Thursday, January 1, 2009

Bruce Bowen and the Evolution of the Box Score

Bruce Bowen's opportunity to win the Defensive Player of the Year Award is past him. The league's premier defender for the bulk of this decade is past his prime. He's still a great defender, but his minutes are down and his feet are a little slower than this time last year. And while it's a safe bet that the Spurs will retire his number, it's doubtful that many outside the Spurs faithful will remember his name in 20 years.

There are at least three reasons such accolades eluded the defensive ace. The first is that Bowen's reputation as a dirty player did not curry favor with end of the season award voters. The second is the tremendous play of other great defenders, such as Ron Artest, who edged Bowen out year after year. The final reason amounts to conjecture, but I'm reasonably confident in it. Some of those who vote on the DPY award are careless and uninformed. If those words seem harsh, please let me explain.

Bruce Bowen is an 8 time All-Defensive Team recipient. On 3 occasions, Bowen has finished as runner up in the DPY voting, losing twice to Ben Wallace and once to Marcus Camby. Staggering facts when cast against the eye-poppingly underwhelming backdrop of his career stats: 2.9 rebounds, .85 steals, and .40 blocks per game. The league's premier perimeter defender is not known to fill up a box score. This, in a nut shell, has deterred voters away from Bowen and into the arms of other dance partners, whose numbers are all glammed-up and bursting out of their evening gowns. In 2007, Marcus Camby's 13.1, 1.1, and 3.6 were attention grabbers, even if, after hindsight examination, there were better looking women at the party.

Unlike Camby, who defensive excellence is showcased in brilliant weakside help, the beauty of Bowen's game is his ability to stick with his man, unphased by screens and whistles, contesting every shot attempt along the way. Bowen is the classic "intangibles" player. What he does doesn't show up in the box score. John Hollinger writes,
Bowen is still one of the best perimeter defenders in basketball; in terms of moving his feet laterally to keep dribblers in front of him he's as good as anyone in the past two decades. Bowen doesn't block shots or get steals at a high rate, but he continually forces opponents into difficult shots and then uses his reach to get a hand in his man's face.
The limited representation of defensive statistics in box scores extends beyond my sympathy for Bruce Bowen, however. Despite what you may think of him, there is much in Bowen's game that is worthy of imitation from the young. Younger players, those in middle or high school, let alone college, won't model their games after Bowen. Why? The basketball universe neither provides the necessary windows through which young players can appreciate what Bowen does nor accessibility to the concrete measurements such players would need to model themselves after Bowen. Statistically speaking, their eyes are elsewhere.

Unlike past seasons, I find myself gravitating away from ESPN box scores and toward Yahoo! or NBA.com stat lines. Why? The latter two sites provide a slightly better box by including +/- and blocks against statistics. This is barely better than what ESPN provides, but it is better. Frankly, I prefer to land on ESPN.com and stay there, but it's not always possible. In fact, on a given day I'm likely to supplement my daily dose of John Hollinger with 82games.com, The Popcorn Machine, Basketball Prospectus, and the Wages of Win Journal. In addition to these, I really like the Comprehensive Stats page offered by Draft Express, except that it only covers prospects and is infrequently updated. But in general, DX uses a format that deserves wider imitation.

As you can see, that's a lot of surfing. Too much, really. Someone is going to have to bite the bullet and reinvent the boxscore(s). The odds that a 15 year old lands on any of those supplemental pages is significantly lower than his perusing the box score at ESPN. It's a pain to trek all over the net to piece together what is increasingly basic statistical information. It's a problem with a relatively easy solution.

Young fans and future basketball players aspire toward numbers because they are inspired by the numbers they read after each game. This is basic psychology. "Hey, man, did you see that Dwight Howard put up a 30-20? I'm gonna do that." This is far more common to hear than a comment along the lines of, "Did you see that Bowen held McGrady to 4-16 and forced him into two offensive fouls?" It would only benefit the world of basketball if young players had more measures to aspire toward, especially helpful defensive measures. In this way, the NBA has some obligation toward box score reinvention--stat hawking plays a large role in the formation of young players, not to mention its contribution to understanding the game. If basketball is religion than its moral values are stats. If, as one example, charges drawn showed up in the box score, players would listen more intently to their coach when he was explaining defensive rotations. But that, of course, would also cause Joe Player to think more deeply about defensive schemes.

It's a new year. What better time is there to suggest that the NBA and its media partners accelerate their movements toward a more comprehensive set of box scores? What follows falls more into the category of suggestive brainstorm than impassioned plea, and I invite our readers to provide their thoughts in the comments under this post. What should the NBA and its media partners consider? To start:
Beyond this, a media partner like ESPN might include Hollinger stats with their box scores or game flow charts.

I realize that some of these things are of zero interest to certain fans, but all the information they are accustomed to sourcing would still be included in new and revised boxes. Although, I suspect that if "new" stats are noticeably included with the more traditional measures then the interest in such things would would climb rapidly.

Usage and efficiency box scores are nice for die hard fans and arm chair GMs, but the real trick is a more detailed basic box, which could be used at all levels of basketball from grade school up and, perhaps, agreed upon internationally. Already standard stats of minutes played (min), field goals made/attempted (FGM-A), three pointers made/attempted (3PM-A), free throws made/attempted (FTM-A), field goal percentage (FG%), three point percentage (3P%), offensive rebounds (OREB), defensive rebounds (DREB), total rebounds (REB), assists (AST), steals (STL), blocked shots (BS), turnovers (TO), personal fouls (PF), and points (PTS) would remain. What could be added to this list to helpfully inform and shape the basketball community? Here is a list of suggested categories I'd like to see considered for a reinvented basic box score:
  • Points in the Paint (PP): self-explanatory and common, PP allows fans the ability to not only gauge a players ability to score within the heart of the defense on high percentage shots, but it also gauges his ability to get there, either by establishing early position near the hoop or by beating the perimeter defenders on dribble drives.

But the statistic du jour in the N.B.A. - the plus-minus statistic - offers an opportunity to remedy many drawbacks of the traditional box score, even if it is limited. More teams are now using plus-minus as an important part of their front-office work.

In their most common form, plus-minus statistics measure how a team performs when a given player is on the floor versus when he is on the bench. When Stephon Marbury is on the floor, the Knicks lose by 0.1 points per 48 minutes. Put Marbury on the bench and the Knicks lose by 12.7 points. Marbury's net impact is a positive 12.6 points per 48 minutes.

Because plus-minus statistics depend on team performance, they capture almost every contribution a player can make. The good pick, the solid help defense, the threat of a 3-point shooter - all these contributions are captured by net plus-minus statistics. In theory, they are an ideal measure of a player's effectiveness.

  • Gretzskys (GRZ): The hockey assist. The pass before the pass that leads to the basket. This is an extremely helpful measure in determining a players overall contribution to sharing the ball with his teammates, or, as some like to say, "playing the right way." In terms of the ability of statistics to shape young players, here is one quantifiable way to encourage the proverbial "extra pass."
  • Points Assisted (PAST): Points assisted is a subtle, but useful alternative to the standard dime. First, it would reflect total points created by assists, including the "extra" point that comes with assisting a 3 pointer and points scored of free throws made from And-1 situations and passes that would have been an assist, if not for an intervening foul. Put differently, it would give a sense of things like effective kick outs, ball reversals, and timely, well placed passes on post entries and fast breaks.
  • Deflections (DFL): With this stat, we start to come full circle in terms of Bruce Bowen. Basic stats need to do a better job of accounting for defense, and here is a start. How many times did a player disrupt the opponent's offense by getting a hand on the ball? Among other things, this is a simple indicator of defensive energy and activity.
  • Offensive Fouls Drawn (OFD): 82Games.com has summoned the call here and everyone would do well to follow their lead. How many times does a defender draw an offensive foul, either by staying in front of his man or rotating in front of someone else's? To my mind, this is a must have defensive measure. Take one subset example: charges drawn. Charges drawn is more helpful as a defensive measurement than, say, steals, which we've tracked since before bread came sliced. Players (see Iverson, Allen or Hughes, Larry) often rack up steals by abandoning their man and rushing a passing lane, or by wandering outside the mandates of their defensive schemes. These bad habits looks nice on a stat sheet. Unfortunately, the stat sheet doesn't record the common result of leaving your man to rush a passing lane--namely, an easy basket for the other team as the abandoned offensive player cuts directly to the hoop. OFDs would actually measure the opposite, the good defensive practice of staying with one's man and playing within the defense.
  • Defensive Fouls Drawn (DFD): In other words, how often does a player force the defense into fouling them? Keeping this Spurs-related, we've touched on this as a key to George Hill's early success.
  • Opponent's Field Goal Percentage (OFG%): This is the Bruce Bowen stat, and is not to be confused with team field goal percentage. What shooting percentage does a defender hold his man to through out a game? In other words, if Tracy McGrady shoots 3-8 while Bowen defends him but improves to 3-4 against Roger Mason, we should have some way to quantify this. Over his career, Bowen has achieved defensive excellence by forcing his opponents into difficult, well-contested shots. Or, as is often the case, Bowen bothers his opponent by not ceding ground. So, if Rip Hamilton likes to shoot off a screen and curl at 17 ft., Bowen occupies the space at 17, forcing Hamilton into a curl and catch at 19 ft. instead. The combination of this and OFDs would give us a way to accurately quantify individual defense.
  • Adjusted Game Score (AGS): This is a forthcoming John Hollinger stat that basically amounts to an unconsolidated, game by game PER. How efficient was Player X in only tonight's game? If LeBron James put up 40, 10 and 10 against the Knicks, what would his PER be for that game? This is the adjusted game score. Helpful stuff from John Hollinger, as usual.
I understand that there are problems with some of these stats or that there may be other or more helpful ways to quantify contributions, which is why I invite our readers to critique these suggestions or add their own in the comments. For example, opponent's field goal percentage coud be a misleading indicator if the sample size was isolated. In other words, a player's percentage for a given game might be down because of injury or a funk and have little to do with his primary defender. So, it's a stat that would require the backdrop of a large sample to be a reliable indicator. But this is the case with most numbers. Charlie Villanueva is not a likely candidate to drop 50, despite one random sign.

Would Bruce Bowen have won DPOY if some of these stats were available at the height of his powers? Maybe, maybe not. Sometimes advanced stats are as surprising in what they disprove as what they prove. I certainly think that players like Bruce Bowen, Raja Bell, Kurt Thomas and Anderson Varejao would be more appreciated if we had better measures of their unique contributions.

Whatever the case, when Bowen finally hangs it up, the Spurs faithful will remember him as a game-changing cog in the championship machine. Bowen's defense was dominant and indispensable, even if there are not measurements available to substantiate the claim. Last season this coffin was nailed shut for me during the Conference Semifinals against New Orleans. The Hornets had spanked the Spurs in 800 kinds of ways, winning by 18 and 19 to take a 2-0 series lead in the opening games. In those games, Peja Stojakavic torched the Spurs for 47 points (60 FG%, 63 3pt%). Bowen was guarding Chris Paul during each of those contests. For the remaining 5 games of the series, Popovich placed Bowen on Peja and the tide of the entire series turned. Over the balance of the series, Stojakavic scored only 43 points (45 FG%, 33 3pt%)---a drop off of nearly 15 points per game and a staggering shooting percentage shift of minus 15 and 30%, respectively. In the first two contests Stojakavic averaged 12 shot attempts per game as compared to only 7 in the back five. Not only did Bowen dramatically reduce his man's shooting percentage, he played him so tight that it was difficult for Peja to simply find space from which to shoot.

Some smart readers will ask, "How can we know for certain that Bowen guarded Stojakavic on these shot attempts? How do we know that he didn't miss shots against Mike Finley on switches or that Ime Udoka was responsible for chasing the 3 point specialist off the arc?" Good question. We can't, and that's sort of my point. But for the record Bowen hounded Stojakavic for the majority of those shots, and he was remarkable. In one sense, he was the series MVP. It's a shame that there aren't better tools to aid our appreciation and memories of these moments.

(HT: Much of this post is collected from memory, and I've appropriated many of the suggestions from elsewhere, whether article, blog post, bar stool, or message board thread. I'm pushing a conversation, but the talking points are not my own.)

1 comment:

Unknown said...

tim, why do you say that draftexpress only has prospects and isn't updated frequently enough? there is a page for every player in the NBA, NCAA and every European league covered. The stats are also updated on a nightly basis.