Thursday, November 20, 2014

The Case for Djokovic as the GOAT

I have a list of the 25 greatest tennis player of the Open Era, which I update every few months or so, and on that is list, Novak Djokovic is currently sitting at the No. 9 spot, ahead of Stefan Edberg and not far behind Andre Agassi. However, there is still a case to be made that Djokovic is the greatest tennis player of all time based entirely on the numbers.

Before determining who the GOAT is, it is crucial to first define what the GOAT is. Some people consider it the person that at his peak can beat any player in tennis history head-to-head and would do so if given the chance. The problem with this definition is that whoever that person is depends on what surface is being used and which type of racket is being used.

The evolution of rackets and use of different surfaces in tennis is always a problem in determining the GOAT, which is why I prefer the following definition for GOAT. It is the singles player that dominated their own era the most and did so for the longest amount of time. That gives us two ways to measure a players dominance: degree and duration, and Djokovic is the leader in both categories, which is what makes him the GOAT.


No mathematical tricks here. Determining that the degree of Djokovic's dominance at its peak is higher than anyone else in tennis history is very clear. Just look at the 2011 season. From the 2010 Davis Cup final to the 2011 US Open, Djokovic went 66-2 with a 43-match winning streak. That year, Djokovic accumulated 13,630 ranking points, which is the most ever in a calendar year.

Of course, the ranking system has changed over the years, so that final stat hardly tells the whole story, but Djokovic also went 11-1 against Roger Federer and Rafael Nadal, who were each ranked ahead of him at the beginning of the year. On top of that, he beat Nadal, who is known as the "King of Clay," in straight sets in both of their matches on clay.

Although there are seasons, such as Rod Laver's in 1969, John McEnroe's in 1984, and Federer's in 2006 that compare to what Djokovic did in 2011, none of them had a stretch that was as dominant as Djokovic's 66-2 run.


This is where things get tricky, because two reasonable people could disagree on how to measure duration. It is certainly more complicated than length of career, because nobody dominates the tour in their first year, and few retire while at their peak. The duration also isn't as short as the number of weeks spent as the No. 1 player in the world, because that perspective treats the No. 2 as equally irrelevant as the No. 200 player in the world.

However, considering the No. 2 player in the world dominant would be wrong, because there is someone better than that player. That doesn't mean it is impossible to be a dominant No. 2 player in the world. The solution lies in playing around with the rankings.

The ATP World Tour does its rankings based on a 52-week system. That means that 52 weeks after a player earns his ranking points, he will lose those points, meaning the duration of ranking is always a 52-week period.

So to measure the duration of a player's dominance, we can play around with the duration of the ranking system. Here is what I did:

I changed the current rankings to 104 week rankings, which is simply the sum of year-end points from 2013 and year-end points from 2014. Since Djokovic got 2000 points from the Australian Open 2013 and 360 points in 2014, the Serb gets a total of 2360 points from that event in his 104-week rankings. That's how it works for all of a player's top 18 events.

Measuring Djokovic.

When I changed the rankings to 104 weeks, I found that Djokovic was still the world No. 1. I then checked the 156-week rankings by adding the 2012 year-end points and Djokovic was still the world No. 1. I kept adding 52 weeks to the ranking and every single time Djokovic was the world No. 1 all the way back to 2007. Then in 2006, Nadal surpassed Djokovic.

(I adjusted the points from pre-2009 year-end point totals so that they would be numerically comparable to the rankings numbers produced from 2009 to 2014.)

That means that Djokovic has been the best player in the world for the last eight years, and no player has been better than Djokovic in a time frame away from this season.

I decided to add one more requirement to the streak, which was that during the streak the player cannot end any season ranked lower than No. 4. The reason for that was that a player can't be the fifth best player for that year and still consider that part of his period of dominance.

That requirement also prevents a potential problem in the future. If Djokovic has just one or two more dominant years, it could drive his point total high enough that his streak of dominating the rankings would go back to 1973, when rankings began. Obviously, years where Djokovic didn't even play shouldn't be considered part of his era of dominance, so the final qualifier prevents that from happening.

Measuring the field.

For measuring another player, I started by finding the latest year that each player finished the year as the No. 1. Then, I started adding 52 weeks backwards in time to that ranking to measure how long their period of dominance had been.

The largest problem with this method of determining duration is simply that there were no points given out prior to 1990. Because of that, it is impossible to compare McEnroe, Ivan Lendl, and Jimmy Connors, who all would have done well in this category if there were a way to measure them.

Nadal and Federer each produced two different streaks, while every other player had just one.

This is the total number of years each player's dominance lasted.

1. Novak Djokovic 8 years (2007-2014)
2. Pete Sampras 7 years (1992-1998)
3. Roger Federer 6 years (2003-2007, 2009)
4. Rafael Nadal 5 years (2007-2010, 2013)
5. Jim Courier 2 years (1991-1992)
5. Lleyton Hewitt 2 years (2001-2002)
7. Stefan Edberg 1 year (1990)
7. Andre Agassi 1 year (1999)
7. Gustavo Kuerten 1 year (2000)
7. Andy Roddick 1 year (2003)

Realistically, Connors is likely the only one who would have been close to Djokovic in this list out of the guys who played before 1990. Connors one season that compares to Djokovic's 2011 is what he did in 1974. That year he went 93-6 with 15 titles, including three grand slam titles.

However, his longest winning streak that year was only 28 games and the other top players in the world that year were Laver, Borg, John Newcombe, Guillermo Vilas, Tom Okker, Stan Smith, Ken Rosewall, Ilie Nastase, and Arthur Ashe. He played all of those players just six times in his 99 total matches. Meanwhile, Djokovic played just Federer and Nadal 12 times in 68 matches. That doesn't even count all the other top 10 wins Djokovic had in that stretch.

I won't move Djokovic from No. 9 on my GOAT list, but that doesn't mean there can't be a statistical case to be made that Djokovic is the greatest tennis player of all time.

*I will post a more complete explanation of my dominance duration stat in a couple days, so that the math behind it can be seen and analyzed and maybe improved upon.


  1. Hello Jared Pine, does this mean that 2014-2006 (of nine years) according to your calculations, Rafa is the number one player? (just like Djokovic is for the 2007-2014 period of 8 years).. Because above, you say that adding back the year-end total points Djokovic is the number 1 player all the way back to 2007 but Nadal takes over if 2006 is included. Again, does that mean Nadal is the number 1 player in your parameters if we take the 9-year frame of 2006-2014?

    1. Over the period of 2006-14 Nadal has the most combined ranking points, but he doesn't get credit for those years, because it isn't consecutive like Djokovic's.

    2. Thanks, though I am more confused now by your use of "consecutive". So did Roger Federer not have the most points combined consecutively between 2003-2008? (Above you say 2003-2007) I am thinking it has to do with your requirement that you have to start counting from the last year that someone was ranked number 1? So if Nadal finishes number 1 in 2015, then he would take it as 2006-15 (nine years)? Just trying to understand..

    3. Potentially yes. However, Nadal would have to be No. 1 by a huge margin to make it consecutive. I was planning on writing another article just to explain the math in this, but I never got around to it, so sorry for the lack of clarity.

    4. No problem. It's a great read in any case and thanks for putting the time into doing it.

  2. Yes indeed does seem like some strange and not quite fully thought out logic and calculations here.