Sunday, April 03, 2005

One Up and One Back

Is there anything to be said for the much-maligned doubles formation known as one up and one back? Can players at any level win using this formation? Are players using this formation condemned to the lower rungs of the tennis ladder?

After seeing Virginia Ruano-Pasqual and Paula Suarez play at Wimbledon a few years ago, I began to question the conventional wisdom that players could not succeed with the one up and one back formation. Right there on the TV, before my very eyes, I saw two women playing in the Wimbledon finals with one player at the net and her partner on the baseline pounding groundstrokes. Neither of the two served and volleyed, nor did they come to the net after returns of serve. I had always been told and always thought that a defining characteristic of better doubles teams was that both players came forward every chance they got, that doubles was played, and more importantly doubles was won, at the net.

This observation was not comfortable for me. I had taught players of all levels to serve and volley or at the very least get to the net at the first reasonable opportunity when playing doubles. If the players were toward the lower end of the tennis ladder, my message was that if they wanted to move up they had to learn to get to the net, to serve and volley. Sure it was tough for them, but if they ever wanted to play better doubles they had to master the transition from the back of the court to the front of the court.

Or did they? Here were Ruano-Pasqual and Suarez succeeding at the highest level of women’s tennis violating this most holy principle. It began to look to me like the serve and volley players, the doubles players who chipped and charged behind returns, were being squeezed from both ends. The lesser players played one up and one back and so did two of the best women in the world.

I began to look around. Sure enough, the winners of local tournaments, the terrors of local leagues at almost all level played a lot of one up and one back. Hmm. Not all, of course. There were still a lot of net crashers winning, but it was hard to claim that one up and one back could not succeed based upon what I saw. I also noticed that many of the best women college players were using the one up and one back formation, too. Hmm. Maybe this craze was confined to top women players, but was not afflicting the men.

Then came the 2004 Athens Olympic Summer Games. Nicolas Massu and Francisco Gonzales of Chile won the gold medal in men’s doubles. They won the gold medal playing one up and one back. Dagger. The dreaded one up and one back system could now claim a major victory in men’s tennis, too. Oh sure, you could claim that this was an unusual event, that most top doubles teams do not pair two players from the same country and therefore the field at the 2004 Olympic Summer Games was weak. Fair enough. But it wasn’t that weak. There is simply no denying that a duo of men from Chile won the Olympics playing in the much-maligned one up and one back formation.

This does not mean that the one up and one back style is necessarily the best formation for everyone or for any specific team. What it does mean is that one up and one back is a valid, successful formation for a significant subset (the majority?) of all tennis players.

In the winter and spring of 2004, my eyes open and with the preceding evidence (with the exception of the 2004 Olympic Summer Games), I set out to investigate just what was behind the success of the one up and one back style. My investigation began with an analysis of the doubles played by my adult tennis students (a couple of dozen adult men and women, NTRP 3.0 - 4.0). Not just a casual observation, but with specific, detailed charting of their points.

My first step in learning about the doubles play of my students was to find out how long their points tended to last and how their points tended to end. To find out how long the points lasted I simply counted the total hits during each point. An ace, service winner, or a double fault was a one-hit point. A service return unforced error, winner, or forceful enough shot to cause an error was a two-hit point, and so forth. That took care of finding out how long their points lasted.

Wanting more information than that, though, I also put a mark on my court diagram indicating the position of the player hitting the final shot in each point. A blue mark indicated a winning shot (winner or shot directly forcing an error from the other team) and a red mark indicated an unforced error. This chart gave me a picture of where the hitter was when the point ended and allowed me to see pretty quickly which areas were more likely to produce winners and which areas were more likely to produce unforced errors.

So what did I find? The graph below (Fig. 1) shows the point lengths, measured by total hits. The left-most column, representing the first hit (serve), shows that 100% of points included the serve. Of course this must be true since all points begin with a serve. The second column, reaching a height of just over 90%, shows what percentage of points included a return of serve. What the 90% figure means is that roughly 10% of the points ended with an ace, service winner, or more commonly a double fault.

Figure 1

Column three shows that less than 70% of points last to the third hit. Put differently, over 30% of points end with the serve or with the return of serve.

As you can see, the probability of a point lasting until subsequent hits drops off sharply. Over half of all points played do not get to the fourth hit (receiving team’s second hit), and less than 10% of all points played went beyond six hits (three per team).

So the points tend to be quick. But how do the points end? Errors. The chart below (Fig. 2) shows how all the points ended, with the red column height indicating points that ended with a unforced errors and with the blue column height indicating points the ended with winners.

Figure 2

Clearly the total number of errors (the red columns) outnumbered the total number of winners. The ratio of errors to winners was roughly three to one for serves and returns of serve. The number of errors was still very large for the third hit, but we see that the winners also began to climb in that third column. The most common point played ended with an unforced error on the return of serve (red bar representing over 80 points). The second most common point played ended with an unforced error by the serving team (red bar representing 60 points). This latter data point is particularly relevant as we examine the wisdom of playing a serve and volley style.

Having learned how long my players’ doubles points lasted, and whether those points ended in unforced errors or in winning shots, I turned my attention to where the last hitter of the ball was on the court when he or she hit the last shot. To do this, I looked to my court diagram with red and blue marks on it. In order to make sense of the marks, I divided the court into three areas: the back court, the transition zone, and the finishing zone. I defined the back court as the area from a step inside the baseline to the back fence. I defined the finishing zone as the area about two steps inside the service line to the net. I defined the transition zone as the area between the back court and the finishing zone.

The chart below (Fig. 3) presents the number of errors (in red) and winners (in blue) from each of the three court areas.

Figure 3

The numbers of errors and winners in Figure 3 do not include serves and returns. In the back court, the number of both errors and winners is low. In the transition zone, the number of errors is very high, and the number of winners is relatively low. In the finishing zone the ratio switches, with winners outnumbering errors.

From this information, it looks like playing balls from the transition zone is a bad idea and that playing balls from the finishing zone is a good idea. But without knowing the total number of hits per zone, it is unclear just how bad and good those ideas are. To answer that question I had to go to video tape.

By video taping the doubles sessions and capturing every hit I was able to expand my knowledge of the playing characteristics to include error and winner rates. The first graph below (Fig. 4) shows what court area all the shots (exluding serves and returns) were played from.

Figure 4

The vast majority of shots were hit from the transition zone. That may explain the large number of errors hit from that zone in Figure 3 above. Having the number of total hits and the number of errors and winners from each zone, determining the error and winner rates was simple. The results of those caluculations for each area are presented below (Fig. 5),again with errors in red and winners in blue.

Figure 5

The error rate is the highest in the transition zone and the lowest in the back zone. The winner rate is the highest in the finishing zone and the lowest in the back zone.

What are the implications of the above data for doubles players and the formations they should assume, the court positions they should strive for, and the goals and responsibilities of each member of a doubles team?

The first implication is that players must avoid making errors in the first few hits of the ball. For the serving team that means getting a serve in on the first shot. Assuming the server gets the ball in, where should the server go? If the goal is to avoid making an error on the next shot, the place to go is nowhere. The server, as the above statistics demonstrate, should stay in the back court. An attempt to serve and volley will virtually guarantee the serving team’s second shot will be played from the transition zone if the net partner cannot pick off the return.

What should the receiving team do? Again, avoid an unforced error is the first priority since most doubles points (in this data set) end with an unforced return error. Aim the ball cross court and get the ball in play. Since an opposing player starts in the finishing zone, once the first priority of getting the ball in play is satisfied, the returner would be wise to avoid that player. The cross court return usually will do that. Having hit a good, cross court return, where should the returner go? Not the transition zone, that’s for sure. The best bet is to stay in the back court, letting the partner move into the finishing zone.

For both the serving team and the returning team, a player who starts back is well-advised to stay back. A partner who starts at the net, or in the transition zone, is well-advised to stay in the finishing zone or to move into the finishing zone. Hey. That’s one up and one back. Maybe that’s why this formation is so popular and so successful. It has a solid foundation in statistical reality.

With one player up and one player back, the responsibilities are divided quite effectively. The player who started back, serving or receiving serve, has the duty to get the ball in play, to keep the ball away from the opposing net player (in the dangerous finishing zone), set up his or her partner in the finishing zone with cross court drives, track down and return any lobs over the net partner, and lob over the attacking opponents if they both choose to come forward.

The player on each team who started in a forward position has the responsibility for staying in or going into the finishing zone and trying to pick off weak replies and put them away. Only the player in the finishing zone has a realistic option of hitting a winner. Therefore, that person should look for and seize opportunities to hit winning shots for his or her team. Most points in tennis are lost and not won, but if they are to be won, the vast majority of the winners will be struck from the finishing zone.

Does all this mean that no player should ever serve and volley in doubles? Not at all. If a player has very good mid-court volleys relative to his or her ground strokes and lobs, or if a player has very good mid-court volleys relative to the opposing returns, then it may be wise to serve and volley. If the opposing team cannot or will not lob, then it also may make sense for a doubles team to get both players into the finishing zone. However, it is probably wise to come in behind a higher shot giving the back player the chance to advance through the transition zone without having to hit a ball there.

Agile players with devastating volleys and smashes may also want to get to the net as often as possible. But if you don’t list hitting scissors-kick overheads, chasing down lobs and hitting winners between your legs among your strongest shots, perhaps you should consider the wrongly-maligned one up and one back doubles formation. If you decide it’s for you, you’ll be in very good company.

No comments: