Friday, January 16, 2015

Evaluating tennis history: Grand Slam Scores

One of the subjects I plan to explore regularly in this space is tennis history. It's a subject I've explored to some extent already, generally through the context of my Melog ratings. While I think the Melog system does a creditable job evaluating tennis performance, it is not without issues.

One of those issues is that the system does not match the evaluation of tennis players in popular opinion. I often find this to be an asset; Melog places value on excelling throughout the year, not merely in big tournaments. But it is inescapably true that most evaluators, including the players themselves, put far more emphasis on Grand Slams than they do on anything else; in fact, “career Grand Slams won,” which obviously ignores achievement in other events, is a commonly-used proxy for all-time ranking.

There is another problem with the Melog system that troubles me far more: It requires a great deal of time to compile the necessary data to evaluate even a single year of tennis. And rewarding as I find that process, it still means that I’ve only got the numbers for 2008-14 so far, which leaves me without a ready means of comparing Roger Federer to Pete Sampras and Bjorn Borg. That would be a nice capability to have.

As it happens, I’ve developed a method that addresses both of those problems at once: it is based entirely on Grand Slam performance, and it is readily calculable for anyone who’s played in the Open Era. And given the content of these introductory paragraphs, you’ve probably figured out that I’m presenting that method here.

The concept is simple. Pick a tennis player and look at his performance in every single Grand Slam he (or she, although I’ve only assembled a database for men so far) plays in his career. Count the titles, finals, semifinals, and everything else back to the first round losses and DNPs, and combine them all into a single number.

How? It probably sounds a bit glib to say “however you want,” even if it is true. With the caveat that “however you want” is a valid answer in its own way, I do have a couple of more specific options in mind.

The first is the simplest: Take the odds of the player winning each Slam right before he’s eliminated and add them up. That sounds more intimidating than it actually is, because I’m not talking about trying to figure out everyone’s exact odds at every stage of the draw. Working under the assumption that each match is a coin flip, you get the following values for each round of a Slam:


All you do from there is count up the number of times a player earns each result, multiply, and add. So looking at, say, Andy Murray, you get:


Murray’s score is, unsurprisingly, very good. To see how good, we have to compare him to other people. The people I’ve chosen fall into two groups: Everyone who has made a Grand Slam final in the Open Era (beginning in 1968), and everyone who has appeared in the top 10 of the ATP Rankings since they began in 1973. (Plus one other guy who we’ll get to shortly.) Out of that group (which includes 174 players, and which I feel easily covers the most noteworthy contenders of the last four decades), here are the top players in order, with intermittent commentary:

1. Roger Federer 25.31

What, you were expecting someone else?

2. Rafael Nadal 18.66
(tie) Pete Sampras 18.66

This tie will be broken as soon as Nadal steps onto the court at the upcoming Australian Open. But for the last six months, Federer's nearest competitors have been precisely even, down to the smallest available fraction of a point. And neither of them is anywhere close to the top spot.

4. Roy Emerson 17.81

Here’s the aforementioned player who has neither been in the ATP top 10, nor made an Open Era final. Emerson was the dominant presence in Slams in the mid-1960s, winning a total of 10 from 1963-67, including all five Australian championships played in that span.

This is the place in the rankings where we start discussing caveats.

The two biggest qualifiers to apply to these numbers are amateurism and the Australian Open. Until the beginning of the Open Era in 1968, professional players were not permitted to play Grand Slams. Many of the world’s best players were therefore restricted from entering the most famous tournaments in the sport. As a result of this prohibition, the fields of the Slams were artificially watered down, and the victories won in those tournaments should be correspondingly discounted.

The same was true to a lesser extent (and for different reasons) of the Australian Open even as the Open Era progressed. For the first couple decades of open play, many of the world’s best players habitually skipped the inconvenient trek to the Slam of the Southern Hemisphere. John McEnroe played only five Australian Opens in his 16-year Slam career – and he showed more interest than his best contemporaries. Jimmy Connors appeared down under only twice, Bjorn Borg just once. Even into the early ‘90s, Australia was sometimes bypassed by some of the game’s best; Andre Agassi didn’t play there until 1995 (and then went on to win four of the next nine editions of the event).

Roy Emerson benefits hugely from both of those factors. He’s not the only player from before 1968 to be helped by the amateur-only rules, but the other great ones of that group (who we’ll see shortly) have a counterbalancing factor: they went pro after some amateur success, and therefore missed out on years of Slam participation. Their triumphs against weaker opposition than they might have otherwise faced can be weighed against the victories they might have secured in later events had they been permitted to participate.

Emerson, on the other hand, retained his amateur status throughout the divided period, and therefore has the diluted draws working unabashedly in his favor. And as an Australian, Emerson always played the Australian national championships (as the event was then known), and was further helped by the fact that many of the world’s best amateurs didn’t make the then-laborious journey around the world to face him.

None of this means that Roy Emerson wasn’t an extraordinary tennis player. But it probably does mean that he wasn’t quite as extraordinary as his ranking here would imply.

5. Ivan Lendl 17.24
6. Jimmy Connors 17.12

Lendl and Connors strike me as two of the more underrated players in men’s tennis history. They won eight Slams apiece, which is a remarkable total – but it’s also responsible for less than half of each of their scores. Lendl famously lost his first four Slam finals, and went 8-11 in finals overall; Connors eked out a winning record in finals (8-7), but was below .500 in semis (15-16).

The thing is, finals and semis are still great results, and these two deserve more credit than they get for reaching as many of them as they did. So I’m giving it to them here.

7. Andre Agassi 16.15

Kind of fits into the Lendl-Connors group, but without having a losing record at any stage. His records in semis and finals are both creditable (15-11 and 8-7), but they don’t quite measure up to those of his most celebrated contemporary; Sampras went 18-5 in semis and 14-4 in finals. But Agassi made more semis, and quarters, and… well, everything else.

Sampras was probably still better. But these rankings at least imply that the difference between them is smaller than might be commonly thought.

8. Ken Rosewall 15.36
9. Rod Laver 15.30

Ken Rosewall made his Grand Slam debut in the 1951 Australian championships at the age of 16. One year later, he played all four Slams and made quarterfinals in two of them. The year after that, he won the Australian and French titles. Over the next three years, Rosewall augmented his trophy case with two more titles (’55 Australian, ’56 US), four finals, and four semis, and then, at age 22, he went professional. As a result, he did not play another Grand Slam until 1968, when he joined the field for the inaugural French Open at age 33.

He won. The next year, he made the French Open final at 34. Age 35 brought a Wimbledon final and a US Open title. 36 and 37 both saw Australian Open trophies. And in 1974, at age 39, Rosewall made the finals at both Wimbledon and the US Open. He restricted his schedule after that, but still made semis in Australia at ages 40 and 41.

Ken Rosewall was a great young player, and he was likely the greatest old men’s singles player of all time. In between, there is a gap of over a decade in his Grand Slam career.

Rod Laver, Rosewall’s countryman, contemporary, and sometimes rival, is generally more celebrated; given that he won calendar Slams in both 1962 and 1969 (separated, of course, by six years of professional play), his plaudits are well-deserved. But in an alternate, pros-always-allowed universe of tennis, I might be more curious to see what Rosewall would have done in his prime years.

10. Bjorn Borg 14.55
11. Novak Djokovic 14.05

Djokovic’s scores for the last five years have been 1.0, 3.25, 2.25, 2.25, and 1.875. Barring severe injury and/or unforeseen collapse in ability, he’s passing Borg later this year – possibly as early as the end of this month. And given that virtually all of Emerson’s work came before 1968 (and much of Laver’s and Rosewall’s as well), Novak is pretty clearly a top-10 player of the Open Era at this point – and climbing.

Borg is one of the players you sometimes see considered for Greatest of All Time status, so his ranking here may not sit well with some. Just remember that this is a career total, and Borg played his last Grand Slam at age 26, and only played the Australian Open once. There are ways of looking at these numbers that make Borg look much better, and we’ll discuss those later. (In-a-different-post later, not in-a-few-paragraphs later.)

12. John McEnroe 12.48
13. John Newcombe 12.22
14. Stefan Edberg 12.17
15. Boris Becker 11.30
16. Mats Wilander 11.20

This group is composed entirely of guys who won 6 or 7 Slams, and makes a pretty clear boundary line between all-timers and guys who were “merely” extraordinary. Nobody below these guys on the list won more than four Slams – which is obviously still an impressive feat. It’s just not going to do quite as much to make you a well-known name among casual tennis fans 20 years after you retire. (Even these guys struggle to reach that level – McEnroe gets there because he was an American and had the rivalry with Borg and threw tantrums, and Becker had the big serve and the Wimbledon title at 17. I was not terribly familiar with the other three before I started paying close attention to tennis.)

17. Guillermo Vilas 8.76
18. Arthur Ashe 8.18
19. Jim Courier 7.75
20. Andy Murray 7.66

As with all active players, Murray’s total is “and counting.” If he makes the quarterfinals in Australia, he’ll pass Courier; if he duplicates last year’s performance at all four Slams (a semi and three quarters), he’ll pass Ashe by the end of the season. But he’s got a lot of ground to cover if he wants to join the group ahead of this one, and I suspect the distance will overcome him rather than the other way around.

21. Tony Roche 6.84
22. Lleyton Hewitt 6.34

Also “and counting,” just not counting nearly as fast as the active guys ahead of him. Hewitt last made a Slam quarterfinal at Wimbledon 2009, and he’s lost in the first round 7 times in the last 3 years. But he’s still going, and I like that about him quite a bit.

23. Andy Roddick 6.12

Roddick made 5 Grand Slam finals, which is an excellent total. He won one of them. The other four times, he played Roger Federer.

If people thought Andy Roddick didn’t live up to expectations, I submit that it may be their expectations that needed recalibrating. He’s right at the edge of being one of the 20 best players of the last 50 years by this metric, and that’s plenty of accomplishment for anyone.

24. Jan Kodes 5.51
25. Ilie Nastase 5.34

Kodes may have had worse timing than Roddick. He won three Slams, but two of them were the 1970 and ’71 French Opens, which occurred just before the ATP rankings debuted. As a result, he ended up with a relatively undistinguished-looking career high ranking of #5, a mark he almost certainly would have exceeded in ’71 (French Open title, US Open final) if anyone had been officially counting at the time. By comparison, Nastase won the ’72 US Open and ’73 French, and as a result, was #1 in the first-ever rankings, which cemented him in popular memory to an extent.

Kodes is also the highest-ranked player on this list to have taken a career-long pass on one of the Slams; he never tried Australia. Nastase played it only once, losing in the round of 64 in 1982 when he was half a decade removed from his last semifinal in any Slam.

26. Michael Chang 5.19
27. Goran Ivanisevic 4.97

I absolutely love the fact that these two guys ended up next to each other. They’re incredibly similar in some ways, and complete opposites in others.

Chang and Ivanisevic were born within 5 months of each other, in September 1971 and February 1972, respectively. They both reached a career high ranking of #2 (behind Sampras in each case), and ended up with relatively similar career win totals (599 for Ivanisevic, 662 for Chang), and as you see, very similar Slam scores. Both of them had records of 1-3 in Grand Slam finals. Chang led in their head-to-head matchups by a slender 6-5 margin. And both of them are now coaches of top-10 players – in fact, their charges faced off in the 2014 US Open final, with Ivanisevic's pupil bettering Chang's on that occasion.

On the other hand… Ivanisevic is a 6-foot-4 Croatian who relied on a booming serve; he is the all-time leader in aces as recorded by the ATP (since 1991). Chang is a 5-foot-9 Chinese American who scampered all over the court and wore opponents down. All four of Ivanisevic’s appearances in Slam finals came at Wimbledon; he made only one semi at any of the other three events, compared to six total at the All England Club. Chang, meanwhile, never made it out of the Wimbledon quarterfinals, but made at least one final at each of the other three Slams.

So, naturally, their lone Slam victories were also similar, as they both came almost completely out of nowhere… and apart from that, they were utterly opposite. Chang won the 1989 French Open, beating the top 2 players in the world along the way; his win over #1 Ivan Lendl is particularly noteworthy because Chang spent much of the match battling cramps, at one point resorting to an underhand serve, but still rallied from two sets down for the win. He was 17 years, 3 months old at the time, making him the youngest men’s Grand Slam champion ever. The early success spurred him on to the aforementioned exemplary career, but he never did win another Slam.

Ivanisevic, meanwhile, seemed well on his way to strong contention for the mantle of “best player never to win a Slam,” having lost Wimbledon finals in 1992, ’94, and ’98, while at least making the quarterfinals of all three of the others, reaching #2 in the rankings, and winning over 20 titles. And then, in 2001, when he hadn’t won a title of any kind in 3 years and was ranked #125 in the world, Wimbledon gave him a wild card into the main draw, and he cut through Carlos Moya (ranked #22 at the time), Roddick (#33), Greg Rusedski (#40), and Marat Safin (#3) to reach the semis. Once there, he narrowly edged past hometown hero Tim Henman (#11) in five sets, then took another five to beat Patrick Rafter (#10) in the final, with the fifth set going to 9-7. He was the most experienced first-time Slam champion ever, having come up empty 48 times before finally nailing down his Wimbledon title; he was also the first wild card ever to capture a Slam. And after this astonishing triumph, he never hoisted another trophy in any singles event.

28. Stan Smith 4.91
29. Marat Safin 4.88
30. Yevgeny Kafelnikov 4.76
31. Patrick Rafter 4.48

We’re clearly into the realm of abbreviated brilliance now. Kafelnikov, Safin, and Rafter all spent time at #1, but it was less than 10 weeks in each case (Rafter managed the startling feat of holding the top spot for exactly one week late in the 1999 season). Smith would likely have been in a similar boat if the rankings had existed a year or two earlier; he won the US Open in 1971 and Wimbledon in ’72, then never made another Slam final.

32. Vitas Gerulaitis 4.30
33. Johan Kriek 4.16
34. Gustavo Kuerten 4.13

Caveats all over down here. Kriek’s two titles were in back-to-back Australians in the early ‘80s, when most of the best players still weren’t attending. Kuerten, meanwhile, won a trio of French Open titles, and never made it as far as the semis in any other Slam (including the other times he played Roland Garros). That makes him a good test case for how important you think titles are in comparison to any other noteworthy result, which is a topic we’ll revisit later.

35. Juan Carlos Ferrero 4.12
36. David Ferrer 3.90

Ferrer's appearance brings us to the answer to the last question I was hoping to answer in this initial post: Who is the best player never to win a Slam? The answer, at least as of now, is Ferrer. It’s “as of now” because, of course, someone could pass him (Tomas Berdych, for instance,  is #4 on the no-titles list), but also because Ferrer’s career is not yet over – and while he’s not likely to win a Slam at this point, he’s probably likelier than Goran Ivanisevic was in 2001.

Next up in Slam scores: An alternate set of weights for each round, along with a look at the differences in how both options evaluate players. That will be followed by breakdowns of the rankings Slam by Slam, a fuller listing of the best players without a Slam title, and a variety of other ways of looking at these results. Oh, and a post-Australian Open update once Melbourne finishes up in a couple of weeks. There's a lot we can do with these numbers, and I plan to do just about as much as I can think of.

No comments:

Post a Comment