Wolves better than you think
"You are what your record says you are," legendary former NFL coach Bill Parcells was fond of repeating, and for some fans that's an article of faith. For those fans, this year's Minnesota Timberwolves are one of the league's biggest disappointments. More than two months into a season that was supposed to snap a nine-year playoff drought, the Timberwolves stand .500, two games behind the Dallas Mavericks for the eighth and final spot in the Western Conference postseason.
Change the perspective, though, and Minnesota is suddenly a very different team. The Timberwolves have outscored opponents by 4.8 points per game, which puts them tied for fifth in the West and seventh in the NBA.
As a result, advanced team metrics built on point differential consider Minnesota a contender. That goes double for the Hollinger Power Rankings, which put more weight on recent performance. Since the Timberwolves have three wins by 20-plus points in their past 10 games, Minnesota has jumped to fifth in the Hollinger Power Rankings, ahead of the Miami Heat and Portland Trail Blazers.
So, are the Timberwolves what their record says they are? Or are they what their point differential says they are?
The meaning of point differential
First, let's consider the importance of point differential. Typically, it closely predicts a team's record. Here's how this season's NBA matches up:
Point differential predicts the record of 28 of 30 teams within two games. The outliers are the Philadelphia 76ers, who have won four more games than their differential would suggest, and the Timberwolves -- five full games behind their expected record. (No surprise that when the Sixers faced the Wolves in Monday's Point Differential Bowl, the result was a 31-point Minnesota win.)
Of course, we don't merely want to predict a team's record, because we already know that. What we want to predict is a team's record going forward, and that's where those outliers are crucial. According to NBA StatsCube, there's a stronger correlation between a team's differential before New Year's Day (.77) and its record the remainder of the season than its record (.75), demonstrating that point differential is the better indicator. That matches my past research showing point differential is the superior indicator for future seasons.
This is most important at the extremes. For most teams, using differential in place of record doesn't change the evaluation much. That's not true for teams like the Timberwolves, who had the second-largest gap between expected and actual record as of Jan. 1 since the 1996-97 season (not counting post-lockout seasons), trailing only the 2006-07 Memphis Grizzlies. Of the 10 teams with the largest difference, nine improved their winning percentage after New Year's Day, and their post-New Year's winning percentage slightly surpassed their expected winning percentage beforehand:
Why differential predicts better
There are a couple of reasons why differential predicts better. One is that blowout wins, which are easy to dismiss as a hot night or running up the score, do tend to indicate a team's ability. (The Timberwolves have won seven games by 20-plus points, most in the NBA.)
More importantly, differential isn't affected as much by record as the outcome of close games. And that's where Minnesota has struggled. The Timberwolves haven't won a single game by fewer than five points this season (0-9) and are 1-11 in games when the score is within three points in the final minute. Nobody else in the league has a record of worse than .250 in such games.
It's easy to craft a logical explanation for why Minnesota has struggled in the clutch. Ricky Rubio's unwillingness to shoot is problematic in half-court sets, especially when defenses are more focused on his weaknesses. Per StatsCube, Rubio is 2-of-16 from the field in the last five minutes of games with a five-point margin or less. And there is a track record here: The Timberwolves were 8-16 in games within three points in the final minute in 2011-12, and 9-17 last season.
At the same time, there's an enormous difference between winning close games a third of the time and winning one in 12 tries. Even three extra wins would be enough to put Minnesota six games above .500 and in the playoffs.
Beyond that, the evidence is overwhelming that the record in close games tends to regress to the mean. Using the same pool of teams dating to 1997, eight other teams have gone 1-8 or worse in games within three points in the final minute prior to New Year's. On average, those teams won 48.2 percent of their close games the remainder of the season despite being far worse than the Timberwolves overall, winning just 23.2 percent of their non-close games. (Better teams do tend to win close games more often than bad ones, though the effect is muted compared to more lopsided outcomes.)
Who the Timberwolves really are
Minnesota's woes in close games are reason for some skepticism that the Timberwolves are really one of the top teams in the Western Conference, and defenses might be able to take advantage of Rubio all 48 minutes in the playoffs when advanced scouting becomes paramount. But for the rest of the regular season, Minnesota is more likely to play like a team that outscores opponents by four points per game than a .500 one. (That's true before considering the boost the Timberwolves' bench could get from the return of Chase Budinger.)
Consider the examples of the teams most similar to Minnesota in the first two months. The 2007-08 Utah Jazz, 1996-97 Trail Blazers and 2001-02 Orlando Magic previously combined a record near .500 with a positive point differential. All three teams finished with at least 44 wins and made the playoffs. The Jazz improved enough in the second half that, combined with better luck, they won 54 games and beat the Houston Rockets in the opening round of the playoffs before losing to the L.A. Lakers.
So there's some precedent to playoff simulations suggesting that the Timberwolves still are heavy favorites to reach the postseason despite the strength of the West. The Hollinger Playoff Odds show Minnesota finishing in the top eight better than 75 percent of the time, and my "true talent" simulations are even more optimistic because of skepticism that the Phoenix Suns can maintain their level of play. So when the Timberwolves host the Suns on ESPN tonight, we're probably watching a playoff team in action. It just isn't the one the records would indicate.