Monday, September 14, 2015

Record vs. Rivals: A Tennis Stat Through the Ages

Roger Federer is the most accomplished male tennis player of all time, absolutely dominating tennis for over six years (making the finals in 22 of 27 Grand Slams, winning 16 of them), and extending his excellence for approximately 13 years. Federer has won the most Grand Slam titles by a relatively wide margin. He was in the most Grand Slam finals by an even wider margin. He was in the most Grand Slam semi-finals. Federer also won on every tennis surface and in every major tournament, completing the career Grand Slam. Federer held the #1 rank for the longest period of time. He has won the most ATP championship tournaments and has even won a Davis Cup championship now. (He has not won an Olympic gold medal, merely silver, but that is nit-picking.) Federer has won tournaments around the world, among the most in history, playing for over fifteen years and yet he still maintains one of the highest winning percentages of all time and one of the highest winning percentages against top-10 opponents ever. His game is basically flawless; he mastered everything from the serve, the return, the forehand, the backhand, the baseline, and the net, with good speed, strength, agility, and stamina. His play seems fluid and regal as if he was born to be the greatest tennis player of all time (or more accurately the greatest male tennis player of the open era). And yet there is a legitimate reason to keep that figurative title from him, and it's a rather compelling reason: Rafael Nadal.

Nadal, who was a clay court specialist, with fewer wins (major and otherwise) than Federer, who largely saved himself for the majors, did not have a flawless game, and started succeeding later and appears to be finishing success earlier, and did not win as much in between, and won most of his major tournaments on one surface, still beat Federer regularly. It is not unusual, nor is it particularly telling that one person has an edge over another person in any given match-up. If Joe Nobody were to beat Federer with regularity, and just merely takes advantage of a match-up disparity (or he somehow figured Federer out), it would not be generally informative, even in such an individual sport. But since Nadal happens to be the second winning-est player of all time, playing in the same era as Federer and beating him most of the time, it does call into question whether Federer is the best player of his own time, let alone of all time.

So, to what extent does Nadal dominate Federer? Well, let's start with the overall record: 23-10 Nadal. Federer is 2-1 on grass; Nadal is 9-6 on hard courts; and Nadal is 13-2 on clay (meaning on non-clay, Nadal is 10-8 against Federer. Now, let's get to Grand Slams: Nadal is 9-2 against Federer in Grand Slams: Federer is 2-1 at Wimbledon, Nadal is 5-0 at the French, and Nadal is 3-0 at the Australian Open (meaning 4-2 on non-clay). Meaning, even on non-clay, Nadal has the better of the tournament match-ups. But, maybe Nadal was beating up on Federer as Federer was moving away from his prime? Maybe prime Nadal was crushing older Federer? There's some of that... but looking just at Grand Slams, the last time Federer beat Nadal at a Grand Slam was Wimbledon in 2007, meaning Federer was about 26 years old and Nadal was about 21. It's hard to discount Federer a year later as past his prime, even if Nadal was approaching his. In 2008, with Federer about 27 and Nadal about 22, Nadal was 4-0 against Federer, beating Federer at Wimbledon as well as the French Open. In 2009, they were 1-1 with Nadal winning their one Grand Slam match-up in Australia.

But, what about that 2007 victory, doesn't that indicate that prime Federer would beat prime Nadal at least at Wimbledon? Well, it's a stretch to believe that 21 year old Nadal who had never won a non-clay tournament was at his peak. But maybe Federer's peak had faded? When exactly was Federer's peak? That's a tough question. First, his personal peak (meaning his greatest tennis) may differ from his accomplishment peak, but certainly in 2004- he won 3 Grand Slams, in 2005, he won 2 Grand Slams, in 2006, he made it to all 4 Grand Slam Finals winning 3 (possibly his peak year), in 2007, he mirrored that accomplishment (possibly his peak year), in 2008, he made it to 3 finals, losing two of them to Nadal. So maybe he was done? But Nadal was then somewhat hurt the following year, and Federer proceeded to make it to all 4 finals again, winning two (and losing one to Nadal). Nadal's peak year did not come for another two years after that, as Federer was still excellent, but no longer maintaining his legendarily consistent Grand Slam performances (so not at his personal accomplishment peak). But, during the 2 years their personal accomplishment peaks coincided, Nadal was fairly dominant against Federer on all surfaces (winning in 3 different Grand Slam finals.)

Perhaps more importantly, which competitors were most dominant against their contemporary elite rivals? This is particularly important because the value of each championship may be somewhat lesser or greater depending on the level of competition. How do we procedurally make such a determination? An early hurdle in such an endeavor is that not only do the greats play in different eras, but even the ones that do play in the same era do not necessarily have overlapping prime years. For example, Connors is substantially older than McEnroe, so in 1985, while Connors was still talented enough to get by good tennis players, he may have diminished more than McEnroe's prime-level game. Thus, if Connors was no longer the mighty Connors we once knew, a McEnroe victory would not be particularly demonstrative of McEnroe's superiority. But age should not define primes because Nadal and Djokovic are the same age (one year apart), but Nadal ripened as a player much earlier than Djokovic did, so comparing their records in 2006 would be of very limited value. And conversely, it appears that Djokovic's prime is going to extend past Nadal's, which means we may not want to compare Djokovic's future years to Nadal's future years either.

First, we devise a method to determine the elite players of all time that is at least largely objective and preferably convenient (even if it is arbitrary). The top 10 ranked players are not historically elite level players. We could create a standard whereby "elite players" won a certain number of Grand Slam tournaments victories, making the victory total sufficiently low to avoid really tiny sample sizes (that would mean 2-3 Grand Slam victories). But, even if we did that, we would would still have to determine a prime for those elite players (who may have won one tournament in 1991 and then another in 1997). The greatest players of all time would have extended periods of accomplishment, particularly Grand Slam achievements (which are critical events, sufficiently plentiful, and well documented). So, if Djokovic won one rogue Grand Slam tournament in 2008 or if Murray did so in 2015, it does not necessarily count as his prime. If the standard for excellence is created with some consideration, this standard could serve both to eliminate the individuals who are not elite all-time greats, and to determine the years in which those greats were in their prime.

An effective method of establishing the "Prime" of each great tennis player works as follows: the player makes it to the finals in 3 Grand Slam tournaments within a two year period (at least 3 finals in 8 Grand Slam tournaments). While winning Grand Slams is a greater accomplishment, getting to the finals with regularity would indicate the player is beating low level competition and high level competition repeatedly over a given period of time. If that player loses to a rival in his prime or someone playing superb tennis, that does not mean they are not in their prime. But, over a two year period, if that historically great player continues to play at a high level, presumably his seeding should improve and he should have increased opportunities to make it to the finals. If he never made it to the finals 3 times in a two year period, then I assert that he is not a historically elite player in his prime. (Obviously, I note that Grand Slam achievement is not the only method of determining a prime. And I note that this method has flaws, e.g., an individual may be playing incredible tennis, possibly winning many peripheral tournaments for 2+ straight years, winning two titles, losing to 3-5 titans in semi-finals (or quarterfinals) of Grand Slams and/or was hurting/injured in 1-2 of those tournaments, and this player would not be considered an all time great in their prime, even though they may have been a top player at the time (see Kriek). [Personally, I think it's informative that a person cannot sustain excellence long enough to make 3 finals in two years. You can also effectively argue that Wawrinka has been better the last two years than Federer has - with two Grand Slam victories, rather than Federer's 3 Grand Slam Finals losses, but can you argue that based upon those two victories Wawrinka is an all time great? If he retires tomorrow, which I hope he does not, he would be a guy who had two hot years when he was one of the best players in tennis, rather than an all time great. I feel somewhat bad omitting Kuerten because he never had a "prime", Kuerten who won 3 French Open titles, making him the the third most accomplished French Open player of the Open era and the only winner of 3 Grand Slams in the open era left all the list.) Similarly, a player could have played far superior tennis in an earlier stage of his career, and yet got a few breaks and played lesser opponents and made the finals and lost 3 times in two years. (Michael Chang, my favorite tennis player, won his one Grand Slam in his teens, well before his "Prime".] But, no criteria would be without fault, and any criteria would be arbitrary. I think this one's logic has merit and it can be relatively easily universally applied objectively.)

This Prime determination method mostly eliminates the very good players and leaves the legends. Notably, between 1974-2015, there were only 19 such players so far. For those individual players, I would also include as a "Supplemental Prime", any period within 5 tournaments in which these players won two Grand Slam tournaments. This is an All Time great player who is dominating a time, or an event ("a hot streak"). (It is notable that placing these arbitrary standards upon peaks is not only useful for the stat, but also for evaluating the competitiveness of particular eras of tennis. But, that is an issue for another time.)

Working in roughly chronological order (note that Agassi, Connors, and Federer had multiple peaks, Edberg and Becker had early hot streaks, and as mentioned, Chang did not win his one Grand Slam during his Prime- he won when he was a teenager- 5 years before his prime):

Primes of the Elite Men's Tennis Players of the Open Era

Connors- 1974-1978, 1982-1984

Borg- 1975-1981

Vilas- 1977-1979

McEnroe- 1979-1985

Wilander- 1982-1988

Lendl- 1982-1991

Becker- 1985-1986, 1988-1991

Edberg- 1985-1986, 1988-1993

Cash- 1987-1988

Agassi- 1990-1992, 1994-1995, 1999-2001

Courier- 1991-1993

Sampras- 1992-2001

Chang- 1995-1996

Ferrero- 2002-2003

Roddick- 2003-2005

Federer- 2003-2011, 2014-2015

Nadal- 2005-2014

Djokovic- 2010-2016

Murray- 2011-2013; 2015-2016

There are 19 players on the list so far with a mean 6+ years of Prime during the 42 years in question. Notable players that did not make this list, but might make someone else's list:

Wawrinka (3-0 in Grand Slam Finals so far),

Kuerten (3-0 in Grand Slam Finals),

Rafter (2-2 in Grand Slam Finals),

Hewitt (2-2 in Grand Slam Finals),

Safin (2-2 in Grand Slam Finals),

Kafelnikov (2-1 in Grand Slam Finals),

Kriek (2-0 in Grand Slam Finals),

Ivanisevich (1-3 in Grand Slam Finals),

Stich (1-2 in Grand Slam Finals), and

Gerulaitis (1-2 in Grand Slam Finals).

So these Primes (1) established the elite historical players and simultaneously (2) established when those players were at their peaks. The next step is to establish their records against each of the other players during the Prime of the other elite players.

 Head to Head during Joint Primes in Majors

Borg v. Connors 1975-1978 GS: 3-2 Connors, T: 5-4 Connors: Overall 15-8 Borg, GS 5-3 Borg

Borg v. Vilas 1977-1979, GS: 1-0 Borg, T: ?: Overall 17-5 Borg, GS 3-0 Borg

Borg v. McEnroe 1979-1981, G: 3-1 McEnroe, T: 7-6 Borg: Overall 7-7, GS 3-1 McEnroe

Connors v. McEnroe 1982-1984, G: 3-1 McEnroe, T: 11-3 McEnroe: Overall 20-14 McEnroe, GS 6-3 McEnroe

Connors v. Wilander 1982-1984, G: 0-0, T: 2-0 Wilander: Overall 5-0 Wilander, GS 0-0

Connors v. Lendl 1982-1984, G: Connors 3-0, T: 6-5 Lendl: Overall 22-13 Lendl, GS 4-3 Lendl

Connors v. Vilas 1977-1978, G: 1-0 Vilas, T: ?: Overall 3-3, GS 2-1 Connors

McEnroe v. Wilander 1982-1985, G:, 3-1 Wilander, T: 4-4: Overall 7-6 McEnroe, GS 3-2 Wilander

McEnroe v. Lendl 1982-1985, G: 3-2 Lendl, T: 11-9 McEnroe: Overall 21-15 Lendl, GS 7-3 Lendl

McEnroe v. Edberg 1985, G: 0-0, T: 1-0 McEnroe: Overall 7-6 McEnroe, GS 2-1 Edberg

McEnroe v. Becker 1985, G: 0-0, T: 1-0 McEnroe: Overall 7-2 Becker, GS 1-0 McEnroe

Cash v. Wilander 1987-1988, G: 1-1, T: 2-1 Wilander: Overall 5-3 Cash, GS 4-1 Cash

Cash v. Lendl 1987-1988, G: 3-0 Cash, T: 3-1 Cash: Overall 5-3 Lendl, GS 3-3

Cash v. Edberg 1988, G: 0-0, T: 0-0: Overall 8-2 Edberg, GS 3-0 Edberg

Cash v. Becker 1988, G: 1-0 Becker, T: 1-0 Becker: Overall 3-1 Becker, GS 1-0 Becker

Lendl v. Wilander 1982-1988, G: 5-4 Lendl, T: 11-6 Lendl: Overall 15-7 Lendl, GS 5-4 Lendl

Lendl v. Edberg 1985-1986, 1988-1991, G: 3-3, T: 11-10 Lendl: Overall 14-13 Edberg, GS 5-4 Edberg

Lendl v. Becker 1985-1986, 1988-1991, G: Becker 5-0, T: 10-7 Becker: Overall 11-10 Lendl, GS 5-1 Becker

Lendl v. Agassi 1990-1991, G: 0-0, T: 0-0: Overall 6-2 Lendl, GS 2-0 Lendl

Lendl v. Courier 1991, G: 0-0, T: 1-0 Lendl: Overall 4-0 Lendl, GS 1-0 Lendl

Wilander v. Edberg 1985-1986, 1988, G: 1-1, T: 6-3 Wilander: Overall 11-9 Wilander, GS 3-2 Wilander

Wilander v. Becker 1985-1986, 1988, G: 1-0 Wilander, T: 5-1 Becker: Overall 7-3 Becker, GS 3-0 Wilander

Edberg v. Becker 1985-1986, 1988-1991, G: 3-1 Edberg, T: 15-6 Becker: Overall 25-10 Becker, GS 3-1 Edberg

Edberg v. Agassi 1990-1992, G: 0-0, T: 3-2 Agassi: Overall 6-3 Agassi, GS 1-0 Agassi

Edberg v. Courier 1991-1993, G: 4-2 Courier, T: 4-2 Courier: Overall 6-4 Courier, GS 4-2 Courier

Becker v. Agassi 1990-1991, G: 2-0 Agassi, T: 4-1 Agassi: Overall 10-4 Agassi, GS 4-1 Agassi

Becker v. Courier 1991, G: 0-0, T: 1-0 Becker: Overall 6-1 Becker, GS 0-0

Agassi v. Courier 1991-1992, G: 3-0 Courier, T: 5-0 Courier: Overall Courier 7-5, GS 4-2 Courier

Agassi v. Chang 1995, G: 0-0, T: 2-1 Agassi: Overall 15-7 Agassi, GS 3-2 Agassi

Sampras v. Edberg 1992-1993, G: 2-0 Edberg, T: 3-3: Overall 8-6 Sampras, GS 2-0 Edberg

Sampras v. Agassi 1992, 1994-1995, 1999-2001, G: 3-3, T: 10-10: Overall 20-14, GS 6-3 Sampras

Sampras v. Courier 1992, G: 1-0 Sampras, T: 2-1 Sampras: Overall 16-4 Sampras, GS 6-2 Sampras

Sampras v. Chang 1995-1996, G: 2-0 Sampras, T: 4-1 Sampras: Overall 12-8 Sampras, GS 4-1 Sampras

Federer v. Nadal 2005-2011, G: 8-2 Nadal, T: 16-9 Nadal: Overall 23-10 Nadal, GS 9-2 Nadal

Federer v. Djokovic 2010-2011, G: 3-4 Djokovic, T: 5-5: Overall 21-21, GS 8-7 Djokovic

Federer v. Murray 2011, G: 0-0, T: 0-0: Overall 13-11 Federer, GS 5-1 Federer

Federer v. Roddick 2003-2005, G: 3-0 Federer, T: 7-1 Federer: Overall 21-3, GS 8-0 Federer

Nadal v. Djokovic 2010-2014, G: 5-3 Nadal, T: 12-9 Djokovic: Overall 23-21 Nadal, GS 9-4 Nadal

Nadal v. Murray 2011-2013, G: 3-0 Nadal, T: 4-2 Nadal: Overall 15-6 Nadal, GS 7-2 Nadal

Nadal v. Roddick 2005, G: 0-0, T: 0-0: Overall 7-3 Nadal, GS 1-1

Djokovic v. Murray 2011-2013, G: 3-2 Djokovic, T: 7-5 Djokovic: Overall 19-8 Djokovic, GS 5-2 Djokovic

Ferrero v. Federer 2003, G: 0-0, T: 2-1 Federer: Overall 10-3 Federer, GS 3-1 Federer

Ferrero v. Roddick 2003, G: 1-0 Roddick, T: 1-0 Roddick: Overall 5-0 Roddick, GS 1-0 Roddick

And FYI: for players with only a Supplemental Prime (2 Grand Slam titles within 5 Grand Slams)- only a hot streak:

Hewitt v. Sampras 2001, G: 1-0 Hewitt, T: 2-0 Hewitt: Overall 5-4 Hewitt, GS 1-1

Hewitt v. Agassi 2001, G: 0-0, T: 1-1: Overall 5-4 Hewitt, GS 1-0 Agassi

Rafter v. Sampras 1997-1998, G: 1-0 Rafter, T: 5-2 Sampras: Overall 12-4 Sampras, GS 2-1 Sampras

Bruguera v. Courier 1993, G: 1-1, T: ?: Overall 4-2 Courier, GS 2-1 Bruguera

Bruguera v. Agassi 1994, G: 0-0, T: ?: Overall 7-2 Agassi, GS 0-0

Brugera v. Edberg 1993, G: 0-0, T: ?: Overall 6-2 Edberg, GS 1-0 Bruguera

Bruguera v. Sampras 1993-1994, G: 1-0 Bruguera, T: ?: Overall 2-2, GS 1-1

Kriek v. Borg 1981, G: 0-0, T: 0-0: Overall 2-1 Borg, GS 1-0 Borg

Kriek v. McEnroe 1981-1982, G: 2-0 McEnroe, T: 4-1 McEnroe: Overall 12-5 McEnroe, GS 2-0 McEnroe

Kriek v. Lendl 1982, G: 0-0, T: 2-0 Lendl: Overall 10-0 Lendl, GS 3-0 Lendl

Kriek v. Wilander 1982, G: 0-0, T: 0-0: Overall 5-0 Wilander, GS 3-0

Kriek v. Connors 1982, G: 0-0, T: 1-0 Connors: Overall 7-2, GS 1-0

Kuerten v. Sampras 2000-2001, G: 0-0, T: 1-1: Overall 2-1 Sampras, GS 0-0

Kuerten v. Agassi 2000-2001, G: 0-0, T: 2-2: Overall 7-4 Agassi, GS 1-0 Agassi

I included additional data in the charts for context of the competition between each two players to be somewhat more thorough and to recognize the disparity at other points in these players' careers. (A fun side note is that while Sampras was evidently more accomplished than Agassi and got the better of Agassi over the course of their careers, when they were both at their peaks (Agassi's Prime was sporadic), Agassi legitimately challenged Sampras, winning 3 of 6 Grand Slam matches.) Please excuse any errors in reporting or calculation; I did the best I could when I compiled the data; this is not my primary profession, and I did it mostly under difficult circumstances. Finally, whereas some of the modern numbers are subject to change including Federer's, Nadal's, or Djokovic's, if either makes one more finals in the near future, these are the current tallies in the records vs. rivals or the RVR.

 RVR

Nadal: 16-5

Djokovic: 12-8

Courier: 7-3

Edberg: 11-9

Wilander: 10-8

Cash: 4-2

McEnroe: 9-8

Becker: 7-6

Connors: 7-6

Sampras: 6-5

Vilas: 1-1

Agassi: 5-6

Ferrero: 0-1

Borg: 4-6

Roddick: 1-3

Chang: 0-2

Murray: 2-5

Federer: 6-14

Lendl: 11-20

 Interestingly, most of these historically elite players pretty much leveled out against their peers (within 2 games of .500) with a few notable exceptions. Lendl played his rivals a lot (far more than any other player), which may speak to his consistence beating the lesser players and/or to the sheer volume of historically competitive players in his extended and hyper-competitive era(s), and Lendl regularly got his behind kicked by these elite players (even though he won many battles). A few players had similarly difficult times with elite competition. A few other players had nice runs against their historic rivals. But perhaps the most noteworthy stat is Nadal's. At 16-5, he has by far the best record against his rivals in their peaks. Nadal's dominance is most demonstrable on clay, but he had very good records against these greatest players ever in their peaks on other surfaces too. And while Nadal's record versus Federer was 8-2 during this period, it was 8-3 against everyone else, meaning he just eviscerated the best players ever. And outside of Lendl, Nadal had the most of these matches so far, meaning he has the second highest sample size. This may mean Nadal raised his game dramatically when faced with the best opponents. Or perhaps more accurately (considering Nadal's winning percentage was one of the best at well over 80% of his matches in general and his winning percentage against historically great players in their primes was over 75%), Nadal beat the best players ever almost as often, if not as easily, as he beat standard professional tennis players, which is extraordinary. Nadal beat Federer, Djokovic and company like they were 3rd round opponents, winning 14 championships along the way. And as for Federer, if you subtract his record versus Nadal, he was an acceptable 4-6 versus his rivals in their prime. If you include Nadal (as you should), Federer was 6-10 against his rivals in their mutual primes.

Given Federer's decent record against other historically great competitors, we probably should not read too much of this deficiency into Federer. Rather, the stat is a statement of the excellence and dominance of Nadal, even among the greatest players of all time in their primes. For that reason, referring to Federer as the greatest men's tennis player of the open era, even with his myriad accomplishments and acclaim, is called into question. Nadal who won only 3 less Grand Slams, but absolutely crushed Federer (and everyone else) in their primes, may be the best player of the Federer era. But, as for who is actually best (factoring in total Grand Slams, record versus rivals, array of important wins, wins on all types of surfaces, number of weeks at the number one spot, total tournament wins, winning percentage, winning percentage against top 10 opponents, caliber of competition, etc.), it still might be Federer, right? I'll leave that decision to you... or to me in another more subjective blog post.

No comments:

Post a Comment