TennisOne - A Quantitative Analysis of WTA Doubles: Part 2

In recent years, doubles has seen a divergence from the traditional serve and volley tactics. Many WTA pros serve and stay back at the baseline. The purpose of this article is to examine the effectiveness of modern doubles tactics.

In part I, we examined the effectiveness of net play and positioning. We compared staying back to closing in and compared the effectiveness of different shots. We found that making the first volley is critical to success. We also found that staggered (up-back) formations are effective in many cases and that two at the net doesn’t always win. This study involved 549 points from eight matches on hard courts from 2003 to 2006 with sixteen different WTA players exhibiting serve-and-volley, all-court, and baseline styles. The players included Martina Navratilova, Alicia Molik, Svetlana Kuznetsova, Kim Clijsters, Ai Sugiyama, Virginia Ruano Pascal, Paola Suarez, Conchita Martinez, Tiantian Sun, Ting Li, Elena Dementieva, Anke Huber, Tamarine Tanasugarn, Janette Husarova, Anna-Lena Groenefeld, and Daniela Hantuchova.

In Part II, we will look at serving, returning, and poaching. We will see that the lack of poaching and net play is a result of service placement and positioning. We will look at the down-the-line return and alternate formations. Again, I will keep this quantitative analysis simple and visual.

Some 25.5% of the 549 points in this study ended on the serve or return. That meant 74.5% of the points went at least three shots. Of the points ending on the serve or return, the server won 60.0% of the points. The 28 aces and service winners were cancelled out by 28 double faults.

Perhaps surprising, 17.4% of the second serves were double faults. You might have been taught to get your second serve in, no matter what. But the WTA pros here did not appear to be doing that. Were they being too aggressive? Or just nervous? Or having problems with the wind or sun? The second serve aces and winners only occurred 3.7% of the time so second serves were unlikely to be aggressive as we will soon see.

First serve percentage was 65.6% and the serving team won 60.0% of the points. On the second serve, the serving team won 57.8%. Although the serve was an edge, the difference between the first and second serve was minimal.

The receiver erred on 20.0% of first serve returns but only 11.3% on the second serve returns. Return winners were much rarer: 3.3% on the first serve and only 1.3% on the second serve. Due to the fewer winners and errors on the second serve, it appears that the receivers were taking fewer risks on the second serve and directing the ball with “high percentage” in mind. However, given the few winners on the second serve return, perhaps the receivers were too conservative. If the receivers were more aggressive, they might be able to produce a better payoff.

The 60% winning percentage on the first serve was marginally higher than 57.8% for the second serve. Why the marginal difference? Why not a higher winning percentage on the first serve?

Figure 1 shows service placement using only the regular starting formation (i.e., not I formation). In Figure 1, the percent service distribution or placement is shown in yellow. Only 13 to 19% of the serves were placed down the center or middle T. Furthermore, the most common serve was at the body.

The percent points won on service is shown in red. The three highest winning percentages of 83%, 73%, and 68% were for serves in the center. Yet, the center serve was underutilized. The most effective server to the center in this study was Alicia Molik. The majority of serves from the deuce court were placed in the middle of the service box. The high number of serves at the body -- unless hit with adequate pace and spin -- makes it easy for the receiver to step around and load up for a strong return. Arguably, most serves did not sufficiently pressure the receiver.

Compared to the ATP Tour, most WTA pros lack big, heavy serves. As a result, good returns can relatively neutralize the serve and service breaks occur more frequently. There are exceptions: Serena Williams, Maria Sharapova, and Alicia Molik. Nevertheless, better placement can turn a solid serve into an excellent serve.

Figure 2 compares the service placement of male and female players in singles play (in the quarterfinals, semifinals and finals) at the 2007 Australian Open. WTA pros are shown in light blue and the ATP pros are shown in red. All players, save Rafael Nadal, were right-handed. On the second serve from the deuce court, the men were more than twice as likely as the women (43% to 19%) to serve down the center usually to the backhand. The men were also more than twice as likely to serve wide to the backhand from the ad court on the second serve (44% to 21%). In other words, the men were more likely to try to exploit the weak side of most opponents, the backhand. In the light green boxes are the serves to the body. In all four green boxes, the women’s percentages were considerably higher than the men’s percentages. Clearly, WTA pros did not place their serves to the corners as well as their ATP counterparts.

In comparison of first and second serves, the gap grew as the WTA pros second served 64% and 55% to the body from the deuce and ad courts, respectively, nearly twice as high as those values for the first serve (34% and 31%). Mediocre placement makes the serve less effective. In addition, the lack of the center serve makes poaching less likely.

Specifically, serving trends vary from player to player. For example, Maria Sharapova liked the center serve from the ad court but served many at the body from the deuce court against Serena Williams, Clijsters, and Chakvetadze. Hingis liked the center serve from the ad court against Clijsters but served randomly in the deuce court. Since statistically Clijsters is weaker on the backhand, Hingis should serve more often out wide to the alley from the ad court to open the court better. Although these are singles trends at a specific tournament, they are examples of top players not properly exploiting opponents’ weaknesses. One might argue the women have better backhands than forehands, but when twelve matches from the round of 16 on were examined, the forehands were more effective (errors outnumbered winners by 105) than the backhands (errors outnumbered winners by 140). There are exceptions: Sharapova is looser on the forehand and Serena’s backhand and forehand produce similar results (and she often prefers the backhand). However, their opponents did not regularly serve more to their forehands.

Shoulder strength allows males to serve harder and heavier than female pros. In the 2007 Australian Open (doubles quarterfinals, semifinals, and final), the females served 90.8 and 76.8 mph on the first and second serve, respectively. The ATP pros averaged 109.9 and 94.6 mph on the first and second serve, respectively. However, there is little reason the WTA pros cannot serve as accurately as their ATP counterparts. So how does one explain the numbers in Figure 1 or 2? Perhaps WTA pros have greater aversion to double faulting or taking risks on the second serve. Normal serving at the forehand or body at 90 mph only results in winning 60.0%. Most WTA pros are quick enough to step around to hit a big return. Second serves in the mid-70s at the body is only easier.

Lack of serves in the center means easier crosscourt returns and less poaching. The geometry of serving and poaching is shown in Figure 3. In particular, in the deuce court, most serves were to the forehand or body. We mentioned earlier, even at 90 mph, it was not difficult to step around and hit the bigger shot – usually forehand -- on a serve at the body. If there were more serves in the center, there could be greater poaching. It is possible to poach off a wide serve but more difficult and rare.

On the first serve, 15% and 19% were placed down the center from the deuce and ad court, respectively (See Figure 1). It is generally easier for a right-handed player to place the serve down the center from the ad court. First, a right-hander can naturally place some side spin on the serve which bounces to the left. Second, few WTA pros can use a powerful kick serve down the center from the deuce court. The few who successfully served down the center from the deuce court in this study were Alicia Molik and Martina Navratilova. Molik has a big kick serve that is accurate down the center of the deuce court. Navratilova is a lefty which helps her center serves from the deuce court.

The third reason is that most WTA pros don’t vary their service position enough. They usually stand halfway between the center hash mark and the doubles sideline where it is difficult to accurately place a serve in the center. Often a serve in the center from the deuce court moves towards the body which explains 58% of the first serves and 53% of the second serves were to the body in the deuce court. The only variation in the serving position occurred in the I formation.

How can WTA pros serve more frequently in the center? They could stand 3-4 feet more to the center as shown in Figure 4B. Most WTA pros serve from halfway between the center hash mark and the doubles side line to divide the court coverage equally. But rather than split the court coverage, they might change the possible service stances as in Figure 4B to increase poaching chances. As mentioned in Part I, it is very important to be the team that makes the first volley. In Part I, we saw the highest winning chances were when the server’s partner volleyed. Unless severely angled, wide serves allow more crosscourt returns go by untouched. Many wide serves also pin the server’s partner to covering the down-the-line return.

Moving to the center will help keep the serve to the center and also make more likely the return up through the center. What goes through the center must come back through the center. The server might have to cover more court but as mentioned it is more important to give your team the first volley chance. In addition, the serve out wide can be more effective if the server moves slightly out to the sideline. Variations in the service position can be used more effectively.

In an earlier article, I suggested for club players to use the planned poach. The WTA pros should use the planned poach more frequently as there are too many missed opportunities by the server’s partner. The server if staying at the baseline is also quick enough to cover the possible down-the-line return. The planned poach was rarely used.

Finally on the psychological side, many pros today don’t feel comfortable with volleying and often bypass poaching chances. Even commentators say the volleys are weaker but we have seen, in Part I, that volleys are still high percent plays which win the point. So the WTA players should go with the odds, not their comfort level.

Poaching occurred only on 11.1% (37 of 333 points) of the first serves using regular starting formations. The poacher won 88% of the forehand volleys and 63% of the backhand volleys. Table 1 breaks down poaching scenarios. In blue are forehand poaches and in red are backhand poaches. There were a total of 75 poaching opportunities off both serves and rallies or on 18.1% of the points that went at least three shots (74 of 409 points). If you count the total in Table 1, it is 74 as there was one poach that resulted in an overhead winner.

Chances appeared much higher on the forehand than the backhand poach on either the return of serve or the crosscourt rally. The backhand poach only won 60% but still a majority, either off the return or off a crosscourt rally groundstroke. On down-the-line rallies – which I often call Australian poaching – the poaching team won all 5 points. On the right side of Table 1, the IND column refers to indecisive poaching, which don’t directly result in winning or losing the point. If the poach volley was made, offense was maintained and the point was decided within a couple shots, it was decisive. If the poach volley was made and the rally continued but returned to a neutral scenario, it was indecisive. Indecisive poaches made up 28.4% of the total opportunities.

Clearly, the WTA pros should be looking to create more poaching opportunities, especially on the forehand volley. That involves serving more accurately down the center from the deuce court as shown in Figure 4B. Poaching after serving from the ad court is apparently less effective as the backhand volleys are weaker.

In this study, 67.9% (226 of 333 total first serves) of the first serves using the regular starting formation were returned crosscourt. Even more, 73.5% (114 of 155) of the second serves were returned crosscourt.

In Table 2, we examine the success rate of the down-the-line return in green. Only points using the regular starting formation (e.g., not I formation) were selected. The cases for first and second serves and for the deuce and ad courts are broken down. Interestingly the first serve success was higher than the second serve success. Perhaps it may be due to the receiver having less time which sometimes results in a more automatic decision and or due to the lack of data. The down-the-line return was not particularly successful except in the case (in blue) of the I formation off the serve from the ad court which won 70%. In addition, off the alley or wide serve, the down-the-line success rate was 44% (also shown in blue) which does approach the success rate for crosscourt returns.

Because the down-the-line return was infrequent (i.e., 41 times), the statistical inferiority of the down-the-line compared to the crosscourt is not certain. The crosscourt return did appear more certain and successful at near 50% as shown in red in Table 2. Much of this success is due to the server staying back and the lack of poaching. On the other hand, if poaching was used more often, the down-the-line might be more effective.

An interesting case is the crosscourt return off a serve in the center shown in violet in Table 2. The success rate was only 34% indicating how difficult it is for the receiving team to win off this serve. Unfortunately, we see few serves down the center.

In an earlier article Creative Doubles Patterns, I recommended several alternative formations: the I formation, the half I and the Australian. In Table 3 are the success rates of the alternative formations. Although data is not extensive, there is enough to suggest that these alternate formations may more successful than the regular starting formation shown in blue at the bottom of Table 3 for comparison. The I formation is further broken down in light blue. As expected the I formation is most successful with the serve down the center (75%). The Australian formation and I formation were exclusively from the ad court except for one point with the I formation from the deuce court.

In part, the alternate formations were successful because they were infrequent and may throw off opponents. It is suggested they be used more often but it is hard to say when they might lose their advantage.

In this Part II, we determined that serving patterns could be retooled. Because serves are rarely placed in the center, poaching on the serve is limited. In addition, second serves are too frequently placed at the receiver which doesn’t create pressure on the receiver. Poaching off the serve occurred only 18.0% off rallies that lasted at least three shots. Considering that forehand poaching had 88.5% success, it should be considered more frequently.

The down-the-line return was probably effective off the wide serve into the alley or off the I formation. In most other cases, however, the crosscourt return is more desirable. Alternate formations, used sparingly, may appear more successful than the regular starting formation.

Finally, it is important to note that quantitative analysis requires generation of lots of data. There were 549 points here but some situations did not occur frequently enough to be conclusive. For example, we cannot be definite about the effectiveness of alternate positions. We can, however, begin to ask questions about certain patterns. In addition, the game and players evolve over the years and may change and improve.

Although you are probably not a WTA pro, you might rethink some parts of your doubles game and try out a few new things. It is important not to assume certain pattern are successful but to objectively note what happens. It is also important to note your feelings of success may not be real; we can fool ourselves. After all, tennis points are real and not wishful thinking. The ball is in or out, not almost in.

Doug Eng EdD PhD coaches men's tennis at Tufts University. During the summer, he directs at the Tennis Camps at Harvard. He has received divisional Pro of the Year honors from the PTR and USPTA and several national award.

Doug completed the USTA High Performance Coaches program and frequently runs educational and training programs for coaches. Doug also writes and speaks on tennis and sport science.