In the last two weeks I came across two different situations regarding the padding of statistics. Both really made me think about how worthy the PPG (points per game) statistic is in terms of measuring a basketball player. While everyone will agree that an MVP is not determined entirely on scoring, we do tend to hold the scoring title as the most important category of statistic when determining a players skill.
Situation #1
While at the University of Washington home game against Oregon two weekends ago, we saw Washington destroy the young Ducks squad. The first half was a complete massacre as Washington stopped nearly every possession and seemed to score at will. They out ran, out shot, out defended, out rebounded, and overall out hustled Oregon.
The second half was a bit different. The game was clearly over with Washington up anywhere from 15-25 points at all times, but the Husky defense got a bit poor as they relaxed or gambled a bit too much. Then, out of the blue, Oregon’s Junior guard from Detroit, Tajuan Porter, slowly began to accumulate points. He is a quick and capable scorer, but not a superstar by any means. Porter began to drop threes from a few feet behind the arc and he took more and more shots as the second half waned on.
We decided to leave the game a few minutes early in order to make our Valentine’s Day dinner reservation, but even as we walked out the door Porter was still dropping shots from deep. With his last shot before our exit, I heard the comment, “typical of Porter to start scoring now.” In other words, only when the game was out of reach would he truly get his points.
Porter finished with 33 points, including a 10-22 shooting performance with 6-13 from behind the arc, while his team was blown out 103-84. Seeing these stats after the game, I was quite shocked. I felt sure Porter must have been shooting 75% from the field as we walked out the door considering how many shots were going in toward the end of the game, but in reality his standard 45% FGP was nothing to jump at. What this means is that during the first half of the game, when it really mattered, he missed many more than he made. In fact, checking the stats, he went 2-10 from the field in the first half.
What does this say about Porter as a player?
While I am sure many people went home thinking what a good game Porter had with his 33 points, in reality he was merely padding his stats when the game did not matter. The Husky defence was less intense and he had easier opportunities to score. Even more importantly, during the first half, when the game was still in reach, he shot a dismal 2-10. The fact that Porter is 10th in the league in scoring, averaging 15 points per game, makes me think about how good of a player he really is compared to the rest of the guys in the league. He is definitely not the 10th best player in the league. 20th? 30th? It is tough to say.
Situation #2
A favorite author of mine, Michael Lewis, is the author of Moneyballand The Blind Side, both studies analyzing players value based on a non-standard statistical analysis for their sport (baseball and football). His quest to do the same for basketball has been long awaited and he has finally done so in a New York Times article called “The No-Stats All-Star” by studying Shane Battier.
The article is extremely rich and in-depth. It covers everything from the stats guys that teams are hiring these days to analyze the sport to ways to statistically analyze defense. It also explains a bit of Battier’s history, much in the way that The Blind Side went into Ogden’s life. But it was Battier’s offense that I found intriguing.
Much of the article analyses a single game where Battier is guarding Kobe Bryant. In the game, Kobe goes for 30 and Battier has 0 points with a few minutes left in the game. Battier managed to drain a three with less than a minute to go, which I found quite impressive considering how cold he must have been shooting-wise.
It is one thing to avoid shooting to let better players shoot. It is another to shoot until you get hot. But to play as many minutes as Battier did and be able to drain a game-changing shot with 44 seconds on the clock is beyond strange. Impressive, yes. Smart, I’m not sure.
While the Lakers won the game on a last second shot, the article calls Battier’s performance a success due to his defense on Kobe. However, I actually question if this is one of those instances where Battier should have tried to score a bit more early on.
If a coach knows he is going to have a player in toward the end of a game that can defend and shoot like Battier, then that player should be at least comfortable scoring that day. Get him a few points. The first bucket is always the toughest. It should not come when the game matters most.
I would not consider this padding the stats, but Battier is very careful of his stats. One of the tricks I found most interesting is that he actually makes sure he shoots the full-court quarter-ending buzzer beater AFTER the buzzer. He knows the odds of that shot going in and the impact it will have on his shooting percentage. Selfish, yes. Smart, in regards to signing his next contract? Extremely.
Stats Conclusion
There is no end all solution to using statistics to analyze players performance. In my opinion both of these situations are needed to explain that analysis. The stats guys in the Michael Lewis article will argue about how the plus/minus of a player can determine his value. But, what about when a player is racking up points during a blowout, such as in the Oregon game?
I do think that there needs to be more statistics work done in studying and educating players. Every player should have the information Battier has at his fingertips. Whether he decides to use them or not is his call. In the end, however, there are situations beyond statistics that will always be more important for analyzing players ability.
*img source: babble.com
{ 4 comments… read them below or add one }
As for the individual player, the points-per-game statistic is still very, and probably the most important criteria for effectiveness DEPENDING on a few factors.
For instance, suppose you observe game-by-game statistics for 2 players that average 20 a game:
1. Throughout the season Player 1, in every game, has finished with a point total within 6 or 7 points of his average. He hasn’t had any games when he has scored less than double digits, but he has also never scored more than 30.
2. Player 2, however, is very streaky and has some games where he as reached as much as 40, while others he has been held below 5 points.
At first glance, it seems like most would certainly pick Player 1 (steady, effective play) before Player 2 (streaky, explosive) to start a team. I would choose otherwise.
Through my observations, Player 1 is usually the type of player that brings the following mindset to the floor – I’m the team’s top scorer, I absolutely need to shoot the ball 20 times for the team to be effective. This can lead to taking contested shots instead of finding the open man, and being impatient throughout the game when you are not getting your “set” amount of touches. This team is much easier to defend, because the defense knows almost for sure that the leading scorer is going to push to get his 20 points. This leads to ineffective offensive positions and bad shots.
Player 2 is much different that Player 1. He is streaky as a shooter and when he is in the zone, he carries the team on his back. Many times, the lead and momentum is created by a player hitting 2,3, or 4 shots in a row. This is exactly what this type of player provides. The downside to this player is the games when he scores 4 points. However, these are the games were the other players on the team are put into a position to step up and carry the responsibility of scoring. Player’s 1 teammates cannot assume this role effectively, because it is almost guaranteed that Player 1 will try to take over the game instead of embracing and exploiting the defense’s weaknesses.
For example, the Laker’s success the past few seasons can be attributed to Kobe’s transformation from “Player 1″ into “Player 2″.
@ohiobsaketball, Interesting analysis, but I would disagree that Kobe has transformed into a streaky player. When was the last time you saw Kobe go for 4 points?
Kobe, like any other star player has to get his touches and his shot.
I think it would be interesting to see a distribution curve of players points. I’m sure this is something some stats guys could put together (maybe I will for another article). I think it will be very interesting to compare the view of players who average 20 ppg and have a standard deviation of 5 points versus someone who averages 20 ppg with a standard deviation of 12 points.
@Ohiobasketball: Like Conor I think this analysis is interesting however a better player comparison might be Mo Williams for the Cavs. Even still that teams success is enviably driven by LeBron but having Williams makes them a much better team a al the game versus the Miami Heat on 3/2/09 when Williams sparked a come from behind victory.
I read that article too. I have been wondering when Miceal Lewis (and GM’s from other sports) would apply Moneyball-ism to other other sports. One of the things i found interesting is that the Rockets shared the scouting report with Battier before every game AND only with Battier. We know Battier is conscious of stats- taking the last second shot after the buzzer, not taking any shots during the game until the last seconds in a close game, etc. Aren’t the Rockets basically telling Battier through that scouting report, “This is how we want you to play and the stats by which you will be measured.” In fact, I think the Rockets have even put in new and different incentives into player contracts based on the “new stats.” It seems like the Rockets are recognizing that players for the most part want to win and want to play team ball to win but that they are also stats driven for themselves individually. At times those two mesh well together but often they don’t. In response, the Rockets, I think, are trying to drive their players behavior to a more team oriented game by measuring how often players take the right angles, get in the correct position, disrupt a pass or shot (that is not quite a steal or a block), etc. If this is true, I am curious why Battier is the only who gets that special stats sheet before the game. Maybe, you only need one Battier on the floor or on your team. If you have 5 at a time, would you ever score? Maybe a GM wants Tracy McGrady padding his stats PPG-wise and Battier padding his with tipped passes. Maybe Oregon should show Porter some stats that show how a little guy can make it in the NBA by shooting a high percentage in the first half…