Time-Saving Shenanigans and the Effect of No-Ad

Yesterday, Colette Lewis reported on another set of possible rule changes for college tennis.  The goals, as always, are to shorten matches, increase television coverage, and systematically ignore the well-informed preferences of those most closely involved with the game.

Colette does a better job of explaining the limitations of the proposed format than I would, so I encourage you to go read her post.

So that we’re all on the same page, let’s summarize the most recent dual-match format:

  • Dual matches–meetings between two schools–have six players to a side. There are three doubles and six singles matches.  The combined results of the doubles matches is worth one point, and every singles match is worth one point, for a total of seven.  The first school to four points wins.
  • First, three doubles matches are played simultaneously.  Each is a single set, first to seven, win by two, and a tiebreak is played at 8-8.  If one school wins two of the three doubles matches while the other is still in progress, the third is abandoned.
  • Next, all six singles matches are played simultaneously.  Each singles match is best of three tiebreak sets.  Once one school has accumulated four points (including the doubles point), the remaining matches are abandoned, and the contest is over.

And here’s the new version:

  • Singles first.  The singles format stays the same, with six simultaneous matches.
  • If and only if a team does not accumulate four points in the singles, a compressed version of the doubles is played: the three doubles matches are reduced to 10-point super-tiebreaks. (This time last year, we were debating the merits of those as third sets.)

The proposed alternative would certainly save time.  It would also effectively destroy doubles as an important part of college tennis.

At last year’s NCAA Men’s Team Championships, 44 of the 63 dual matches were decided by a score of 4-0 or 4-1.  While it’s impossible to know how the abandoned singles matches would have turned out, it’s safe to assume that almost all of those meetings–along with many of the 4-2 outcomes and a few of the 4-3’s–would have been decided before any doubles was necessary.

Since length is such an important part of these debates, I ran some numbers to see what else might be done.

No-ad, no-overtime

The most popular device for speeding up tennis is “no-ad” scoring.  You’re probably familiar with it, as both the ATP and WTA tours use it for doubles.  Once a game reaches 40-40, the receiving team decides whether to play a final point in the deuce or ad court, and the outcome of that point decides the game.

So, how much time does it save? To get a rough idea of the answer, I looked at roughly 3600 ATP Challenger singles matches from this year.  (It’s not the most relevant dataset, I realize, but better “available” than “ideal.”)

In those matches, 24.2% of games went to deuce.  Those games averaged 9.7 points each, meaning that a switch to the no-ad format would save 2.7 points per deuce game.  Overall, such a change would save about 0.65 points per game across the board.  The average best-of-three-sets match lasts about 22 games, so switching to no-ad scoring would reduce the number of points in the typical match by 14 or 15.

At the ATP level, each additional point within a game–that is, one that doesn’t add to the number of changeovers or set breaks–adds about 33 seconds to the length of the match.  So the switch to no-ad scoring would shorten the length of an average match by about eight minutes.

Switching doubles to no-ad would have a lesser effect, because the matches are already shorter.  Figuring an average match length of 10 to 12 games, that’s another four minutes saved.

The impact is a bit more ambiguous than I’ve made it out to be, because no-ad scoring makes service breaks more common.  If the server has a 65% chance of winning a point (typical for male tour pros), he or she has only a 65% chance of holding from deuce in the no-ad format.  The server’s chances might be a bit worse, assuming the returner chooses the side which favors him or her.  In an ad game, that same server has a 77.5% chance of holding from deuce.

It would take a much more in-depth simulation (informed by much more college-specific assumptions) to know the impact of that difference.  Some additional breaks would speed up matches, making 6-0 and 6-1 outcomes more likely, while others would push sets to tiebreaks.

But college tennis takes longer

So far, I’ve been forced to use numbers from the pros to evaluate proposals for collegiates.  Somewhere along the line, the numbers don’t add up.

According to the advisory group responsible for the “hide-doubles-in-the-attic” proposal, the average dual match time at last year’s NCAA championships was over three and a half hours.  What’s taking so long?

On the ATP tour, the average best-of-three match is just over 90 minutes.  Doubles generally moves more quickly, so the first-to-seven matches should be less than half as long.  Plus, since dual matches are often decided while the longest singles matches are still going, the average completed match must be shorter than the average match at the college level.

Thus, even accounting for less serve dominance, longer rallies, and assorted factors such as the absence of ballkids and the higher number of lets (remember, these matches are played on adjacent courts), how are we getting so far beyond the magic three-hour time frame?

One explanation is simply poor data collection and analysis.  The numbers the advisory group cites are from last year’s team championships, a particularly small sample.  And by using an average, and not a median, one or two very long matches can skew the numbers–especially with such a small sample.  The five-hour dual matches are surely beyond saving, so why give them so much weight?

An alternative explanation is that college tennis really is that much slower, in which case many of the numbers I cited above don’t tell the whole story.  Are there far more deuce games in college than the 24.2% on the Challenger tour?  Are interminable, 15- to 20-point deuce games much more common?  Do points take considerably longer?

If so, the effect of moving to no-ad scoring would be greater than the twelve-minute conclusion I reached above.  Twelve minutes is a little less than one-tenth of my estimate for equivalent ATP matches, assuming a 90-minute average singles match and 45 minutes for doubles.  So if dual matches are really lasting three hours and 40 minutes, the equivalent time reduction would be almost double–better than 20 minutes.

Purists may hate no-ad scoring, but given a choice between losing 15-point deuce games and losing college doubles, I’d ditch dramatic deuce games in a second.

About these ads

5 Comments

Filed under College tennis, The Rules

5 responses to “Time-Saving Shenanigans and the Effect of No-Ad

  1. I will confess that I have never watched college tennis – pretty much by accident, not by design. The only good thing for me about all this is I suddenly have the desire to see if I could attend some college tennis tournaments in the NYC area. I’ve enjoyed watching small college football matches up close and in person, even when I had no rooting interest in either team.

    Having said that, I will add that as a casual fan of professional tennis, it seems to me a terrible idea to downgrade doubles. While I don’t watch many doubles matches at regular tournaments, I do tune in fairly regularly to Davis Cup action (not so much Fed Cup), and doubles in that kind of atmosphere, where the partisan crowds get involved & the team benches on both sides do a lot of cheerleading, is a treat. From the comments in Colette Lewis’s blog post, I gather that the same is true for college doubles as well (e.g. someone named Mal Taam comments that “The crowd [in doubles] and the teams/teammates cheering each other on cannot be found in any other aspect of the game”).

    I gather (perhaps wrongly) from the phrasing of the letter as reprinted in Lewis’s post that this is not a done deal, merely recommendations being passed onto others to consider. Hopefully it will get shot down and something more intelligent will be proposed. Do you know whether or not something as simple as no-ad was considered by this advisory group? Or did they just go ahead & jump out on the weakest possible limb? As a fan I personally don’t mind no-ad all that much, and it is easy to get used to.

    Beyond that you mention something rather simple & obvious that should have been born in mind and apparently wasn’t – namely, the difference between average and median (a.k.a. mean). I forget when it was that I first learned the difference – probably a piece of reporting I was doing for some publication or other. Regardless, it was an important thing to learn & should obviously be considered here. Where is the raw data lurking?

    • Yep, median vs mean is so important, especially in mainstream discussions of income inequality and stuff like that. Imagine the difference between the median and mean match times of men’s Wimbledon matches the year of Isner-Mahut…

      This is just a proposal, but it’s a proposal sponsored by the group (the USTA) that is funding broadcasts of these events. So while an ITA (college coaches association) counterproposal is forthcoming, not sure how much weight that will have.

      It does seem sometimes like the unpopularity of doubles is a self-fulfilling prophecy. Maybe people really don’t/won’t watch it, but it so rarely gets any kind of chance. Everyone involved in college tennis thinks it is the heart of the college game, and as Colette has pointed out, it is the most likely career path for collegiates going pro.

  2. I think you are worng when you mention that no ad would make more 60 or 61.
    Because when you say that you are assuming that the player serving got 50% or + to win a point but if the player receiving had 50% or + to win a point then it s the opposite.
    Basically you can just say that no ad makes every game closer. One you have said that I am not sure you can say more 6 0 especially for WTA.

  3. David

    This will be very long, so feel free to skip it if you want to.

    I have attended the last 6 NCAA’s tournaments and the last 8 or 9 ACC conference tournaments. I do not attend many dual matches as my favorite team (UVA) is a 3 hour drive away each way. I only watch the men play and do not go to any women’s matches. BTW if UVA was not really good, I would not have attended any of these NCAA tournaments. Very few will travel to see their team lose in the round of 16 to a top 4 seed, which is almost a certainty. I have heard of many 5+ hour matches (mostly women) at the NCAA’s, often with the temperature around 90.

    The longest regular season dual match I ever watched was over 6 hours, when a couple of expert men moon ballers/retrievers played each other at #4 or #5 as I recall with the match tied 3-3. It lasted over 4.5 hours with 30 to 40 point shots being routine. There was no streaming or TV back then, but that was a very long day counting my 3 hours of round trip driving to UNC from Greensboro, NC. Not many will devote that much time to watching a college tennis match. One ad/no ad would have shortened this match by at least 1.5 hours.

    Women’s tennis has many more moon ballers/retrievers and these matches do last a very long time, especially with the lack of clinching rule in dual matches. The NCAA tournament and the NTI have the clinching rule, which shortens matches, unless it is very close and there is a long three setter that has to be played to conclusion, because it decides the match, which is kind of rare these days.

    The real problem at the NCAA’s is not necessarily the length of the individual matches, but trying to squeeze the first four days of the men and women tourneys onto just 12 courts. The days start at 9 AM and often end after midnight. Aside from the fans, it must be tough on the umpires and other officials who have to be there. There are ways to fix that, but no one wants to change it.

    I think that the NCAA women’s final this year was extremely long this year I watched the NCAA men’s final 4-3 win by UVA over UCLA and because my team was playing, I did not care how long it took. I did not time it, but my guess is it took about 4.25 to 4.5 hours. I thought it was great, but maybe a fan with no rooting interest would think it was too long. It was streamed and many watched it that way. I will say that the weather was unusually pleasant, cool, and cloudy at the team NCAA’s this year in central Illinois, but had the fans been asked to sit for 4.5 hours in the sun when it is typically 85+ degrees, most could have not lasted that long. This years site Athens is almost always very hot and humid in mid May. The worst heat one I ever saw was the semis at Tulsa with 95 degree air temperatures and 110+ degree hard court surface temperatures. Luckily, Tulsa had awnings for the fans and they provided some comfort or no fan would have lasted.

    If this change is all about fitting matches into an approximate 3 hour window for in season dual matches for some as yet unannounced TV sports network deal or conference network like the B1G, SEC, etc network, then it almost has to happen. Maybe a few (maybe 50 or so) more fans will show up or stay to the end if the matches are shorter, but I doubt that is the reason for the change. College tennis is just not that popular and never has been and fans will ONLY come to watch it when their own team is playing. Oddly enough a close college tennis match with 6 singles going on at the same time is IMO the most exciting college sport to watch. I have played tennis for 60+ years and know a whole lot of people who also play the game. I am the only college tennis fan among the hundreds that I know and many attended area colleges/universities. They never go to watch their own team play. Almost all love to play the game, but do not like to watch tennis unless it is a Grand Slam and the big guys are playing.

    How to accomplish the “shortening” is a very tough call with no easy answers, but something will obviously have to go. I do not think it is the average roughly 3.5 hour match that is the real problem. Some blow outs end in 2.5 hours or less with a clinching format. The TV problem is with the 4 to 5+ hour matches. I think nearly all college tennis fans, players, future players and coaches do not accept fan convenience or TV as a good enough reason for the matches to be shorter. All of these constituents have plenty of time to play all day, except sometimes airline schedule are impacted for the away team.

    Many years ago there were 9 points in college tennis and doubles was two out of three sets with no clinching rule and everybody rode buses or vans, which meant that matches routinely took 5 or 6 hours to finish. That is why doubles was changed to three separate 8 game sets and only one point was allotted to doubles and it was played first instead of second like in the old days. Nobody or almost nobody would stay that long, especially if the weather was not good (too cold, too hot, or too rainy) on that spring day, which is still true today, except perhaps in Southern California.

    Few people currently attend live college tennis matches and if it is only an average or below average team almost no one except family and friends attend. Even the NCAA tournament draws poorly (IMO maybe 1,500 people were there this year for the men’s finals and much fewer in the early rounds) unless the home team is playing and lately the men’s home teams have not made it into the 16 field or have exited early. I mean no offense to women’s tennis, but it is much more poorly attended than the men. People always look back at the old UGA days when their students were still in school during the NCAA’s (not true now) and UGA was playing for the title in Athens and maybe some beer/tailgating was happening. They probably had 5,000 plus back then, but only if UGA was playing.

    Somebody or something is driving this shortening the matches “thing” and even the ITA coaches themselves have agreed to experiment this year, but only during the indoor season, which I found very surprising as they no doubt like it the way it is. A cynical person might think that they may just be trying to prove that changing things during the unpopular indoor season with very few good matches will be proof that everything should stay the same.

    I think the ITA proposal is much better than the USTA advisory committee one where doubles is essentially eliminated. You covered the average time gained from one/no ad and the fewer games very well, but the no/one ad will also shorten the time much more in multiple deuce games. It is common for evenly matched players, especially women, to play 5 or even more deuce games with many shots per point.

    Obviously, the clinching aspect change also shortens dual matches.

    I am probably the only huge college tennis fan around that has no problem with the ITA proposal. I think the changes do hurt those top players who are trying to use college tennis to get into professional tennis, especially doubles. However, unless you are close to a NCAA championship level and a top 5 college player, you will not make it in professional tennis and 98% of the college players are not at that level. I am not convinced that college tennis should be structured for the few who can play it at the pro level.

  4. Dan

    I played D1 tennis at a school formerly in CUSA, now the Big East. I don’t think it should come as a surprise that instant line call challenges, timed changeovers, and an endless supply of ball boys creates faster matches.

    My high school team state tournament format was seven players, three singles and two doubles simultaneously, one point each. First team to gets to three, wins. You were permitted to make strategic lineup construction – you did not have to put your best players in 1-7 order. Your 1 singles had to be better than 2 singles, etc., but your 1 doubles could comprise of your teams two best players. The result was lots of teams with very strong 1st singles players would move their 2 and 3 players to first doubs, bump first doubs to second doubs, and sacrifice 2 and 3 singles. The high school format was, in basically every way – viewing, playing, etc. – better than the college format.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s