Second Serve Return Win %: Would You Believe Tomas Berdych Ranks First? [Stats]

Does this take into account number of matches played? Ferrer and Berd seem to enter everything and go quite deep.

March 9th, 2015 at 2:10 pm

There is a problem with these stats, as the piece on the ATP Page says, they do not include the Australian Open. Only the most important tournament played in these last two months.
So these numbers are skewed.
I commented it to Greg Sharko and he said that the numbers for Grand Slams are done in a different format and they were not available.
I suggested to him waiting for them instead of printing misleading stats, but he seems to think that the fact that he stated in the article that they were not included makes it all right.
I disagree. If I rule on a case by stating beforehand that I did not look at all the available evidence, the fact that I came clean doesn’t make the ruling fair. LOL
I suggest waiting to have all the numbers with us.

March 9th, 2015 at 8:57 pm

i agree yolita; i found the article and stats misleading. many fans and definitely players consider slams the most important events, yet they’re not included? what? seems weird.

March 9th, 2015 at 9:39 pm

You have more reliable stats on tennisabstract, for the previous year. It is here:

http://www.tennisabstract.com/cgi-bin/leaders.cgi

Just click on the heading of the column you are interested to get the ranking.

So, a few stats from there:

best returner: Nadal, Djokovic, Murray, Ferrer.
But the key stat here is the median rank of the opponent: for Rafa it is 44, for Novak 17…

This stat is very important, and I checked it for the last few year (since TAbs. exists). Here is the ranking: Djokovic 17, Murray 25, Federer 25, Nishikori 26,5, Klizan 31,5, Ferrer 33, Wawrinka 33, Berdych 33, Cilic 35 (I omited players with less than 20 matches), then a few well ranked players: Dimitrov 43, Nadal 44, Tsonga 45, Raonic 48.

The “median” ranking means that the player played the same number of matches with players ranked below and above that number.

The average ranking of the opponent was the following:

Djokovic 24,5; Federer 37,2; Nishikori 39,1; Murray 40,4; Wawrinka 42,1; Dolgopolov 45,2; Ferrer 48,2; Tsonga 49,4; Nadal 50,6; Gasquet 51,5; Berdych 53,5.

Those stats are very, very interesting, and are easy to check from 2012. Since then, Novak has always had the toughest competition (although he mostly didn’t have to play the no 1 ranked player…).

March 9th, 2015 at 10:29 pm

Yolita, they include AO, otherwise Djoko wouldn’t have 15 matches (3 in Doha, 7 in AO and 5 in Dubai), Federer wouldn’t have 12 (4 in Brisbane, 3 in AO and 5 in Dubai). It includes AO.

March 9th, 2015 at 10:29 pm

Another interesting stat is the number of seconds by point.

The ranking in 2012 (but only the top 10 players):

1. Nadal, 50,7;
2. Murray 45,3;
3. Berdych 45,3;
4. Djokovic 45,2;
5. Nishikori 43,7;
6. Cilic 43,7;
7. Ferrer 43,5
8. Wawrinka 41,6
9. Raonic 40,7
10. Federer 39,6

and now:

1. Nadal, 46,5
2. Murray 42,9
3. Djokovic 42,2
4. Ferrer 42,1
5. Berdych 41,5
6. Cilic 40,9
7. Nishikori 40,7;
8. Raonic 40,2
9. Wawrinka 39,3
10. Federer 37,8

March 9th, 2015 at 10:40 pm

I suggest also to check to “domination ratio”, a very important stat. There are others, interesting too: mn per set, % of sets won, % of games won, etc.

March 9th, 2015 at 10:48 pm

wow mat4, “But the key stat here is the median rank of the opponent: for Rafa it is 44, for Novak 17…”

amazing! but it makes sense.

i posted recently how many matches nole played versus top 10 over his 4 best years on tour versus the same stat for fed and rafa over their 4 best, and nole had about 20 more top 10 wins.

March 9th, 2015 at 11:23 pm

daniel, perhaps now but initially they didn’t; it was written as such on the ATP site.

March 9th, 2015 at 11:25 pm

That made much more sense Mat4. Cheers.

March 10th, 2015 at 2:43 am

It is easy to check matches against top ten opposition, and we see that Novak has played a lot against top 10. It is also interesting to note that his results, since 2012, have steadily improved — after that exceptional 2011 season, 2012 was below average, while 2013 and 2014 are at the same level (about 80% of wins).

I tried to quantified the impact of Boris Becker, estimating that his results in the last 52 weeks could be representative enough. But…

Comparing the last 52 weeks with 2013 (when Novak played quite well), I had to put it all in context first: Novak, in 2013, had a median opponent ranked 22 and an average opponent ranked 37,4 while the number for the last 52 weeks are 17 and 22,4. It made a big difference, and Novak’s stats are a bit worse now than in 2013.

So I decided to focus only on his matches against top 20 players. Novak played 41 matches against top20 in 2013, 30 in the last 52 weeks; median/average ranking were 7 and 8,9 in 2013, 8 and 8,4 respectively in the last year. Let’s see here:

Hold rate: 84,9 (2013), 85,8 (last 52)
Aces: 8,5% (2013), 7,6% (last 52)
Points per Serve game lost: 1,9 (2013), 2,0 (last 52)
First serve in: 65,9 (2013), 66 (last 52)
First serve won: 73 (2013), 73,8 (last 52)
Match winning %: 82,9 (2013), 80 (last 52)

Very difficult to notice anything in particular. So I changed the data, focusing on hard and taking the results against top 50 opposition to have enough matches.

Number of matches: 47 (2013), 30 (last 52)
Mean opponent ranking: 10 (2013), 11 (last 52)
Average opponent ranking: 18 (2013), 16,4 (last 52)

Hold rate: 84,8 (2013), 87,5 (last 52)
Aces: 8,1% (2013), 7,6% (last 52)
Points per Serve game lost: 1,8 (2013), 1,9 (last 52)
First serve in: 65,3 (2013), 65,7 (last 52)
First serve won: 74,9 (2013), 75,2 (last 52)
Match winning %: 89,4 (2013), 86,7 (last 52)

So, stats… while they can show a lot, don’t show everything. The sample, for one, is too small. Then, here we don’t have a few key numbers.

March 10th, 2015 at 5:35 am

But what can we see in those stats?

We can check a few myths about Novak (Fed and Novak are in the focus of my tennis explorations).

First, about the “economy” of play. We can see that Novak’s game is physical, and the toll on his body is important. Let’s compare Rafa’s, Fed’s and Novak’s numbers.

First, the key distorting factor:

Median ranked opponent:

Novak: 17 (overall), 15 (hard), 21,5 (clay)
Rafa: 44 (overall), 44(hard), 30 (clay)
Fed: 25 (overall), 323,5 (hard), 29,5 (clay)

Avg ranked opponent:

Novak: 24.4 (overall), 23.1 (hard), 25.1 (clay)
Rafa: 50.6 (overall), 52.0 (hard), 43.7 (clay)
Fed: 37.2 (overall), 36.7 (hard), 36.8 (clay)

Min. per set:

Novak: 41,3 (overall), 40,6 (hard), 41.6 (clay)
Rafa: 43,8 (overall), 42,5 (hard), 45,2 (clay)
Fed: 34,8 (overall), 35 (hard), 33 (clay)

Novak is ranked 39 in the top 50, Fed is among the very best, and the avg time is 39 mn.

I checked also the second per point stat, but the problem is the time spent BETWEEN points. In that category, Rafa is ranked first, Andy second, Novak third (most time per point).

What about the number of points played?

Avg. points per match:

Djokovic: 152
Federer: 149
Nadal: 144

OK, you’ll say that Novak wins a lot more on opponent serve but… it is not the case. The domination ratio is about the same for all three players (Fed 1,37, Rafa and Novak 1,35), and when we see that they lose from 1,9 to 2,1 points on their serve… Anyway, stats are easy to check there too: Novak and Rafa play 6,6 points per return game, Fed plays 6,8. They win between 2,7 and 2,9 points per return game. The difference quite irrelevant.

March 10th, 2015 at 6:07 am

Finally, let’s look at the deciders. Novak spoke of that very often, so, is a Becker impact on key moments noticeable?

Deciders (name, number of matches, %, median, avg)

Djokovic 11 11-0 100.0% 22.0 23.5
Federer 18 13-5 72.2% 22.0 23.2
Nadal 8 5-3 62.5% 14.0 19.5
Murray 14 11-3 78.6% 21.5 26.9
Nishikori 19 17-2 89.5% 10.0 28.9
Raonic 16 9-7 56.3% 37.0 37.4
Wawrinka 14 5-9 35.7% 21.0 37.5
Ferrer 22 11-11 50.0% 22.5 37.7
Berdych 15 8-7 53.3% 38.0 46.6
Cilic 15 9-6 60.0% 31.0 45.2

The results for Novak are very good, but it seems to be a statistical aberration. In 2012 he vas at 16-3, in 2013 at 15-6, in 2014 at 15-2.

But here, we can see why Wawrinka, Berdych, Raonic or Ferrer are not better ranked.

March 10th, 2015 at 6:40 am

Just hope that somebody will notice these posts.

March 10th, 2015 at 6:42 am

Finally, Novak stats for 2011, career and last 52.
match Tiebreak Ace% 1stIn 1st% 2nd% RPW DR

2011 70-6 (92%) 9-10 (47%) 6.3% 65.3% 74.1% 55.6% 44.6% 1.38

Career 653-154 (81%) 182-109 (63%) 7.2% 64.5% 73.3% 55.0% 41.9% 1.26

Last 52 62-8 (89%) 15-11 (58%) 8.0% 66.3% 74.8% 56.9% 42.7% 1.37

March 10th, 2015 at 6:59 am

Just to mention that everything is the work of Jeff Sackman from Tennisabstract.

March 10th, 2015 at 7:00 am

@mat4: I absolutely loved your posts. It has been obvious to many of us that Novak has had tougher competition than his main rivals. It is very nice to have that feeling supported by numbers.

I had only counted top 10 opponents over the past 4 years. And Novak has played many more than Roger/Rafa/Andy, both in absolute and in relative terms, as percentage of their total matches. Nole has climbed the mountain through the most difficult path and he sits alone at the very top.

All these numbers just add to Novak’s stature. :-)

@Daniel: Thank you for the observation. In the original Sharko post the Australian Open was not included. He himself said that in the first paragraph. I had a discussion with him on Twitter…I’m glad the Australian Open numbers have been added.

March 10th, 2015 at 3:50 pm

@Yolita:

It seems that Novak’s stats from the AO were not included in the general data, while they are present in his own stats.

Anyway, thank you. I was afraid no one will read these posts.

March 10th, 2015 at 4:15 pm