12/15/2004 04:26:00 PM|||Kurt|||Not long ago, Dean Oliver, author of a watershed basketball numbers book, "Basketball on Paper,” was hired as a consultant with the Seattle Supersonics. The move was reminiscent of Bill James — a man who had for decades been thinking outside the baseball box — becoming a Red Sox consultant. I throw that out there not to point out the obvious winning trends in both places, but rather to point out that the use of detailed, relatively complex statistics to measure player performance is something becoming part of the sports establishment. Baseball’s new wave of statistics (well, a bunch go way back to Branch Rickey, but that’s a story for another day) started coming into the light when the book “Moneyball” came out and shot up the best sellers list. The new wave of basketball stats are not nearly as well known — PER has not come near the level of OPS — but their day is coming. And, if I and some other bloggers can push that envelope along, all the better. My use of those stats in this blog has led to a host of questions to me (well, host may be an overstatement) along the lines of “what is going on here?”. So, following is a series of definitions of the stats you’ll see here most, with links to longer definitions and equations. (If that’s your thing, start here.) First, one more thought. In the same way that baseball statistics are valid because of the number of trials — a starting major leaguer gets more than 500 at bats per season, for example — basketball stats are also valid. There are 82 games a season and that means most teams have upwards of 7,400 offensive possessions per season. A star player like Kobe Bryant may take upwards of 1,100 shot per season (1,178 last year) but even a role player, such as Chris Mihm last season in Boston, will take hundreds (381). The volume is there to make valid analysis. PER, or Player Efficiency Rating. This statistic is one you’ll see a lot and comes from NBA stat guru John Hollinger, who has his own site and puts out a Basketball Prospectus book every season. I’ll let him explain PER (from his site):
The formula, which I call the Player Efficiency Rating (PER), adds the good (made shots, steals, assists, rebounds, blocked shots, free throws), and subtracts the bad (missed shots, turnovers, fouls) by assigning a point value to each item (I arrive at the point values in a fairly tortuous way, and that's one of the parts I'm saving for the book). The rating for each player is then adjusted to a per-minute basis (so that, for example, you can compare subs with starters in the frequent 'he should start ahead of so-and-so' debates), and also adjusted for the team's pace. In the end, one number sums up the players' accomplishments (the statistical ones, anyway) for that season. I've set it up so that the league average, every season, is 15.00, which produces sort of a handy reference guide: A Year For the Ages: 35.0 Runaway MVP Candidate: 30.0 Strong MVP Candidate: 27.5 Weak MVP Candidate: 25.0 Bona fide All-Star: 22.5 Borderline All-Star: 20.0 Solid 2nd option: 18.0 3rd Banana: 16.5 Pretty good player: 15.0 In the rotation: 13.0 Scrounging for minutes: 11.0 Definitely renting: 9.0 On next plane to Yakima: 5.0
PER is a very good measure of what a player contributes offensively, and that can also be extrapolated to what a team gets from a position on the floor. I think the best evidence of how well PER works is to see who is at the top of the list this year, so far: NAME............ PER Tim Duncan...... 30.51 Kevin Garnett... 30.32 Dirk Nowitzki... 30.05 Amare Stoudemire 29.14 Dwyane Wade..... 26.82 LeBron James.... 26.40 Kobe leads the Lakers with a PER of 23.7, a number held down by his shooting percentage and turnovers. oPER or Opponents Player Efficiency Ratings. This is the same as PER, but done to whomever a player is guarding to give you an idea of the defensive performance of a player. This is a flawed system when applied to one player because of zone coverages, switch offs and the rest of the way defense is played in the NBA. That said, it does work to give you a general idea of what a player is doing. However, oPER it works pretty well when applied to a position. For example, the Lakers this year are allowing an oPER of 10.9 against opponents shooting guards, well below the league average of 15. However, they are getting burned at the three (18.9) and the four (18). This basically matches up with what we see — Kobe (playing 88% of the Laker minutes) is shutting down two guards, but inside the Lakers are no match defensively. eFG% or Effective Field Goal Percentage. The problem with basic/traditional field goal percentage, particularly when talking about guards, is that it counts a made three-pointer the same as a made two pointer — the equivalent of a football stat that counted touchdowns and field goals as the same. Clearly those are not worth the same amount, and neigher is a two=ponter and a three-pointer in basketball — a three-pointer is worth 50% more on the scoreboard. So eFG% gives players that bonus (50%) for making the more difficult shot (the equation can be seen as eFG% = (2PM + 1.5*3PM) / FGA or eFG% = (FGM + 3PM/2)/FGA). It’s not a perfect measure, but it gives you a better idea of how a player is really shooting, if they are taking a number of three pointers. eFG% can also be applied to how well a player or team is doing defensively. Roland Rating. This one is a measure of a players value to his team and is easy to find at 82 Games. The explanation:
The best gauge of a player's worth to a specific team comes from looking at the difference in how the team plays with the player on court versus performance with the player off court. The on court +/- number represents the team's net points with the player on the floor per 48 minutes, while the off court number is the team's net with the player off the floor per 48 minutes. The Roland Rating is the difference between the two, with a positive number indicating the team has played better with the player than without.
For example, so far this season Kobe leads the Lakers with +21.2 followed by Jumaine Jones (+16.2) and Brian Cook (+5.6). What’s interesting is that with Kobe on the court (averaged out for 48 minutes) the Lakers are only +2.3, but with him off they are -18.9. Points Per Possession. I haven’t used this too much yet, but it's a favorite of basketball sabermetrics people, so you will as the season wears on. It is just what it sounds like — the average points per possession a player or team gets. This is always worked out to an averave of 100 possessions. A thought from 82 Games:
An accepted formula for calculating possessions amounts to Field Goal Attempts minus Offensive Rebounds plus Turnovers plus Free Throw line trips earned. We believe there is some inherent unfairness in this scheme on many levels (i.e. a bad pass costs you, an assist gets you nothing...or giving free possessions to prodigious no-shoot offensive rebounders).
Top Five-Man Floor Units: This uses the +/- system applied to various five-man units, and how they perform against the five-man units they are out against. Part of this is to create a percentage that is the number of times (out of 100) they have bested the opposing five. There are other stats out there, applied to passing and the like, but lets save those for another day.|||110315680285802271|||What Are Those Numbers?