Peak level of play (Federer, Nadal, Djokovic & Co.)

Discussion in 'Former Pro Player Talk' started by krosero, Jan 5, 2013.

  1. NatF

    NatF Talk Tennis Guru

    Joined:
    May 10, 2012
    Messages:
    27,572
    Location:
    Cretaceous
    Yes unfortunately we never could get that subforum.

    I would say serve is more vitally important to a top tennis player in general. You have a head start in point construction that you won't have in a return game. No player can afford to simply play loosely on their own service games because they're not in control of what the opponent bangs down and will inherently be at a disadvantage most of the time on return. Also because you can get free points on serve but can't on return - except DF's.
     
    Meles likes this.
  2. -NN-

    -NN- Hall of Fame

    Joined:
    Jan 10, 2017
    Messages:
    4,500

    Maybe this really depends on the conditions and a bias for the service game is justified in faster conditions, and for the return game in slower conditions. This comes with some assumptions, namely that those with very good return game stats tend to be better baseliners than those with very good serve game stats. Obviously some players are very good at both, but Nadal and Djokovic have had the clear edge on Roger as soon as the conditions get slower than medium, and they both have consistently better return game stats than Roger. At some stage, it becomes hard to gain percentage points on service games. Federer can be a substantially better server than Nadal or Djokovic but only be ahead by a few percentage points on service games held. At that point, I guess there are more gains to be made on return. There's a sort of threshold where diminishing returns occur for the service game at the very top level of the game, and nailing the ground-game and the returns is massively more important. Just a thought.
     
    Last edited: Feb 6, 2017
  3. NatF

    NatF Talk Tennis Guru

    Joined:
    May 10, 2012
    Messages:
    27,572
    Location:
    Cretaceous
    I will have to investigate these return game stats on slower courts - perhaps just look at runs where the final was reached. Djokovic will not doubt be the best of the 3, Nadal will rule on clay. Purely out of interest it's besides the point really.

    I would think that a small difference in service game numbers is more significant than a small difference on return. As long as you hold serve you only need one break to secure a set. You will more rarely win a set on any surface holding serve just once.

    Those diminishing returns will also be relative to who you're playing. It also depends when those points come.
     
  4. Gary Duane

    Gary Duane G.O.A.T.

    Joined:
    Jun 5, 2014
    Messages:
    10,876
    I never bothered to post my own data with my own formula, which actually was very simple. DR states a very important principle, that a dominant serve is hugely important because it always a player to "coast". This results in what I call the "Sampras effect", Sampras easily being the best example from the last 25 years or so, but it applies as much to Federer and even to Nadal on clay when he was fully GOATING on clay. What I did was this: I created a column for weighted game%. It doesn't matter what I use for a neutral level.

    I arbitrarily picked 88% of service games won as the neutral point, so if a guy wins 88% of his service games, his WTG% remains unchanged. 92% he gets an extra 1%, an extra 2% at 96%. 80% loses 2%.

    Someone better in math could come up with something more nuanced, but what it means is this:

    80/44 or 62% is only worth 60%. 88/32 or 60% remains 60%. 96/30 or 58% is worth 60%.

    Of course, 80/44 is nearly impossible, but so is 96/30. However, Fed got to close to 95/35 on grass in 2004. Confirmation from TA that shows Roger won 57.4% of all points, which is not human. But he did it. That's the gold standard on grass. Using this weighted number, Nadal edges that number out by a hair, on clay, in 2012. Without the weighting, Nadal wins by a couple %, a bit more.

    In other words, weighting actually brings clay and grass rather well into sync. It also rewards great servers but does not do what DR does, which rewards servers so hugely that a guy like Raonic will outscore a guy like Hewitt.

    Hewitt loses about 1/2% in his slam year - he only won around 87% of his service games, not bad, but a bit low.

    Fed gains about 1.5% in his best year, serving at 95%.

    Fed's biggest gain is in 2007, serving at whopping 97% on grass (you get such high % figures because of so few matches), but his return fell to around 24%. That's still winning over 61% of games but goes to close to 63.5% weighted.

    Most people will not realize that Nadal won 95% of his service games on grass in 2008.

    Here are the top 20 on grass - I have not double-checked every year with TA, but usually the ATP figures on grass are pretty good. I didn't post this data last year after the ATP screwed up everything because I was so disgusted and started to distrust their figures for every year, on every surface:

    Federer
    Federer
    Sampras
    Federer
    Federer
    Djokovic
    Djokovic
    Sampras
    Murray
    Federer
    Hewitt
    Nadal
    Sampras
    Murray
    Sampras
    Sampras
    Federer
    Nadal
    Federer
    Djokovic

    This simply reflects that idea that Fed and Sampras were the best on grass. Djokovic, Murray, Nadal and Hewitt are rather competitive, not surprising. Guys who peaked before 1991 don't show up because I have not data for their peaks.

    @Meles @-NN-
     
    Meles and -NN- like this.
  5. -NN-

    -NN- Hall of Fame

    Joined:
    Jan 10, 2017
    Messages:
    4,500
    @Gary Duane

    It is very interesting and I'd like to hear more views on this from others. @pc1 seems quite dismissive of such claims. Check further back in this thread if ya wanna see.
     
  6. Meles

    Meles Talk Tennis Guru

    Joined:
    May 23, 2014
    Messages:
    29,424
    Location:
    Baby Zeballos TV
    Serve game is vitally important in slams as said players usually get through the draw more efficiently. At some point with return you have issues. Roddick would be a prime example. His poor returning put him under duress. Sampras probably had the minimal return game for a great player. It could be argued that Murray has the greatest return game and we see what that's done for him in the last two slams despite some nice improvements on serve to boot.
     
  7. pc1

    pc1 G.O.A.T.

    Joined:
    Jul 18, 2008
    Messages:
    14,434
    Location:
    Space/Time continuum alternative reality
    I'm not necessarily dismissive but over the course of a year with many matches, GW% generally tend to have somewhat similar winning percentages. This is what I've found over the years. I am quite curious of @NatF new stat which I find fascinating.
     
  8. Gary Duane

    Gary Duane G.O.A.T.

    Joined:
    Jun 5, 2014
    Messages:
    10,876
    @pc1 may be having problems with DR, which is a very different matter. I utterly reject DR as meaningful because it totally distorts a principle.

    Remember, DR is return won/service lost. Granted, using points is not as ridiculous as using games, but the principle is still horribly flawed.

    If someone is winning 97% of service games and 6% of return games, DR suggests this is equal to 80/40.

    6/3 =80/40

    And that's simply absurd.

    The absurdity become obvious towards 100%, but there is still a built exaggeration that makes this not work.

    I might rate 97/6 more like 53%, but not like 60%.

    That's the problem with DR.
     
    Last edited: Feb 7, 2017
    NatF and -NN- like this.
  9. Gary Duane

    Gary Duane G.O.A.T.

    Joined:
    Jun 5, 2014
    Messages:
    10,876
    Sorry, I can't unpack that statement. Could you be a bit clearer?

    GW% is most definitely not the same for a whole number of players, only fairly close for the very best who are repeat slam winners.

    I can take one look at % of points won, % of games won, % of service games won and get a pretty good idea of who will rank high AND be likely to win big tournaments in this era.

    It's too early to make conclusions for 2017, but of the younger players Dimitrov is off to a great start and will most likely be a major player this year if those stats stay up there.

    You can't see what I see because I've spent years collecting this data where I can see at a glance the best players of the last 25 years for all games. The ATP has this data fragmented, so you have to add it up yourself, but weighted the GW% for Djokovic, Nadal and Federer are neck-and-neck, all over 58%. And I may not have weighted them quite enough for serve.

    I can't line this up with the 90s because the eras are different, and that changes the game dynamics.
     
  10. pc1

    pc1 G.O.A.T.

    Joined:
    Jul 18, 2008
    Messages:
    14,434
    Location:
    Space/Time continuum alternative reality
    Maybe so but I have to examine it. However I see no reason why we cannot adjust this to see what numbers we can use to get appropriate ratios in relation to winning. I see no reason to reject it out of hand. I think the concept is great.
     
  11. Gary Duane

    Gary Duane G.O.A.T.

    Joined:
    Jun 5, 2014
    Messages:
    10,876
    Sampras won about 24% of return games, all surfaces, career. His return number some years, mostly earlier in his career, were exceptional. As we both know, Sampras coasted more than any other player from the last 25 years and MAY be an example of how earlier champions approached the game. In a lot of ways he was a throw-back, in the best sense of the word.
     
    metsman and Meles like this.
  12. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    Fed vs Nadal, IW 17 4R, Fed won 6-2, 6-3

    [​IMG]

    federer : 26 W, 17 UEs, won 57 points
    Nadal : 10 W, 15 UEs, won 42 points

    fed's AM = (57-32)/99 = 25/99 = 25.25%


    For 12, it was :

    2012 Indian Wells sf, d. Nadal - 21.4% ( from Krosero )

    though that was on a highly windy day , so its not a fair comparison ...
     
    krosero likes this.
  13. krosero

    krosero Legend

    Joined:
    Dec 3, 2006
    Messages:
    7,001
    So Nadal AM was 10.1%. It was 10.7% in the '12 match.

    Do you have the net stats for the recent match?
     
  14. Waspsting

    Waspsting Semi-Pro

    Joined:
    Mar 21, 2017
    Messages:
    633
    Have you considered differentiating "unforced errors" into two groups?

    Call them "Casual UEs" (CUEs) and "Forceful UEs" (FUEs)

    The former would be run of the mill UEs, like dumping a regulation backhand into the net.

    The latter would be UEs that occur when a player attempts to hit a winner or force an error, like sending a down the line backhand wide while the other player is on the other side of the court.

    I think FUEs get to the heart of what Aggressive Margins is looking to measure, while the CUEs component serves as a baseline for, shall we say 'natural' errors.

    Would require some interpretation and it could get complicated, but I think this expansion would give a clearer picture in numbers of whats going on on the court.

    What do you think?
     
  15. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    yep, even I noticed it just now. Its just in that picture itself.

    at the net,

    nadal : 1/1
    federer : 4/7
     
    krosero likes this.
  16. pc1

    pc1 G.O.A.T.

    Joined:
    Jul 18, 2008
    Messages:
    14,434
    Location:
    Space/Time continuum alternative reality
    That's because I have discussed this with a man who is perhaps the leading sports statistician in the United States and the official statistician for MLB.

    I'm not dismissive. I keep an open mind on this.

    Incidentally DR to me has great potential and I think it could be a wonderful stat.

    @NatF
     
  17. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    krosero and NatF like this.
  18. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    Federer def Nadal Miami 2017 final : 6-3,6-4

    got the W/UE stats from here :

    Federer : 29 W, 19 UEs
    Nadal : 17* W, 23 UEs

    Federer won 71 points
    Nadal won 56 points

    Federer's AM = (71-42)/127 = 22.83%
    Nadal's AM = (56-42)/127 = 11.02%



    Breaking the W/UE stats further :

    FH wing :

    federer : 19W, 13 UEs
    nadal : 8W, 10 UEs

    BH wing :

    federer : 5W, 5 UEs
    nadal : 5W, 12 UEs

    service :

    federer : 5 aces, 1 DF
    nadal : 4 aces, 1 DF


    http://www.atpworldtour.com/en/news/federer-brain-game-miami-2017-final

    *it should be 17 W for nadal, not 15 ( 8 fh, 5 bh, 4 aces )
     
    Last edited: Apr 3, 2017
    NatF and krosero like this.
  19. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    NatF likes this.
  20. NatF

    NatF Talk Tennis Guru

    Joined:
    May 10, 2012
    Messages:
    27,572
    Location:
    Cretaceous
    Even Murray's AM doesn't look that low considering...
     
    abmk likes this.
  21. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    Yeah, its surprising.

    Sent from my MotoG3 using Tapatalk
     
  22. NatF

    NatF Talk Tennis Guru

    Joined:
    May 10, 2012
    Messages:
    27,572
    Location:
    Cretaceous
    I guess because both kept their unforced errors relatively low the AM's are going to be fairly high. Murray obviously served poorly which isn't reflected by a stat like this.
     
  23. Steve Joseph

    Steve Joseph Banned

    Joined:
    May 17, 2017
    Messages:
    12
    Big four will be on tour until the next five years, with federer going first
     
  24. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    yeah ..
     
  25. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    Nadal vs Thiem : Nadal def Thiem : 6-3,6-4,6-0

    23 W to 22 UEs from nadal
    21 W to 34 UEs from thiem

    Points won : 94 for nadal, 61 for thiem

    nadal's AM = (94-56)/155 = 24.52%
    thiem's AM = (61-56)/155 = 3.23%

    Stan vs Murray : Stan def Murray : 6-7,6-3,5-7,7-6,6-1

    87 W to 77 UEs for stan, 36 W to 36 UEs for murray
    Points won : 179 for stan, 160 for murray

    stan's AM = (179-113)/(339) = 19.47%
    murray's AM = (160-113)/(339) = 13.86%
     
    Last edited: Jun 10, 2017
    NatF likes this.
  26. Moose Malloy

    Moose Malloy Legend

    Joined:
    Dec 27, 2005
    Messages:
    8,541
    Would anyone care to calculate the aggressive margins for today's women's finals for me?
     
  27. NatF

    NatF Talk Tennis Guru

    Joined:
    May 10, 2012
    Messages:
    27,572
    Location:
    Cretaceous
    That AM for Thiem is it the lowest you've seen?
     
  28. Moose Malloy

    Moose Malloy Legend

    Joined:
    Dec 27, 2005
    Messages:
    8,541
    @krosero has collected all the AM's we have. I believe the lowest we have are for Connors in his losses at RG in 82 & 85, Vilas in the 78 RG final, and Austin in the 80 USO SF. All were in the negative.
     
    NatF likes this.
  29. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    Points won : Ostapenko - 102, Halep - 94
    54/54 W/UE for Ostapenko
    8/10 W/UE for Halep

    AM for Ostapenko = (102-62)/196 = 20.4%
    AM for Halep = (94-62)/196 = 16.32%
     
  30. Moose Malloy

    Moose Malloy Legend

    Joined:
    Dec 27, 2005
    Messages:
    8,541
    ^thanks! Despite so much emphasis on her UE's here, I think that shows she played pretty well.
     
  31. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    agreed ...
     
  32. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    The men's final :

    nadal def stan : 6-2,6-3,6-1

    points won : nadal = 94, stan = 57
    nadal : 27 W to 12 UEs
    stan : 19 W to 29 UEs
    masterclass from nadal, mediocre from stan

    total UEs = 41, total points = 151

    nadal's AM = 35.1%( 94-41)/151
    stan's AM = 10.1% (57-41)/151
     
    Last edited: Jun 12, 2017
  33. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    Federer's AMs at 2017 Wimbledon so far :

    1R vs Dolgopolov: 32.56% -----(50-22)/86 --- Won 50 points out of 86, total UEs = 22
    2R vs Lajovic : 41.81% ------(98-29)/165 --- Won 98 points out of 165, total UEs = 29
    3R vs Zverev : 47.37% -----(106-16)/190 --- Won 106 points out of 190, total UEs = 16 (net rushing from Zverev cut down on # of UEs in this match)
    4R vs Dimitrov : 35.54% ------(95-36)/166 --- Won 95 points out of 166, total UEs = 36
    QF vs Raonic : 45.95 % ----- (105-20)/185 --- Won 105 points out of 185, total UEs = 20
    SF vs Berdych : 37.34% -----(126-39)/233 --- Won 126 points out of 233, total UEs = 39
    F vs Cilic : 40.625% ---- (96-31)/160 --- Won 96 points out of 160, total UEs = 31
     
    Last edited: Jul 16, 2017 at 8:41 AM
    NatF likes this.
  34. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    Dimitrov's AM in the 4R match was 21.09%
    Raonic's AM in the QF was 32.43%
    Berdych's AM in the SF was 29.18%
    Cilic's AM in the F was 19.375%
     
    Last edited: Jul 16, 2017 at 8:41 AM
  35. Moose Malloy

    Moose Malloy Legend

    Joined:
    Dec 27, 2005
    Messages:
    8,541
    @abmk @NatF

    Was this the lowest winner total for Fed in a major final? 23 seems awfully low.
     
  36. NatF

    NatF Talk Tennis Guru

    Joined:
    May 10, 2012
    Messages:
    27,572
    Location:
    Cretaceous
    Possibly, maybe FO 2008 might be around the same? Official stats put him at 35 but I think there's an issue with those - I'm sure @abmk will know though.
     
  37. abmk

    abmk Talk Tennis Guru

    Joined:
    Dec 9, 2008
    Messages:
    24,470
    Location:
    U.S
    The FO 2008 stats have # of first serves by the player in the winners column.

    But # of UEs seem fine.

    Federer won 52 points. 7 were UEs from rafa. So total of 45 winners+forced errors.

    Tennisabstract has Federer at 16 winners - which seems reasonable.

    http://www.tennisabstract.com/charting/20080608-M-Roland_Garros-F-Roger_Federer-Rafael_Nadal.html

    This was Federer's 2nd shortest slam final after RG 08 final in terms of points, even shorter than the double bagel final in USO 04. ( This had 160 points to 164 in the USO 04 final).
    Federer hit 40 winners in that one.
     
    NatF likes this.

Share This Page