Heavy Topspin

Italian translation at settesei.it

Everyone seems to agree that a huge part of tennis is mental. It’s less clear exactly what that means. Pundits and fans often say that certain players are mentally strong or mentally weak, attributes that help explain the gap when there’s a mismatch between talent and results.

Here are three more adjectives you’ll hear in ‘mental game’ discussions: clutch, streaky, consistent. I’ve frequently railed against commentators’ overuse of these terms. For instance, hitting an ace facing break point is ‘clutch,’ in the sense that the player executed well in a key moment. But that doesn’t mean the player himself can be described as clutch. Just because he sometimes performs well under pressure doesn’t mean he does so any more than the average player. Same goes for ‘streaky’–humans tend to overgeneralize from small samples, so if you see a player hit three down-the-line backhand winners in a row, you’ll probably think it’s a hot streak, even though such a sequence will occasionally arise by luck alone.

Some players probably are more or less clutch, more or less streaky, or more or less consistent than their peers, even beyond what can be explained by chance. At the same time, no tour pro is so much more or less clutch that their high-leverage performance explains a substantial part of their success or failure on tour. Most players win about as many tiebreaks as you’d expect based on their non-tiebreak records and convert about as many break points as you’d predict based on their overall return stats. Nothing magical happens in these most-commonly cited pressure situations, and no player becomes either superhuman or completely hopeless.

If you’re reading my blog, you’ve probably heard most of this before, either from me or from innumerable other sports analysts. I’m not taking the extreme position that there is no clutch (or streakiness or consistency), but I am pointing out that these effects are small–so small that we are unlikely to notice them just by watching matches, and sometimes so tiny that even analysts find it difficult to differentiate them from pure randomness.

Still, we’re left with the unanimous–and appealing!–belief that tennis is a mental game. In trying to explain various simplified models, I’ll often say something like, “this is what it would look like if players were robots.” Even though some of those models are rather accurate, I think we can all agree that players aren’t robots, Milos Raonic notwithstanding.

Completely mental

An extreme version of the ‘mental game’ position is one I’ve heard attributed to James Blake, that the difference between #1 and #100 is all mental. (I’m guessing that’s an oversimplification of what Blake thinks, but I’ve heard similar opinions often enough that the general idea is worth considering.) That’s a bit hard to stomach–does anybody think that Radu Albot (the current No. 99) is as talented as Rafael Nadal? But once we backtrack a little bit from the most extreme position, we can see its appeal. At the moment, both Bernard Tomic and Ernests Gulbis are ranked between 80 and 100. Can you say with confidence that those guys aren’t as talented as top-tenners Kevin Anderson or Marin Cilic? Yet Tomic often excels in pressure situations, and Cilic is the one known to crumble.

The problem with Tomic, Gulbis, and so many of the innumerable underachievers in the history of sport, isn’t that they fall apart when the stakes are high. We can all remember matches–or sets, or other long stretches of play–in which a player seems uninterested, unmotivated, or just low-energy for no apparent reason. Even accounting for selection bias, I think the underachievers are more likely to provide these inexplicably mediocre performances. (Can you imagine Nadal appearing unmotivated? Or Maria Sharapova?) In a very broad sense, I could be talking about streakiness or consistency here, but I don’t think it’s what people usually mean by those two terms. It operates at a larger scale–an entire set of mediocrity instead of say, three double faults in a single game–and it offers us a new way of thinking about the mental aspect of tennis.

Focus

Let’s call this new variable focus. There are millions of potential distractions, internal and external, that stand in the way of peak performance. The more a player is able to ignore, disregard, or somehow overcome those distractions, the more focused she is.

Imagine that every player has her own maximum sustainable ability level, and on a scale of 1 to 10, that’s a 10. (I’m saying ‘sustainable’ to make it clear that we’re not talking about ninja Radwanska behind-the-back drop-volley stuff, but the best level that a player can keep up. Nadal’s 10 is different from Albot’s 10.) A rating of 1, at the bottom of the scale, is something we rarely see from the pros–imagine Guillermo Coria or Elena Dementieva getting serve yips. The more focused the player, the more often she’s performing at a 10 and, while she may not be able to sustain that, the more focused player remains closer to a 10 more of the time.

This idea of ‘focus’ sounds a lot like the old notion of ‘consistency’, and maybe it’s what people really mean when they call a player consistent. But there are several reasons why I think it’s important to move away from ‘consistency.’ The first one is pedantic: ‘consistent’ isn’t necessarily good. If you tell a player to be consistent and she hits nothing but unforced errors on her forehand, she has followed your directions by being consistently bad. More seriously, ‘consistency’ is often conflated with ‘low-risk’, which is a strategy, not a positive or negative trait. A player like Petra Kvitova will never be consistent–her signature level of aggression will always result in plenty of errors, sometimes ugly ones, and occasionally in ill-timed bunches. Even an optimized strategy for a highly-focused Kvitova will appear to be inconsistent.

If you’re the type of person who thinks a lot about tennis, you probably see the limitations in my definition of consistency. I agree: The concept I’ve knocked down is a bit of a strawman. If I could do a better job of consisely defining what tennis people talk about when they talk about consistency, I would–again, part of the problem is that the term is overloaded. Even if you mean ‘focus’ when you’re saying ‘consistency,’ I think it’s valuable to use a separate term with less baggage.

Chess

Is ‘focus’ any better than the other mental-game concepts I’ve knocked down? We can objectively measure clutch effects, but it’s a lot harder to look at the data from a match or an entire season and quantify a player’s level of focus.

Nonetheless, I strongly suspect that at the elite level, focus varies more than, say, micro-level streakiness. Put another way: The difference in focus among top players has the potential to explain much of their difference in performance.

I started to think about the importance of focus–again, the ability to sustain a peak or near-peak level for long periods of time–while following last month’s World Chess Championship between Magnus Carlsen and Fabiano Caruana. (I wrote about the chess match here.) Chess is very different from tennis, of course. But because it doesn’t rely on physical strength, speed, or agility at all, it has a much stronger claim to the ‘mental game’ moniker than tennis does. While flashes of brilliance have their place in chess, classical games require sustained concentration at a level that few of us can even fathom. One blunder against an elite player, and you might as well give up and get some extra rest before the next game.

A common stereotype of a chess grandmaster is an old man, whose decades of knowledge and savvy help him brush aside younger upstarts. Yet Carlsen and Caruana, the two best chess players in the world, are in their mid-20s. The current top 30 includes only four men born before 1980. 12 of the top 30 were born in the 1990s, two of them since 1998. The age distribution in elite chess is awfully similar to that of elite tennis.

The aging curve in tennis lends itself to easy explanations: Players can start reaching the top when they hit physical maturity in their late teens, they continue to improve throughout their 20s as they gain experience and enjoy the benefits of physical youth, and then physical deterioration creeps in, beginning to have an effect in the late 20s or early 30s and increasing in severity over time. There’s obviously some truth in that. No matter how important the mental aspect of tennis, it’s hard to compete once you’ve lost a step, and even harder with chronic back or knee pain.

Yet the chess analogy persists: If tennis were mental, with much of the variation between elites explained by focus, the aging curve would look about the same. As modern science has improved training, nutrition, and injury recovery–thus reducing the effect of physical deterioration–tennis’s aging curve has developed a flatter plateau in the late 20s and 30s. In other words, as physical risks are mitigated, the elite career trajectory of tennis looks even more like that of chess.

Thinking ahead

For now, this is just a theory. Maybe you agree with me that it’s a very appealing one, but it remains untested, and it’s possibly very difficult to test at all.

If sustained focus is such a key factor in elite tennis performance, how would we even identify it? The most direct way would be to avoid the tennis court altogether and devise experiments so that we could measure the concentration of top players. I doubt we could convince the ATP top 100 to join us in the lab for a fun day of testing. There is some long-term potential, though, as national federations could do just that with their rising stars. Some might be doing so already; some professional baseball and American football teams administer cognitive tests to potential signees as well.

Unfortunately, we can’t make the best tennis players in the world our guinea pigs. If we looked instead at match-level results, we could try to measure focus using a similar approach to what I’ve done before in the name of quantifying consistency (oops!). My earlier algorithm attempted to measure the predictability of a player’s results–that is, is the 11th best player usually losing to the top ten and beating everyone else, or are his results less predictable? That’s not what we’re interested in here, because by that definition, ‘consistency’ isn’t necessarily good.

We could work along similar lines, though. Given a year or more or results, we could estimate a player’s peak level, perhaps by taking the average of his five best results. (His absolute best result might be the result of an injured opponent, an untimely rain delay, or something else unusual.) That would indicate the level that marks a ’10’ on his personal scale of 1 to 10. Then, compare his other results to that peak. If most of his results are close to that level–like the ‘consistent’ player who loses to the top ten and beats everyone else–he appears to be focused, at least from one match to the next. If he has a lot of bad losses by comparison, he is failing to sustain a level we know he’s capable of.

That sort of approach isn’t entirely satisfying, as is often the case when working with match-level stats. Perhaps with shot-level or camera-based data, we could do even better. Using a similar approach to the above–define a peak, compare other performances to that peak–we could look at serve speed or effectiveness, putting returns in play, converting opportunities at net, and so on. It would be complicated, in part because opponent quality and surface speed always have the potential to impact those numbers, but I think it’s worth pursuing.

If I’m right about this–that tennis isn’t just a mental game, it’s a game heavily influenced by sustained concentration–the long term impact is on player development. Academies and coaches already spend plenty of time off court, talking tactics and utilizing insights from psychology. This would be a further step in that direction.

The mental side of tennis–and sports in general–remains a huge mess of unknowns. As the next generation of elite players tries to develop small technical and tactical improvements in order to find an edge, perhaps the mental side is the next frontier, one that would finally enable a new generation to sweep away the old.