General TennisRoger Federer

The Speed of the Courts in 2017

We've discussed court speed many times on the blog so I thought it would be interesting publish this year's Hawkeye CPI data which compares the speeds from all Masters 1000's in 2017 and gives a like for like comparison from the 2016 season. Most of the numbers come from screenshots I've managed to grab when watching TennisTV, screenshots others have grabbed and tweeted and this chart posted by TennisTV commentator Nick Lester.

Unfortunately, I don't have the data for the Grand Slams in 2016 to compare but as you can see only one court was faster in 2017 than it was in 2016 – Monte Carlo. Although I think we have to assume this year's Australian Open was faster than 2016 too but I don't have the data.

Court Speed Index for 2017 (Main Show Court Average)

Tournament CPI 2016 CPI 2017 Year on Year Difference
Australian Open Unknown 42
Indian Wells 30 27.4 -2.6
Miami 33.1 30.3 -2.8
Monte Carlo 23.7 24.9 +1.2
Madrid 22.5 20.9 -1.6
Rome 24 22 -2
Roland Garros Unknown 21
Wimbledon Unknown 37
Montreal N/A 36.3
Cincinnati 35.1 33.6 -1.5
US Open Unknown 35.7
Shanghai 44.4 42.9 -1.5
Paris 39.1 37.5 -1.6
London 42.1 41.4 -0.7

So are the courts still getting slower?

Well, they certainly are based on 12 months data but we need to look over a longer period to really know. Fortunately, I do have the historic numbers for some specific tournaments but not all. Sadly we'll never have the data for pre 2012 when there were some big changes to the speeds but nonetheless you can see the numbers for the last 5 years of Miami (except 2012 which is unvailable), the BNP Paribas Masters and the World Tour Finals below:

  2012 2013 2014 2015 2016 2017
Miami 31.5 29.8 31.2 33.1 33.8
Paris Bercy 32.2 31.2 31.8 29.9 38.9 37.5
World Tour Finals London 33.9 32.8 33.6 34 40.6 41.4

So based on the above the speeds have gone up a notch in recent years across those three events. Paris tries to match court speed in London so they'll likely be similar this year as well. The problem for me is that there are no speed standards to adhere to, tournament directors can play around with it willy-nilly. Ideally, there should be some sort of control in place so that we get a mix of speeds throughout the year across all 4 surfaces as they're all very similar across the season.

Any suggestions on how the ATP should enforce it? Allocate a set court speed between X & Y to each tournament per year that they have to fall within and ensure there's a mixture of speeds across the Calendar?

Problems With This Data

For those eagle-eyed readers you'll notice that in my second table both Miami and Paris have different ratings to the 2016/2017 comparison. E.g. Paris 2016 has a CPI of 38.9, but in the chart above has one of 39.1. The 38.9 comes directly from a TennisTV screenshot taken mid-tournament, and the 39.1 comes from the latest numbers that Nick Lester got directly from Hawkeye during the Paris tournament. The discrepancy is because the speed of a hard court changes over a tournament depending on conditions and the roughness of the surface underfoot (it gets smoother the more it's played on). So the 38.9 number is the average speed at a certain point in the tournament whereas the 39.1 is the average speed once the tournament is done and dusted. Same for Miami.

That is a problem because we'll only know the true average speed once the tournament is finished. That means we'll only get the info midway through the final, or if they mention it at the next Masters 1000 tournament. Otherwise, it's a 12-month wait until the tournament rolls around again and last years court speed is brought up on screen.

The other problem is there's no real archive or source for the data – we are reliant on Hawkeye revealing it, TennisTV publishing it or one of the commentators doing some research and getting the numbers from a Hawkeye operator. And then, of course, a viewer has to screenshot it or make a note of it. Robbie Koenig does seem to be the on the ball at putting these numbers out but he's not commentating on every tournament so there's always going to be holes in the dataset, unfortunately.

Let me know what you guys think in the comments. Should the speed of the courts be regulated?


Huge fan of Roger Federer. I watch all his matches from Grand Slam level right down to ATP 250. When I'm not watching or writing about tennis I play regularly myself and have a keen interest in tactics, equipment and technicalties of the sport.

Related Articles


  1. We is SCANDALOUS. Before explaining what I mean, we will use the following as a guideline*.

    Fast: 45+ (CPI)
    Medium Fast: 40+
    Medium: 35+
    Medium Slow: 30+
    Slow: <30

    Using that as a guideline, here is a percentage split of events based on CPI, across the 14 big titles. This is outrageous! There are no fast courts out there at all. None! 50% of them are below medium, and just 21% above. The courts are heavily tilted towards the slower categories, with Roland Garros, Madrid, Monte Carlo, and Rome virtually at snails pace. You're welcome, Nadal.

    Again, there are no fast courts. AO and Shanghai are are the closest we get to a fast court. What surprises me is that Indian Wells is slower than Miami. Hard to believe.

    Fast: 0%
    Medium Fast: 21%*
    Medium: 29%
    Medium Slow: 14%
    Slow: 36%

    **Assumption: WTF is placed in the Medium Fast category

    1. My conclusions: –

      – At least two more grass events must be added (preferably Master 1000’s).
      – At least two more big titles must be moved from the Slow to Medium Fast speeds for balance.
      – At least one medium pace title needs to us a fast court.
      – If the WTF must be played on clay once in a while, then it must be played on grass too. Fair?
      – If Wimbledon is going to be forced to stay in the medium category, then US Open should at least be moved closer to the medium fast range. Right now it just outside medium slow.
      – And yes, tournaments should adhere to a range of court speeds for a specific year. Speed depends on other variables too, so that needs to be accounted for.

      1. Alison, that was funny. I’m sure you don’t have to go down under Maybe somewhere in Southern Texas? Why not? What I mean is, if you are going to push for clay for the WTF, why not talk about grass?

    2. The problem with that is those categories you have given are pretty much the CPR ratings from the ITF.

      We don’t know how close CPI and CPR are in terms of calculation though as one is done by Hawkeye, the other by the ITF with either a Sestee or a Sprite. Hawkeye factors everything as it’s taken from tournament play. CPR is taken usually at a lab in London, not courtside.

      If they’re calculated similarly then you’re right… no courts are fast.

      1. True. But even if you assumed that AO and Shanghai are fast, my argument is more about the distribution. The speed is more tilted towards the slower side (in terms of percentage of events).

        The other thing to keep in mind is that there is such a thing as a fast court. Remember the Asian exhibition matches between Sampras and Federer sometime in 2007? Their last match was played on what I consider a fast court. Sampras won that one by the way but lost the series. I can tell you that AO2017 was nowhere near what I consider fast. It’s not a myth. There is something called a fast court.

      2. Yeah, like I said below I guess we gotta assume that CPI and CPR are pretty close in how they are calculated. As the formula for measuring speed is going to be identical or near identical regardless of whether it’s hawkeye data or ITF sestee data? Which means your categorisation is correct.

    3. Before I start, I just wanna say that I agree with you, Sid – I find it stupid how little variety there is on tour nowadays and I really hope change is forthcoming. For me, there are a few things worth noting, though.

      1. I don’t know how arbitrary your use of boundaries as a guideline for fast are. Seems to me the more statistically significant way is to plot the mean court speed over the last five years and checking the skewness coefficient -I.e compared to tour average, how many courts fall below that number. I haven’t run them, but I suspect it’s astoundingly bad.

      2. I think it’s very easy to get caught up in the CPR as an accurat reflection of court speeds, whereas in truth it’s probably dependent on wind, conditions, and even the temperature on the day the measurement was taken. I’m betting there’s no real standardised testing involved, and an increase from a year to the next might be explainable by something other than conspiracy.

      I think the first step is to make the CPR an official parameter for courts – encourage commentary to use it to illustrate conditions on a specific day etc. Have a full specification of how it is measured and controlled, and make this transparent. Then we can start looking at imposing limits on tournaments.

      1. I can only do what I can with the data available. One of the assumptions is that 42.9 (Shanghai) isn’t exactly what you call a fast court. It didn’t look that fast to me to be honest. Faster means comparatively faster. I’ve added a link to a table at the bottom which defines boundaries and assumes that there must be fast courts up to a 50 CPI. Can it be lower than 20? Sure, I mean you would have to be playing on a few inches of sand.

        My point is, court speeds are definitely skewed towards the lower side.

      2. CPR is irrelevant here though really.

        This is CPI – it comes from Hawkeye data during the tournament. So it takes into account all variables as it’s measured during matches.

        CPR is the ITF measurement that applies purely to the surface when it was tested. This is rarely done on site, usually, manufacturers send a sample of the surface and it’s tested in a lab. I believe they do test on site where practical but their equipment is bulky. Used to be a sestee which is very cumbersome. Now use a Sprite or something which is a bit more portable.

        Ideally, we need Hawkeye to give us some categorisation of speed. Otherwise, we are looking at CPI numbers and comparing to ITF CPR categories.

        However, I guess we gotta assume that they two are pretty close. As the formula for measuring speed is going to be identical or near identical regardless of whether it’s hawkeye data or ITF sestee data?

    1. You’re contradicting yourself. You do not want regulation, yet you want court speeds to be faster. How do you go about enforcing that? What is to stop for example, some of the faster courts from being slowed down?

      1. What I meant is (that I didn’t include) that generally court speeds should be regulated, but the tournaments themselves should have the right to modify their court speed as long as they don’t break the required limit. This would require rigorous testing, and I’m all for it! Like I said, these general limits, should have the purpose of making courts going faster. Now for details, I would let the tournaments decide for themselves as long as they are in the limit zone.

  2. Ubitennis had a chart of the court speed index for the 2017 Masters tourneys and Slams that was in a tweet on Nov. 2. I don’t know if they do this every year. I could not find information on their website.

  3. Its a bit sad the entire sport is now rigged to extend match time in order generate more money. In the end players will suffer with injuries etc. Its fair there should be balance type of court speed instead slowing everything. I am sure viewer will enjoy faster matches rather than endless grinding.

  4. Off topic here BUT did you see that at the tournament in Milan, the fans can walk around freely. Five setters with tie-break at deuce. What the hell is that. The ATP is assuming we all have ADD and can’t sit still anymore!

    1. The way I see it, it’s their generation and they can do it their way. I hope all that monkey business doesn’t
      happen while Federer is still playing. Once he retires, and I hope not any time soon, I’m going to stop following tennis and watch my invaluable collection of Federer matches, exclusively. These NextGen fucks can take off their shirts and go at each other like fucking gaytards on the court and I won’t give a flying fuck! Pretty soon this whole thing is going to turn into the fucking, retarded MMA.

  5. I couldnt Agree more with what Sid had to say. Seems like there is not clarity in concluded the accuracy of the data provided, nonetheless great effort J.
    But looking at the Court Speeds of IW, Miami, Montreal, W&S – they all average at 31-32???? its absolutely risible!!!!!! they have to push it to 40s atleast if 45s. And then I wonderwhy a novak djokovic dominates the masters club- simply because like a nadal enjoys clay , a novak enjoys a clayish-hardcourt club. it honestly makes tennis that much monotenous and boring tbh. Paris at 38? I seriously doubt it – it looks a green clay court to me . maybe its the camera angles or just me. The Australian Open and Shanghai are comparatively ‘quick’ and not fast like Sid rightly mentioned. Yes Federer has advantage on fast courts but Nadal Rafa came out the second best on em. As a fed fad I am honestly not being greedy , it ed just be nice to see player adjusting to different conditions – hell it would make the game a lot more interesting. Look at the Wtf in 2012-2015 court speed for an indoor court, laughable. 2016 was slightly quicker and we had a different winner. Interesting there.

    Anyways dont see that happening… lets hope London stays atleast somewhere between 40-45 for heavens sake.

    1. The camera “angles” discussion is really about the camera position. In Paris, the camera is very far away from the court, so the relative distance between the camera and each of 2 baselines is very similar. That’s why the court looks so flat: the perspective looks “compressed” and the ball barely seems to move. By contrast, in smaller courts, like the ones you see in the Challenger circuit, the camera is closer to the court, so the farthest baseline is can be 3 or 4 times farther away than the closest one, so it appears 3 or 4 times smaller. What you get is a very deep perspective, where the ball seems to fly at you at supersonic speed and fade into oblivion when it travels back.
      (if you watched “Vertigo”, check the scene where James Stewart looks down the stairs and nearly collapses)
      The greeks noticed this some 2500 years ago, long before tennis matches were televised 🙂

    1. Pfft… HaHaHa. What a gross show. Really, I like the sentence (“our execution of the proceedings was in poor taste”)
      This is typical political talk. It’s not the “execution”. It’s the idea, the planning and the decision. And the chump who took it. The “execution” is right in that it must follow THE script, so it is beyond blame.
      Who are they trying to imitate? The announcement of round X in a boxing fight? The starting grid in motorcycling races?

    2. Things are going from bad to worse? Nopes. It’s a gimmick for popularity. Now, the whole world will be talking about it, even those who do not know what tennis means. All this was well planned and executed.

  6. I think the #nextgen new rules are fun and I’m really looking forward to seeing Nadal or Marin Cilic trying to play in these timespans It’s good to change things around, and this is the #nextgen, they can cope. I like the no-ad rule. I also think they should do away with the second serve; if they can’t get it in the box the first time then TUFF. Federer would have no trouble at all with these new rules. Andy Murray would have a fit, he gets upset if Someone in the arena is using a phone.

    1. Sheila,

      Are you fucking serious?!?

      It’s the fact that you have to win a game by two points that makes getting a break so meaningful, and so satisfying to the fans. Think of that deuce game between Federer and Nadal at the Australian open. A long 26 shot rally did not give Federer the break. Nadal erased it with a superb serve. Federer had to work hard again to earn the break. What about the tug of war in the Nadal serve game before that when Federer broke to move to 2-3? That’s what makes tennis interesting. We might as well be watching WTA if we want to go with no-Ad scoring. Sorry, no offense to the women.

      By taking away the second serve, you are erasing the joy of a big serve, a well placed serve. Every serve will then become a second serve. I guess finally we can see men and women playing together and put to rest this equality debate once and for all.

      I think you were trolling, right? 🙂

      1. Hi Sid. I never troll. I was just enjoying something different. I can’t imagine all (any?) of the changes will be absorbed, but it’s interesting to have a change now and then – think of the Laver cup. I’ve been living and watching tennis for a long long time but I can still enjoy changes and yes new things for the sake of them sometimes. As to only one serve, every serve would be a surprise -will it be a “first” serve or a “second” serve?
        Now I’m going to watch The Fed play Andy Murray.

      2. The Laver cup had a different format in terms of how points are scored for a match win. But scoring within the game was the same. So yes, while an overall format change is welcome, they should keep the points scoring system pristine.

        Good luck trying to go for a first serve knowing there is no second serve.

        You want change and excitement? Make court speeds more variable and you will enjoy a variety of playing styles.

  7. I agree, Sid. Some of the most exciting times in tennis is battling for a break. Think about it. Need two serves, etc. I don’t get 4 games with 5 sets. What’s wrong with 6 games and 3 sets. Fed says a match is like an exam, you should be prepared and when you’re on court, it’s all up to you. No coaching during the match.
    What tennis needs to change is Davis Cup. And add a grass 1000 event.

    1. [Some of the most exciting times in tennis is battling for a break]

      Correct. Tennis is a game of breaks. Whether it’s a game break, or a mini-break.

      1. Im not a fan of the new rules either, nor or the service suggestion. The whole format of tennis is built around the fact that it’s difficult to either break or get a break back. If it were not serve reliant every game would be 50 50 – and there would be no such thing as an important game until the end of sets.

        I don’t mind the let cord thing either way, but I actually think hitting the net is banking on luck so would be happy for it to be an automatic fault even if it dropped in.

        What we saw yesterday was simple. You win a game early, that’s the set won. Shortening sets is ridiculous.

      2. I wouldn’t necessarily mind shortening it to 5 games per set, but anything below that is just nonsense. And what’s this about the *server* being able to choose sides at deuce? Doesn’t this tilt the balance significantly in favour of the server? There’s a reason why it’s the receivers who choose in doubles.

  8. Waiting on the O2 draw out, I am reading some statistic this year
    Roger is 11-1 against Top 10 this year (damn Montreal)
    Rafa is 12-5
    Zverev is 6-5

    The other five have tie (goffin) or even negative stat against top 10, yet they are the best 8 in end of year rank. So rank is just an absurd ilusion?

    Anyway, the fact that out of 5 defeats of Rafa against top 10, 3 was made by Roger make me smile 🙂

  9. Well if Rafa doesn’t win the WTF this year he never will. I would be very surprised if his knees keep him
    out of this one. Still I hope to be surprised x

Leave a Reply

Your email address will not be published.

Back to top button