In The Numbers Game: Why Everything You Know about Soccer is Wrong by Chris Anderson and David Sally, they wrote about the Maldini Principle, otherwise titled, dogs that don’t bark. For the soccer illiterate, Paolo Maldini is one of the greatest central defenders to play soccer, starring for Italy and Milan. However, despite his acclaim as a defender, Anderson and Sally reported that he made a tackle once every two games. Defenders often are measured or compared by the number of tackles made, with a greater number of tackles reflecting well on the player. How then, did one of the universally-accliamed best defenders make so few tackles?
Positioning, mostly, but also anticipation, game awareness, and a host of other ephemeral qualities that are best categorized as cognitive and perceptual skills. How does Shane Battier continue to take so many charges despite his age and slowing foot speed? The same cognitive and perceptual skills.
Anderson and Sally argued (to me) that scorers tend to be overrated, and defenders underrated (and underpaid) because it is harder to see and measure their skills. The same is true in basketball, as there remains no great way to measure a defensive player’s effect on the game, as blocks and steals are not perfect measures, and not all blocks and steals are equal (a blocked shot that goes out of bounds is certainly not the same as one that starts a fast break in the other direction). More to the point, Anderson and Sally argued that many of the things that happen in a game, and effect its outcome, go unnoticed by statistics. The argument in basketball tends to be that PER and other advanced stats capture these things that go beyond the traditional statistics, even though we cannot necessarily identify them, and to a large extent that is true. However, what are those things that go unnoticed?
The following is a real-life example that I witnessed over the course of a season. This is my memory of the season; obviously, the head coach had a different memory. I did not do advanced statistical analysis, as I had no role in the team, though I did watch every home game.
At Salt Lake Community College (SLCC) two seasons ago, the women’s team had a 6’3 center who became eligible at the end of the first semester. If I remember correctly, nobody in their league played a center much taller than 6’0, so she had a substantial advantage. Nobody else on the SLCC team was over 6’0, so she was clearly the most imposing player on the team. If I remember, she played between 14 and 18 minutes per game, and her stats were unremarkable. She was often substituted for not getting a rebound or missing a shot close to the basket. However, her total effect was ignored by the stat sheet.
When she posted on the block, most teams fronted her and rotated the bottom weak-side help defender into the middle of the key, so it was almost like she was double-teamed before she received the ball (otherwise, teams played a sagging zone that had a similar effect). In one game, due to this defensive deployment, the team’s PF had open shot after open shot from the free-throw line because it was her defender who was sagging in to double team the post. Despite creating open shot after open shot with her mere presence, she was taken out of the game because she wasn’t doing anything. To her coach, her lack of points, assists, and offensive rebounds suggested that she was ineffective and possibly lazy.
The same effect happened on defense. Teams started games driving to the basket or looking for cutters in the key or post-ups. When the 6’3 girl entered the game, the driving largely stopped: Guards did not drive into the key and attempt to shoot over a 6’3 girl. Her mere presence changed the opponent’s game-plan because she was adept at blocking shots. However, as with the offensive side, when the opponent stopped driving, there were no shots to block, so her effect was not obvious unless you noticed the pattern of the opponent’s offense. Inevitably, if a bricked three-point attempt caromed over her head, the coach would take out the 6’3 girl because she wasn’t doing anything.
Her was an instance when a player’s presence on the court changed the game on offense and defense. This is not exactly the same thing as the Maldini Principle, as she showed no special skill of anticipation or reading the game. However, just as Maldini’s effect did not show up in statistics, neither did this player’s. Her effect was creating open shots for teammates or dissuading opponent’s from attacking the basket. In Basketball on Paper, Dean Oliver asserts that field goal percentage (and field goal percentage defense) is the most important statistic in basketball: If you shoot better than your opponent, you are more likely to win. Creating open shots for a teammate, and preventing lay-ups for the opponent, would seem to have a big positive effect on FG% and FG% defense, and ultimately wins. However, because presence does not show up in the box score, and players do not get assists for drawing double teams or blocks for dissuading shot attempts, her positive effect went unnoticed in many games when, in my opinion, she should have played far more than 18 minutes due to the effect that she had on the game.
[For the record, I was not the only one to notice; a rival coach used this to negative recruit SLCC and sign a 6'3 high-school player by pointing out that despite her presence on the court and effect on the game, she played less than 18 mpg.]
Another example based more on skill than size occurred when I watched a player who I had trained previously play with her AAU team prior to her senior year of high school. I grew frustrated watching the coach’s substitution pattern in the first game, so I decided to keep a running +/- for the second game for this player, Jo, and the girl who played ahead of her, Anne. In the first game, I noticed that Jo did all the little things that coaches preach: get back on defense, move your feet, don’t foul, pass ahead to the open player, etc., whereas Anne made the more spectacular, noteworthy plays like driving coast-to-coast through three players. However, I felt that the ratio of her amazing plays to her turnovers while trying to make amazing plays was the reason for the team’s loss.
This was several years, so I forget the exact numbers, but in the second game, their team lost by roughly 10. However, Jo had a +/- of +6-10, whereas Anne’s was a -15+. Again, these numbers are not exact, but the discrepancy was very large. Why did the coach not see this?
I think this is an issue in coaching. We preach that we want certain things, but we reward others with playing time. When I worked summer camps, every well-known coach, whether college or high school, who spoke to the campers preached about the importance of effort, hard work, defense, etc. If you are the best defender on your team, your coach will find a way to play you. I never heard a coach tell a group of campers that if you are the best shooter on your team, your coach will find a way to hide you on defense, but you can name a player like that on nearly every NBA team. I’m not saying that the coaches were wrong or that they were lying to the players; I’m arguing that in a very real way, oftentimes what coaches say they value most, and who they play the most, are at odds with each other.
When I watched Jo play, I noticed, among other things, that she managed to prevent a basket in numerous 2v1 situations. She was an elite soccer player before switching to basketball, and I attributed her performance to her soccer background. Several times in one half, the opponent had a 2v1 against her and did not score even though she was often the shortest player on the court. When it happens once, it could be an offensive mistake; when it happens twice, it could be a fluke; when it happens several times in one half, it is a trend. However, does this get noticed? When she is out of the game, and they give up a 3-point play in transition almost immediately, is it noticed or is the difference attributed to her defense?
Offensively, she did not over-dribble; she passed ahead. This enabled her teammates to get open lay-ups or have a greater advantage when attacking the basket. When she was out of the game, Anne tended to dribble the length of the floor. When she did pass, she passed to players who were not wide open or who did not have an advantage to attack the basket because all five defenders were back on defense. Of course, if they did score off of her pass, she was credited with an assist, and everyone noticed the player who created the shot for the teammate. However, when you take one dribble and throw the ball 40-feet down court to a player who passes to a teammate for an open lay-up, people tend to forget that 40-foot pass that started the break. It is a relatively easy, uncontested pass: anyone could make the pass, so what is there to notice?
These are the types of plays that go unnoticed by statistics but have an effect on the game. As I wrote earlier, PER, +/-, and other advanced statistics capture these types of plays in a general way: Looking at +/-, it was obvious to me that Jo was the more effective player, even though she was shorter and had a slighter build. However, if you only see the noticeable plays, the remarkable plays, or the traditional box score, you might come to a different conclusion. This is the Maldini Principle. Sometimes, it is not what you do, but what you prevent your opponent from doing, whether it is due to mere presence or skill or reading the game or anticipating or positioning or a host of other possible and related factors that are not tabulated by traditional statistics.