ploto
12-13-2007, 10:53 PM
Pretty nice article:
Statistics are like bikinis,'' noted San Antonio humorist Brent Barry said the other day. "They're nice to look at but they don't tell you the whole story.''
With that, Barry, the Spurs' witty backup shooting guard, gave a quick smile and a nod as he headed to the court, confident that in sharing that little bit of wisdom from an old coach, he more or less had nailed the topic at hand.
Statistics in sports are what you make of them, based on the validity of the data collected, the general acceptance of their worth by a specific user group, the means to compare the material not only within the current sample but also against those from the past in a historically accurate way ... and the ability of a player, coach, team executive, agent, sports journalist, fantasy-league operator or bookie to make a buck off of them.
(Bet you thought we were headed into some tweedy, snooze-inducing, statistical analysis "think'' piece, didn't you? Nah, we'll leave that to the wonks and all the Billy Beane and Bill James wannabes infiltrating the basketball world these days.)
Look, stats are great. Stats are fun. Stats span the ages, helping little kids learn long division courtesy of their favorite players' batting averages and helping big kids calculate the risks vs. rewards of that second cocktail or slice of pumpkin pie.
It's cool that Wilt Chamberlain scored precisely 100 points in that 1962 game in Hershey, Pa., not 98 or 99, and even cooler that he averaged 50.4 over an entire season. That same year (1961-62) -- back when Magic Johnson was 2 years old -- Oscar Robertson essentially invented the triple-double by averaging one, start to finish. Heck, add up Robertson's points (11,620), rebounds (3,993) and assists (4,076) from his first five seasons (384 games) and you realize he averaged a triple-double from Eisenhower's last term into LBJ's first full term.
The NBA, like the NFL, the NHL and every other league or sport, can't touch Major League Baseball in its dazzling array of gaudy, astounding or otherwise memorable statistics (or at least couldn't, until baseball blindly or willingly put the credibility of its numbers in jeopardy in the so-called Steroids Era). Still, it is worth remembering that Kareem Abdul-Jabbar scored 38,387 points over 20 regular seasons, this sport's equivalent of Pete Rose's 4,256 hits. It's neat to know that all-time assists leader John Stockton (15,806) set up 53 percent more baskets than No. 2 man Mark Jackson (10,334), a Gretzky-like dwarfing of the competition.
And you probably could win a few bar bets by realizing that Michael Jordan would have had to average 11.1 points or worse in his final forgettable 2002-03 season in Washington to drag his career scoring average below 30.0; as it was, Jordan averaged 20.0 points that year (never falling below that over a full season) and finished at 31.123 points per game, best all time ahead of Wilt's 30.066.
What, after all, is the scoreboard if not a measuring tool of the most important stat? Points matter, more rebounds are better for a team than fewer, Dikembe Mutombo's finger wag was his way of marking an imaginary tote board and, if you're Spurs coach Gregg Popovich, a low defensive shooting percentage combined with a high deflections total can be a beautiful thing.
"Deflections is a really good indicator of how active a team might be defensively,'' Popovich said. "It probably speaks toward steals and stops and those types of things. So yeah, I think deflections is pretty good. All coaches talk about those. And contested shots.''
The bad news for NBA fans is that deflections and contested shots are stats traditionally kept behind the curtain, counted up by coaching staffs and used to educate and inspire players defensively but rarely, if ever, made public. The statistics that are readily available to the average fan are the familiar old box scores, team stats and league leaders, based on decades of newspaper agate type that has changed little since, well, George Mikan roamed the lanes.
The good news is that, just like media options, stats are changing. A sterling example: At NBA.com, box scores for the 2007-08 season now include plus/minus ratings for each player and a category labeled "BA'' for blocks against. Even better news is that deflections and contested shots are being studied this season to see how trackable and reliable they would be, as two more stats worth adding.
"Statistics tell stories that lead people to understand the game,'' said Steve Hellmuth, executive vice president of operations and technology for NBA.com.
The league's official Web site, headquartered in Secaucus, N.J., has far more on its plate than statistical testing and recording -- it is not to be confused with the Elias Sports Bureau or Stats, Inc. -- but it is well-positioned to incorporate the sorts of numbers and trends that players and coaches find most valuable.
In recent years, the NBA has embraced the efficiency rating, an attempt to capture a player's versatility or overall contribution on the court not unlike the NFL's quarterback rating equation. The plus/minus stat, an offshoot of what hockey coaches have used for decades, is intended to reflect a player's (or combination of players') tendency to outscore or be outscored while in the game. Though NBA.com had been featuring the "Lenovo plus/minus'' stats calculator on its Web site, it worked with coaches, including Seattle's P.J. Carlesimo, to tweak the standard in adding it to box scores. Plus/minus ratings are crunch in real time, which means you can check on your favorite team during its game to see how each man is doing.
While it still is not a perfect tool -- garbage time in blowout games can skew reserve players' totals, good or bad -- it can be handy and, what the heck, hurts nothing.
"You really appreciate it when you've seen the game,'' Hellmuth said.
Blocks against? Let's just consider this a stat that is kept because it can be.
"Some guys in the Euro leagues suggested, 'You should have symmetry in your box scores,' '' Hellmuth said. "Since we were recording blocked shots, we could record the blocks against.''
In practice, though, a couple of players in any game will have one or two attached to their name, while most will show zeroes. A guy with four, five or six probably should be sitting on the bench.
"Blocks against just tells you which guys take bad shots,'' Barry said.
The possibilities for more statistical analysis are almost endless, if that's the way basketball wants to go. Some fans have talked of adding the "hockey assist'' -- that is, the pass that leads to the pass that sets up a basket -- because a superb outlet pass or the first pass on a fast break can be the key move in a scoring play. Then again, the game as played in the United States already might be suffering from assists inflation.
"In Europe, if you pass to the guy and he makes a dribble, it's not an assist,'' Minnesota guard Marko Jaric said. "If you pass to the guy and he makes a jump shot, especially from far out, that's not an assist either. An assist, basically, is, he has to lay it up or dunk the ball. That's why, over there, you don't see many guys with more than four or five. Or eight or nine assists there would be like 14 or 15 here.''
"Dribbles taken'' might be a slick way to find out which players yo-yo the ball too much while the shot clock ticks away. "Times doubled'' could be a helpful measure of the franchise guys, especially big men in the low post, although it might be too open to interpretation. It also could confuse Knicks fans, who could mistake it for a stat chronicling all those 64-32 and 89-44 deficits their team has dug.
It would be fairly easy to track "charges taken'' as another hustle-type of activity to motivate defenders, though that could encourage more flopping at a time when the league claims to want less.
"People are already flopping,'' Popovich said. "I don't think they'd be flopping any more.''
With advances in technology and people willing to track and interpret them, additional statistics are cheap and available, if someone wants them. Basketball has always been the jazz of sports, improvisational, five players interdependent. It isn't baseball, with a parade of guys to the plate in search of numbers. So the question becomes, Should someone want them?
"If somebody out there in the audience will talk about it and enjoy it, why not?'' Popovich said. "That's what it's about, having fans appreciate the game and understand what's going on. So if you can make somebody a little more a part of it ... it always feels good for a fan to know what's going on in a coach's head or a player's head.''
Sounded a little surprising, coming from a rather traditional coach. So Popovich relies on statistical breakdowns more than his gut, instincts and experience?
He shook his head.
"It's all a feel to me,'' he said. "There are certain stats that I look at, but I do things more on feel. If you told me that our team was a poor offensive rebounding team, I would already know that without looking at the stats. I watch the film. I look at stats as a helpful icing-on-the-cake sort of thing.
"I'll give you a 'for instance.' Before our last game, we were 27th in the league in field-goal-percentage defense. Now normally we're first, second or third by the end of the year. So that stat would tell you, 'Your defense isn't up to snuff and you've got a long way to go.' Or, 'Maybe it's just a statistical thing because there's not a big sample yet, not very many games.' But we were also third in the league in points allowed.
"Now those two things seem to be statistically opposed to each other,'' Popovich said. "You think about that and you go, 'How the hell can that be?' One stat says you suck and the other one says you're pretty damn good. Those kinds of things interest me. But other than that, I go more by feel for my team and what's going on.''
http://sportsillustrated.cnn.com/2007/writers/steve_aschburner/12/04/statistics/index.html
Statistics are like bikinis,'' noted San Antonio humorist Brent Barry said the other day. "They're nice to look at but they don't tell you the whole story.''
With that, Barry, the Spurs' witty backup shooting guard, gave a quick smile and a nod as he headed to the court, confident that in sharing that little bit of wisdom from an old coach, he more or less had nailed the topic at hand.
Statistics in sports are what you make of them, based on the validity of the data collected, the general acceptance of their worth by a specific user group, the means to compare the material not only within the current sample but also against those from the past in a historically accurate way ... and the ability of a player, coach, team executive, agent, sports journalist, fantasy-league operator or bookie to make a buck off of them.
(Bet you thought we were headed into some tweedy, snooze-inducing, statistical analysis "think'' piece, didn't you? Nah, we'll leave that to the wonks and all the Billy Beane and Bill James wannabes infiltrating the basketball world these days.)
Look, stats are great. Stats are fun. Stats span the ages, helping little kids learn long division courtesy of their favorite players' batting averages and helping big kids calculate the risks vs. rewards of that second cocktail or slice of pumpkin pie.
It's cool that Wilt Chamberlain scored precisely 100 points in that 1962 game in Hershey, Pa., not 98 or 99, and even cooler that he averaged 50.4 over an entire season. That same year (1961-62) -- back when Magic Johnson was 2 years old -- Oscar Robertson essentially invented the triple-double by averaging one, start to finish. Heck, add up Robertson's points (11,620), rebounds (3,993) and assists (4,076) from his first five seasons (384 games) and you realize he averaged a triple-double from Eisenhower's last term into LBJ's first full term.
The NBA, like the NFL, the NHL and every other league or sport, can't touch Major League Baseball in its dazzling array of gaudy, astounding or otherwise memorable statistics (or at least couldn't, until baseball blindly or willingly put the credibility of its numbers in jeopardy in the so-called Steroids Era). Still, it is worth remembering that Kareem Abdul-Jabbar scored 38,387 points over 20 regular seasons, this sport's equivalent of Pete Rose's 4,256 hits. It's neat to know that all-time assists leader John Stockton (15,806) set up 53 percent more baskets than No. 2 man Mark Jackson (10,334), a Gretzky-like dwarfing of the competition.
And you probably could win a few bar bets by realizing that Michael Jordan would have had to average 11.1 points or worse in his final forgettable 2002-03 season in Washington to drag his career scoring average below 30.0; as it was, Jordan averaged 20.0 points that year (never falling below that over a full season) and finished at 31.123 points per game, best all time ahead of Wilt's 30.066.
What, after all, is the scoreboard if not a measuring tool of the most important stat? Points matter, more rebounds are better for a team than fewer, Dikembe Mutombo's finger wag was his way of marking an imaginary tote board and, if you're Spurs coach Gregg Popovich, a low defensive shooting percentage combined with a high deflections total can be a beautiful thing.
"Deflections is a really good indicator of how active a team might be defensively,'' Popovich said. "It probably speaks toward steals and stops and those types of things. So yeah, I think deflections is pretty good. All coaches talk about those. And contested shots.''
The bad news for NBA fans is that deflections and contested shots are stats traditionally kept behind the curtain, counted up by coaching staffs and used to educate and inspire players defensively but rarely, if ever, made public. The statistics that are readily available to the average fan are the familiar old box scores, team stats and league leaders, based on decades of newspaper agate type that has changed little since, well, George Mikan roamed the lanes.
The good news is that, just like media options, stats are changing. A sterling example: At NBA.com, box scores for the 2007-08 season now include plus/minus ratings for each player and a category labeled "BA'' for blocks against. Even better news is that deflections and contested shots are being studied this season to see how trackable and reliable they would be, as two more stats worth adding.
"Statistics tell stories that lead people to understand the game,'' said Steve Hellmuth, executive vice president of operations and technology for NBA.com.
The league's official Web site, headquartered in Secaucus, N.J., has far more on its plate than statistical testing and recording -- it is not to be confused with the Elias Sports Bureau or Stats, Inc. -- but it is well-positioned to incorporate the sorts of numbers and trends that players and coaches find most valuable.
In recent years, the NBA has embraced the efficiency rating, an attempt to capture a player's versatility or overall contribution on the court not unlike the NFL's quarterback rating equation. The plus/minus stat, an offshoot of what hockey coaches have used for decades, is intended to reflect a player's (or combination of players') tendency to outscore or be outscored while in the game. Though NBA.com had been featuring the "Lenovo plus/minus'' stats calculator on its Web site, it worked with coaches, including Seattle's P.J. Carlesimo, to tweak the standard in adding it to box scores. Plus/minus ratings are crunch in real time, which means you can check on your favorite team during its game to see how each man is doing.
While it still is not a perfect tool -- garbage time in blowout games can skew reserve players' totals, good or bad -- it can be handy and, what the heck, hurts nothing.
"You really appreciate it when you've seen the game,'' Hellmuth said.
Blocks against? Let's just consider this a stat that is kept because it can be.
"Some guys in the Euro leagues suggested, 'You should have symmetry in your box scores,' '' Hellmuth said. "Since we were recording blocked shots, we could record the blocks against.''
In practice, though, a couple of players in any game will have one or two attached to their name, while most will show zeroes. A guy with four, five or six probably should be sitting on the bench.
"Blocks against just tells you which guys take bad shots,'' Barry said.
The possibilities for more statistical analysis are almost endless, if that's the way basketball wants to go. Some fans have talked of adding the "hockey assist'' -- that is, the pass that leads to the pass that sets up a basket -- because a superb outlet pass or the first pass on a fast break can be the key move in a scoring play. Then again, the game as played in the United States already might be suffering from assists inflation.
"In Europe, if you pass to the guy and he makes a dribble, it's not an assist,'' Minnesota guard Marko Jaric said. "If you pass to the guy and he makes a jump shot, especially from far out, that's not an assist either. An assist, basically, is, he has to lay it up or dunk the ball. That's why, over there, you don't see many guys with more than four or five. Or eight or nine assists there would be like 14 or 15 here.''
"Dribbles taken'' might be a slick way to find out which players yo-yo the ball too much while the shot clock ticks away. "Times doubled'' could be a helpful measure of the franchise guys, especially big men in the low post, although it might be too open to interpretation. It also could confuse Knicks fans, who could mistake it for a stat chronicling all those 64-32 and 89-44 deficits their team has dug.
It would be fairly easy to track "charges taken'' as another hustle-type of activity to motivate defenders, though that could encourage more flopping at a time when the league claims to want less.
"People are already flopping,'' Popovich said. "I don't think they'd be flopping any more.''
With advances in technology and people willing to track and interpret them, additional statistics are cheap and available, if someone wants them. Basketball has always been the jazz of sports, improvisational, five players interdependent. It isn't baseball, with a parade of guys to the plate in search of numbers. So the question becomes, Should someone want them?
"If somebody out there in the audience will talk about it and enjoy it, why not?'' Popovich said. "That's what it's about, having fans appreciate the game and understand what's going on. So if you can make somebody a little more a part of it ... it always feels good for a fan to know what's going on in a coach's head or a player's head.''
Sounded a little surprising, coming from a rather traditional coach. So Popovich relies on statistical breakdowns more than his gut, instincts and experience?
He shook his head.
"It's all a feel to me,'' he said. "There are certain stats that I look at, but I do things more on feel. If you told me that our team was a poor offensive rebounding team, I would already know that without looking at the stats. I watch the film. I look at stats as a helpful icing-on-the-cake sort of thing.
"I'll give you a 'for instance.' Before our last game, we were 27th in the league in field-goal-percentage defense. Now normally we're first, second or third by the end of the year. So that stat would tell you, 'Your defense isn't up to snuff and you've got a long way to go.' Or, 'Maybe it's just a statistical thing because there's not a big sample yet, not very many games.' But we were also third in the league in points allowed.
"Now those two things seem to be statistically opposed to each other,'' Popovich said. "You think about that and you go, 'How the hell can that be?' One stat says you suck and the other one says you're pretty damn good. Those kinds of things interest me. But other than that, I go more by feel for my team and what's going on.''
http://sportsillustrated.cnn.com/2007/writers/steve_aschburner/12/04/statistics/index.html