Are unforced error statistics useful to anyone?

Embed from Getty Images

Every time a professional tennis match finishes, a litany of statistics is proffered by either the TV company or the website of the tennis tour. Some of these could be argued to be useful for supporting technical or tactical development, such as the first serve percentage and the number of points won from it. Others give a picture of the pattern of the match, such as the number of winners each player has hit, which, if there is a significant imbalance, fairly unequivocally identifies the aggressor. There is one statistic, however, which is always present, yet which many feel is at best of limited use, and at worst misleading: ‘unforced errors’. Several coaches are sceptical about the accuracy and value of the unforced error count, and Roger Federer has described the distinction between forced and unforced errors as ‘extremely tricky’. So what exactly does constitute an unforced error?

It is generally accepted that if a player does not have time to set him or herself for a shot, any resulting error is forced. However, if they are in position to play the ball, an error is typically categorised as unforced. In the latter case, there is nothing to reflect the ferocity of the shot they are facing, the spin, or the movement of their opponent (who may, for example, be rushing to the net). It therefore will be classed as an unforced error if you miss from a balanced position, irrespective of whether you are trying to counter an aggressive topspin drive from Dominic Thiem on an uneven clay court, a vicious sliced approach on a grass court from an onrushing Federer, or a gentle moonball from an out-of-position Sara Errani on a smooth hard court! I would contend that only one of these errors would be in any sense unforced- but the statistics do not reflect the situation, just whether or not you made the ball. What this means is that the calculation of the number of unforced errors is reasonably objective- anyone trying to record it should get a fairly similar answer. The fact that a statistic can be calculated easily, however, does not automatically imply that it is helpful!

It is fairly obvious to any regular watcher of tennis that players with differing styles will routinely produce error counts of differing magnitudes. For example, Maria Sharapova will, even on a good day, produce quite a number of ‘unforced’ errors, due to the fact that she hits flat and hard and tries to hit winners. Her shots are risky, but if she makes enough of them she will generally win. On the other hand, smaller, more defensive players like Sara Errani or Simona Halep will generally appear to make low numbers of unforced errors because they try to hit a much smaller number of winners. Crucially, if they are playing Sharapova, they will have to make Maria ‘play the extra ball’ by making difficult defensive shots, in the hope that she will miss. What this means is, when they miss a ball, due to the type of shots Maria hits it will almost always be categorised as a forced error. On the other hand, Maria is faced with the pressure of someone retrieving for all they are worth, giving her a variety of paces and spins, and putting psychological pressure on her by making it very hard for her to finish the point off. Her errors are going to be categorised as unforced, but is this really a fair description? A typical example of this is seen in the statistics from Sharapova’s 6-4, 4-6, 6-3 win over Halep at this year’s US Open. In 221 points, Maria is recorded as making 64 unforced errors, as opposed to Simona’s 14. So, it would appear that Maria is making an unforced error almost every 3 points, yet still winning, whilst Simona makes one approximately every 16 points. I would argue that neither player would be remotely interested in the unforced error count afterwards. Halep will know that she didn’t lose as a result of making roughly one unforced error every two games, just as Sharapova would have been happy to make 100 or more ‘unforced’ errors if she made enough winners to win the match. Similarly, when Federer plays Nadal, Roger knows beforehand that he is going to make more ‘unforced’ errors than Rafa, as his style is far more positive and adventurous. It is highly unlikely that, in the event of a defeat, Roger is going to worry about the ‘unforced error’ count.

So, why do unforced errors still feature so prominently on summaries of match statistics? One reason is that journalists like the idea of players losing for an easily identifiable reason, such as ‘he/she made too many mistakes’. For example, a number of articles attribute Sharapova’s loss to Anastasija Sevastova at the US Open to the number of unforced errors she made. However, this is misleading in many ways. You will recall that her match with Halep lasted 29 games, and she was recorded as making 64 unforced errors. The match with Sevastova was one game longer, although it contained 16% fewer points, and she made 51 unforced errors. It is clear that her error rate was almost identical, and for her, not untypical. It could perhaps be argued that the astute play of Sevastova on the big points made it difficult for Maria, and she was perhaps more likely than normal to miss on these. In any event, it would be very lazy to attribute her defeat to notionally unforced errors, and this would not give enough credit to Sevastova.

Given that the unforced error statistic is misleading, is it simply something that is of use to coaches? Again, I would argue that it is rarely useful even to professionals. Part of being a top player is having the courage to attempt difficult shots and take risks at the right time. These risks will show up as unforced errors if they do not come off- but this should not be used to discourage the player from trying things in future. In general, very few attacking players will benefit from worrying about making mistakes- so how helpful can it be to them, psychologically, to present them with an unforced error count?

Some have suggested that the error count during a match could be split into categories such as mental, technical, etc. This could be useful to coaches, as they would be interested in having a record of when their player was making errors and why. Whilst a player does need to be unafraid of missing shots, careless mistakes and technical flaws can be significant. However, it is questionable whether non-experts could make such a categorisation with any consistency, so this may be something a coach needed to do for their player, rather than something which could be provided in the official statistics.

In conclusion, the basic unforced error count provided in most official statistics is of little use. Furthermore, if coaches focus on this statistic, it could have a negative effect on players by discouraging them from attacking. In the limit, it could even have a negative impact from a spectator’s viewpoint, as players who worry about making mistakes do not produce the greatest spectacle! A more informative measure would look at the reason for the error, and categorise it accordingly- but this would require some techical input, and would be a lot less easy to produce. At junior level, I would suggest that a lot of psychological harm could be done by encouraging players to focus too much on limiting unforced errors. I once worked with a junior player with a Sharapova-like aggressive, flat-hitting style who had been led to think that the way to turn the game around when the momentum was against her was to tell herself to stop making (unforced) errors. Surprisingly, this only made things worse! Overall, I would suggest that positive thinking is a major weapon in a tennis player’s arsenal, and dwelling on unforced errors can only hinder this. Perhaps this statistic should no longer be produced as standard?

 

Embed from Getty Images