Hockey stat, with a twist, useful in NBA, too
It never stops. Both fans and teams constantly crave more ways to measure how players are performing. Thus, it was perhaps inevitable that a "catch-all" stat would appear on the horizon, and recently one has.
It's a powerful tool with a simple premise: Add all the points the team scores when a player is on the court, and subtract all the points the team allows when he is on the court. Subtract the latter from the former, and you end up with the player's "plus-minus" – how many points better or worse (i.e., plus or minus) the team is with that man on the court.
As that dejected lot known as hockey fans can tell you, NHL teams have used plus-minus for several years to figure out which players have the greatest impact. In a sport where scoring is so infrequent and defensive evaluations are so difficult, plus-minus is both easy to track and, once you know how to interpret it, highly informative.
For years, basketball aficionados hoped for something similar, but because scoring is so much more frequent in basketball, tracking plus-minus for every player over a full season was an incredibly laborious process.
Fortunately, the information age has arrived, and with it has come folks like those at 82games.com. They've computerized all the info in play-by-play sheets and churned out plus-minus marks
for every player in the league.
As hockey fans already know, plus-minus by itself isn't very telling. The leaders in this category are all the players on the best clubs, while those at the bottom saw a lot of minutes on awful teams. In other words, the quality of a player's teammates does more to determine his plus-minus than the quality of the player himself.
Nonetheless, we can glean useful information by comparing a player's plus-minus when he's on the court against his plus-minus when he's off it.
For example, Chris Andersen doesn't appear to have been a valuable player for the New Orleans Hornets this year based on his on-court plus-minus, which is -4.1 per 48 minutes. However, compare that to when he's off the court, and it's clear that Andersen is a fairly important player, because the Hornets plummet to -8.9 when Birdman is grounded. Using that comparison, Andersen's net plus-minus becomes a +4.8, which means he's been a lot more valuable than the average Hornet.
Of course, the key word there is "Hornet." On any team, about half the players will have a net plus-minus above zero and half will be below, because they're only being compared to each other. That makes it difficult to compare the net plus-minus of two players on different teams.
Moreover, even on the same team we can find an interesting phenomenon, one I call the Charlie Huddy Effect. For those who don't recall the immortal Charlie Huddy, he was a solid, unspectacular defenseman for the legendary Edmonton Oilers hockey teams of the mid-1980s. That squad had two of the greatest players in NHL history, Wayne Gretzky and Mark Messier. But because Gretzky and Messier played on separate lines, and hockey defensemen stay on the ice longer than other players, Huddy was about the only Oiler who got to play with both great players. As a result, in 1982-83 it was the modestly talented Huddy, not Gretzky or Messier, who led the league in plus-minus – even though he was nowhere near as important or talented as his two Hall of Fame teammates.
This happens in basketball, too. Look at the Miami Heat , for instance, and you'll notice that Damon Jones 's net plus-minus is better than Dwyane Wade 's. Is this because Jones is better than Wade? Of course not – it's because of Stan Van Gundy's substitution pattern. Because Jones is usually the first starter to get subbed, he is always on the floor with both Wade and Shaquille O'Neal . But because Van Gundy wants one of his stars on the court at all times, Wade isn't always on the floor with Shaq. In fact, because Wade plays so many minutes, he sometimes shares the court with four substitutes. In other words, Wade often has to share the floor with bad players while Jones plays almost exclusively with great ones.
Finally, there's one other important effect to consider when viewing net plus-minus, and I call it the Backup Effect. Net plus-minus looks not only at a player's on-court plus-minus, but also his off-court plus-minus. That second data point is heavily impacted by a variable outside the player's control: the quality of his backup. If a player has a terrible backup, then his net plus-minus will look good because the team will play so poorly with the sub on the court. Similarly, players with high-quality backups won't show nearly the same disparity.
A good example of how this works in practice is with Stephon Marbury . When he's on the court, the New York Knicks are a break-even proposition, scoring just as mujch as their opponents. When he's off it, however, the Knicks are worse than the Hawks, getting outscored by a whopping 12.2 points per 48 minutes. The reason, as you might have guessed, is that New York sends in CBA talents when Starbury checks out. So he can thank Moochie Norris and Jamison Brewer for that +12.2 net plus-minus of his, which ranks 11th in the league.
So those are some of the limitations of plus-minus, but let's be clear: This still is a very powerful tool. Perhaps the best application is evaluating players on the same team who play similar positions. That's especially true with a team like the Memphis Grizzlies that tends to use a set substitution pattern and alternates almost exclusively between two players at a given position.
Glancing at plus-minus shows us what a mistake the Grizzlies made at the beginning of the season with the backcourt, and how rectifying it has helped improve the team.
The Grizzlies started the year with a pairing of Jason Williams and Bonzi Wells . However, a quick check of net plus-minus shows that both have been significantly outplayed by their backups this season. At point guard, Earl Watson 's net plus-minus of +5.4 crushes Williams's -1.4. And at off guard, Mike Miller 's +4.8 is a huge improvement on Wells' -5.8.
Since opening day, of course, Miller has overtaken Wells at shooting guard while Watson continues to cut into Williams's minutes at the point. That's helped the Grizzlies turn things around after a slow start, although they still could take things a step further. There's no excuse for Watson to continue to sit while Williams starts, especially considering the Grizz are 11-13 when Williams plays at least 30 minutes. Perhaps Sunday's home loss to New Orleans, in which Williams played 31 minutes, will help Mike Fratello connect the dots.
Overall, plus-minus can be an extremely helpful tool for both fans and coaches. But like almost any other stat, it can be misused and abused by those who misinterpret the information. The key is recognizing that three factors exert powerful influences on a player's net plus-minus: a player's team, the players with whom he shares the court, and the man who plays when he sits.
However, once we acknowledge its limitations, a great deal can be gained from studying plus-minus. Using this data, a team like the Knicks can immediately see the overwhelmingly negative impact their lack of a backup point guard has had on their season, while the Grizzlies can identify a position where the backup is badly outplaying the starter. Nuggets of knowledge like that are available to us thanks to the plus-minus tool, giving us one more reason to be excited about the new frontiers that basketball analysts are discovering.
By John Hollinger
John Hollinger, author of "Pro Basketball Forecast 2004-05," is a regular contributor to ESPN Insider.