Friday, December 29, 2006

Statistics I'd Like to See

One of my pet peeves is to hear someone state that statistics are meaningless. Anyone familiar with this blog or my website OrangeHoops.org, would easily deduce that I do place value into the numbers of the game. I do not think that statistics tell the whole story, but you can certainly tell a heck of a lot about a player from his statistics, and given the right statistical information you can predict a lot of future performances.

Let’s say you have two college players. Player A scores 25 points a game with 10 rebounds, 7 blocked shots, and 5 assists a game, while shooting 60% from the floor and 85% from the free throw line. Player B scores 9 points a game, with 2 rebounds, 0.5 blocked shots, 1 assist a game, while shooting 35% from the floor and 50% from the free throw line. Assuming no injury is impacting Player B’s statistics, I don’t think anyone would argue that Player B is better than Player A, even though you had not seen either player perform. Clearly, the statistics have some meaning.

Similarly, if Player C scores 17.9 points a game, and player D scores 17.1 points a game, there’s no conclusion at all that can be reached about the value of those two players, other than the fact they both average roughly 17 points a game. In this case, they have little meaning.

I think there are three reasons why a segment of fans disregard statistical evidence. First, is what I would call the Jeter Effect. I plan to write more about this concept at some other time, but suffice it to say it is when you place extreme value on the intangibles of a player and credit him with every good thing that ever happened, and throwing out all statistical evidence because it doesn't measure the intangibles.

The second is a segment of our population is poor at math, and anything to do with numbers to them is inherently bad. There’s really no way to convince these individuals that numbers can be good (there’s probably some of them who would argue that I can’t tell that Player A is better than Player B given my scenario above).

The third reason is that we are often given the wrong set of numbers, or incomplete information. And basketball is full of false numbers. If you don’t look at statistics in the proper context, they give a false impressions. I’m amazed when fans are impressed by a NBA player scoring 40 points in a given game when he’s shot 18 for 48 for the night. Heck, you take 48 shots, you had better score 40 points! Now 40 points on a 20 for 26 shooting effort is quite impressive. Extremely impressive.

When Wilt Chamberlain scored his record 100 points, the overlooked statistic is that The Stilt shot 28-32 from the free throw line that night. That's absolutely amazing, for any player, much less a guy who shot 51% from the line for his career. He was also 36 of 63 from the floor that night, a solid 57%. His 100 point night was a truly amazing night, even for him. Anyhow, I am sidetracked at the moment.

Cuse Country did an excellent job a week ago showing how offensive rebounding can be quite misleading, so I won’t go too much into it here. But simply put, the number of rebounds in a game are a combination of an individual’s ability to get a rebound combined with the number of missed shots. If there were no missed shots in a game, there would be no rebounds. So if a team shots 60% from the floor there will be less opportunity for rebounds than if a team shoots 30% from the floor. And you do need to factor that in, somehow. Unfortunately, to my knowledge, nobody tracks that information in college basketball.


There are other statistics that I think could be useful in providing us all with a clearer picture of what a player has accomplished. And as such, here are the five statistics I would like to see in college basketball.

Earned Rebound Average (ERA): I figure there are roughly 25 missed baskets a game, by each team [We could use any number, but 25 will do; it is all relative]. So in a given game, there are approximately 50 opportunities for a rebound. The ERA would be [number of rebounds] / [missed baskets by both teams] * 50. So if Player A had 10 rebounds in a game where both teams missed a combined 42 shots, then he would have an ERA of 11.90.

Assists In the Paint (AIP): Assists were designed to help show how another player, especially a point guard, has helped set up his teammates for baskets. I would like to see assists broken down a few ways. One way would be to show Assists in the Paint. This is the number of assists a player has to players who have scored while they were inside lane (‘the paint’). I think an assist to a player taking a perimeter jumper is valuable, especially if the point guard has driven into the lane and pulled defenders off the perimeter shooter, but sometimes that feels ‘cheap’ to me. AIP would at least allow us to quantify the different types of assists a player gets.

Free Throw Assists (FTA): We have all seen the spectacular pass to the inside player who goes up for the basket and is fouled hard by the opposition, missing what would have been a sure basket. The player gets to the free throw line, but the player who passed him the ball gets no credit. I think the purpose of the Assist is to measure how well a player sets up his teammates; and clearly, in these situations, he set them up well, and he should be given some credit. I would not tie the statistic to whether or not the player makes the free throws [why penalize a guy because his teammate is poor free throw shooter?]; if the guy gets fouled in the act of shooting, the player passing the ball should bet an FTA. Can you imagine how many FTA Sherman Douglas would have had?

Defensive Plus / Minus (+/-): I’m stealing a concept from hockey here. Defense is extremely difficult to measure statistically. Guys who block a lot of shots often are well out of position and actually playing poor defense, while other guys are holding their ground and preventing the other team from scoring by playing smart defense. That never gets measured. You cannot really look at how a guy does statistically against the guy he is ‘matched up against’. First of all, that has no value in a zone defense… the defender is guarding an area, not a specific player. Second of all, double teams occur, players switch off, etc. A +/- would be the difference in points scored by a team against points given up by a team during the time the player is on the court. Sure, a player would be hurt by being on the court with bad defensive players, but it would give us an additional measurement of how he is doing. A basketball player is 1/5 of the defense when he is on the court, so he’s going to have some impact, plus or minus.

Free Throw Percent broken down by First & Second Effort: We know a guy shoots 70% from the free throw line. But what we do not know is does he usually make the front end of a one and one, or does he miss it. There’s a big difference between shooting 60% on the front end, and 80% on the back end, versus 80% on the front end and 60% on the back end.

Assume two players; both are fouled 100 times in one and on situations. Player E shoots 65% on first effort, 80% on second. He would make 65 free throws out of 100 first attempts (65% of 100), and then 52 out of 65 (80% of 60 second attempts). He would score 117 points and shoot 117 of 165, or 71%.

Player F shoots 80% on the first effort, and 60% on the second. He would make 80 free throws out of 100 first attempts (80% of 100), and then 48 out of 80 (60% of 80 second attempts). He would score 128 points and shoot 128 out of 180, or 71 %.

Both players shoot 71% from the line, but Player F scored 11 more points for his team, which his clearly more valuable.

Now I do not know what the difference between shooting the front end and back end of a one and one are. But I think I would believe there is a difference. The second free throw attempt the player has the confidence of making the first one, less pressure since a second shot is not dependent upon it, and the player is more relaxed since more time has passed since he was actively playing.

I’m sure there are more statistics we could come up with. Let me know if you have any. I would love to have access to these five. With today’s technology, I could go throw line by line of every Syracuse game and derive these numbers myself. But frankly, I don’t have the time. I think it would be revealing. These aren't the answers to everything, but sure do help to answer some questions.

At the least, it would help paint a better picture of what the player was actually doing on the court. Who knows, we may actually find out if Craig Forth was indeed a better player than Jeremy McNeil? Or if Paul Harris or Eric Devendorf is a better two-guard.
RY

No comments: