Behind the Numbers
Sportswriter Allen Barra is a stats fetishist whose worship of numbers often takes him into the world of fantasy.
11:01 PM, Jan 16, 2002 • By JONATHAN V. LAST
IN THE END, it was Andre Miller who drove me over the edge. Miller is a nice little point guard out of Utah who's in his third year with the Cleveland Cavaliers. He averages 15.8 points and 10 assists per game; he makes good decisions.
But last Friday, the Wall Street Journal's Allen Barra suggested that Miller is the third-best player in the National Basketball Association.
It wasn't the nuttiest thing Barra has come up with. He writes a column in the Weekend Journal (easily the best newspaper section in America) called "By the Numbers," where he pores over sports statistics and tries to divine truths. Barra is part of a larger sect of sportswriters, led by baseball numbers guru Bill James, who in recent years have attempted to de-mystify sports through statistical analysis. It's fun and it makes for good conversation, but it's also frequently ridiculous.
Barra's celebration of Andre Miller wasn't the first time he'd confused a good player with a great one. At last year's NBA All-Star break, Barra chided fans for their selections for the Eastern Conference team, claiming that Miller, Jerry Stackhouse, Marcus Camby, Elton Brand, and Jahidi White--Jahidi White!--should have been All-Stars.
Where does he get this stuff? From the numbers. Barra determines a basketball player's worth by examining his HGS stats. The HGS (HoopStats Grading System) calculates a score for players based on--take a deep breath--(Field Goals Made x 1.4) + (Blocked Shots x 1.4) + (Free Throws Made x 1) + (Assists x 1) + (Steals x 1) + (Offensive Rebounds x 0.85) + (Defensive Rebounds x 0.5) - (Turnovers x 0.8) - (Field Goals Missed x 0.6) all divided by minutes played, and then multiplied by 48.
Barra places his faith in the HGS alone, because, as he says, "If you are doing something to make your team win, you are leaving a paper trail that will reveal it." Which, of course, is quite insane. For instance, all of Barra's "overlooked" All-Stars last year, with the exception of Camby, played for cellar-dwelling teams. And I could go on and on and on about how the HGS favors big men because it places so much emphasis on shooting percentages and blocked shots and counts two-pointers and three-pointers the same way. But I won't.
The purest proof of the HGS's worthlessness is the results. At the end of last season, Barra wrote a paean to Shaquille O'Neal asserting that Shaq was the obvious pick for Most Valuable Player because he had the highest HGS score. And HGS aside, you could make a reasonable argument that he was the MVP, since his Lakers won the NBA championship, but it's worth looking further down Barra's MVP list: In third place is Karl Malone, in fifth is David Robinson, ninth is John Stockton, eleventh is Elton Brand. Andre Miller clocked in at thirteenth.
Malone couldn't get his team out of the first round of the playoffs, even with the help of Stockton. Robinson is over the hill. Brand, a promising young forward, was on a Chicago Bulls team that finished the season 15-67, raising the question: His marvelous HGS score aside, how valuable was he? Would the Bulls have lost 68 games without him?
And absent from Barra's list entirely was Allen Iverson, the Philadelphia '76er shooting guard who led his team to the second-best record in the league and the NBA Finals. Barra dismissively noted that Iverson didn't even crack the top 15 HGS scores. Iverson went on to win the MVP award.
In another telling column, Barra argued that Hakeem Olajuwon, the great Nigerian center, is the best player in NBA history, saying, "With a switch of supporting casts, Mr. Olajuwon's team probably would have won about the same number of games, and probably the same number of championships as Michael Jordan's."
Statistically, Olajuwon may be the best center of the modern era, but to argue that the talent of his Rockets teams was inferior to Jordan's Bulls is nonsense (see "The Third Coming"). After Jordan's retirement, the Bulls were exposed as a group of men barely capable of making an NBA roster, while Olajuwon's teammates, particularly Robert Horry and Sam Cassell, have gone on to great success without him. In other words, Jordan's supporting cast only looked great because he made them great.
But Barra gets even nuttier with his Olajuwon numbers, saying, "I will suggest this: Mr. Olajuwon, cloned five times, would beat five clones of anybody else." While it's impossible to test Barra's argument, it seems theoretically obvious that the most successful team of NBA clones would consist of power-forwards who could handle the ball and run the court, like Kevin Garnett, not centers like Olajuwon. Either way, it's a strange concept that Barra seems to give a lot of weight to. In last week's column, he noted that "if you multiply a player's HGS by five, you'll get the theoretical points per game for a team made up of players of similar skill." Shaq's HGS is 37.02, meaning that in Barra's world, 5 Shaqs would score 185 points per game. I rather think they couldn't get the ball past mid-court.
Barra's numbers fetish is so all-consuming that we can't be far away from his arguing how badly Michael Jordan is performing. Last October Barra wrote a column positing that Jordan's comeback was doomed. If Jordan played "his highly optimistic 30-32 minutes a game, and his percentages continue to drop," Barra lamented, he would only shoot 43.5 percent from the field and 75 percent from the free-throw line, with his steals and rebounds dropping as well. Which, to Barra, spelled disaster.
All of his dire predictions are coming to pass. Jordan is playing 36 minutes a game and shooting 40.8 percent from the field. His free throws are a little better than expected, at 80 percent, but his rebounds are off from his last season in 1998, from 5.8 per game to 5.2 per game, as are his steals, from 1.7 per game to 1.56 per game.
But despite all of that, Jordan has turned his team into a winner. Last season the Washington Wizards won 19 games. This season, with the addition of Jordan and not much else (except Brendan Haywood and a lottery pick who doesn't play), they are on pace to win 42 games and go deep into the playoffs. Although the season is young, Jordan seems a lock for MVP.
Excellence in sports is about more than numbers. It's about subtleties, like drawing double teams, and intangibles, like leadership. At the start of the NFL season, Barra called the Denver Broncos' Brian Griese "the AFC's best quarterback," and noted that according to "our Pass-Efficiency Rating, he has better efficiency numbers than John Elway ever had." Everyone in the room who would take Griese over Elway to QB a game that their life depended on, raise your hand. But to Barra, it's always obvious. Just check the HGS or the Pass-Efficiency Rating or the SloB (Slugging times On Base).
By reducing athletes to numerical automatons, Barra takes sides against the Great Man theory of sports: At their highest levels, sports are about men doing things that are beyond themselves. It's why the Packers have never lost a playoff game at Lambeau Field and the Tampa Bay Buccaneers are winless when the temperature goes below freezing. It's why the Yankees beat the Mariners and the A's in the playoffs last year. It's why Jordan is a winner and Olajuwon is just a great center. It's why Iverson was the MVP and Miller is a role player.
If you get too caught up in numbers, you lose the ability to recognize even the most elemental truths. Last August Barra wrote: "If the Cubs stay sharp enough in the season's final two months to get the home-field advantage for the playoffs, that would make them the NL's favorite." Well, it probably made sense on paper.
Jonathan V. Last is online editor of The Weekly Standard.