning at Indian Wells, her level has proven to be a bit lower than expected, so she has lost some 100 EPL-points over the last months, dropping to #65. You can expect that rankings will only start to resemble each other again after a year back on tour.

Arantxa Parra Santonja: WTA 14 (856th), EPL 4721 (150th)

Arantxa Parra Santonja is a player who has focused primarily on playing doubles over the last few years. Her 14 WTA-points come from just one tournament: the qualifying tournament of Nottingham in 2015. Playing very few singles matches will lead to a low WTA-ranking. In the EPL-system, things work differently. As long as Parra Santonja is an active doubles player, she keeps her EPL-rating. The few singles matches she plays will still be processed. The reason she keeps her EPL-rating is that it is assumed to be a reasonably reliable estimate of her level if she would choose to play another singles tournament. For the same reason, a player such as Sania Mirza, current #1 in doubles, still has a singles EPL-rating even though she hasn't played a competitive singles match in five years. For the EPL-rating to change, you need to play matches. If you don't play, your EPL-rating stays the same until you're entirely out of competition (singles and doubles) for over half a year. Thus, Parra Santonja's EPL-rating is still an indication of the level she used to play at when she still played many singles matches. Her WTA-ranking, based on a single tournament, cannot fulfil that function.

Conclusion

One needs to keep in mind that the difference between both ranking systems results from a difference in their intended purpose. The WTA-ranking is in essence a measure of merit, where points are awarded as a prize for achieving a result. The resulting ranking is used for determining which players are allowed into tournaments, and for seedings. The fact that all similar tournaments award the same number of WTA-points ensures that all tournaments remain equally attractive in terms of defending or improving your ranking. My EPL-rating has a different purpose: it is intended to measure the strength of a player at this time, and is therefore designed to be more sensitive to the form of the week. The more recent a result, the more it will factor into the EPL-rating. In the WTA-rating, all results of an entire year factor in at the same level, and after a year the points gained the year before are all lost. The EPL-ranking gives an indication of where a player should be in the WTA-ranking if she would play at the current level for a whole year, and is therefore indicative of future rises and falls.

Given this analysis, one can expect players such as Bertens and Peng Shuai to rise on the WTA-ranking, whereas Vinci, Schmiedlova and Hibino can be expected to fall down the WTA-ranking in the coming months. Unless, of course, their level suddenly changes again, but that's the sort of uncertainty every prediction contains.

Michiel van de Sande, May 16, 2016