Surprisingly, the NBA's No. 1-ranked first-half offense shares something in common with its No. 30 second-half offense: They're both the same team -- the confounding Minnesota Timberwolves.
What gives? How can Minnesota be so good or so bad offensively depending on the half? And what does it tell us about how the Timberwolves might play going forward?
Tale of two halves
Amazingly, before halftime this season the Timberwolves are outscoring a Golden State Warriors offense that might well be the best in NBA history, averaging 115.1 points per 100 possessions to 113.9 for the third-place Warriors, according to NBA.com/Stats.
Yet while Golden State gets better in the second half of games (improving its offensive rating to 115.3), Minnesota's carriage turns back into a pumpkin. After halftime, the Timberwolves' 95.2 offensive rating is the worst in the league.
Comparing Minnesota's offensive factors in the two halves reveals a completely different team. The Timberwolves are top 10 in three of the four key components of offense before halftime, but don't rank that high in any of them after the break.
Note that while Minnesota's free throw rate goes up in the second half, that's true for teams generally due to intentional fouls. The Timberwolves' ranking is a better indication of their relative foul-drawing.
Though the decline takes place across the board, far and away the biggest difference is Minnesota's shooting. Not only is effective field goal percentage (eFG) the most important one of the four factors -- accounting for almost half of the variation in a team's offensive rating from game to game last season -- it's also the area in which the Timberwolves have cratered the most. So let's take a closer look at Minnesota's shooting.
Similar shots, different results
Breaking down the Timberwolves' shot attempts into the NBA.com zones by half shows something surprising: Their second-half shot chart actually looks a lot better.
Yes, Minnesota has gotten to the basket a bit more frequently before halftime. However, the team is taking a quarter of all shots between the paint and the 3-point line. After halftime, that drops to 21.5 percent, with those extra shots all funneling beyond the arc.
Given 3s are much more efficient attempts than 2-pointers outside the paint, we'd actually expect the Timberwolves to have a better eFG in the second half if they shot with the same accuracy from each zone. Clearly, that hasn't been the case, as a breakdown of shooting percentages by zone and half reveals.
Minnesota has been less accurate from every single zone, often shockingly so. Take the difference in shooting on non-corner 3-pointers. In the first half, the Timberwolves are making 44.4 percent of these attempts, second in the league and a hair behind the San Antonio Spurs. After halftime? Well, stop me if you've heard this before, but Minnesota's 27.2 percent accuracy is worst in the NBA.
Maybe it's who's taking the shots? After all, Karl-Anthony Towns has attempted 9.7 shots per first half, which dips to 7.5 per game after halftime. Nonetheless, when you apply players' game-long eFG to their shots attempted by half, the difference is marginal: .500 in the first half, .499 in the second.
Now, none of the statistics that offer half-by-half splits on NBA.com/Stats account for the proximity of nearby defenders or whether the shot was taken off the dribble or in a catch-and-shoot situation. There is some evidence for the theory that Minnesota is settling for more shots off isolations after halftime, as the team's assist rate drops from 60.1 percent of made field goals (11th best in the NBA) to 55.8 percent (18th).
Given that good offense tends to create more opportunities for offensive rebounds, the decline on the glass in the second half also suggests increased stagnation. Still, the underlying process doesn't seem to be nearly as different as the results, which point toward another, more likely culprit: randomness.
Timberwolves better than second halves, but not as good as first halves
Whenever a team like Minnesota plays better in one half than the other, there's a tendency to assume that the better performance reflects the team's true ability while the weaker performance is fluky. Zach LaVine's comments to Tim Bontemps of the Washington Post about why the Timberwolves have struggled in second halves reflect this line of thinking.
"I think it's just our focus," LaVine said. "We don't come out as strong as we should. We give up points and we're not scoring, the ball's sticking and we just have to play like we do in the first half."
Realistically, Minnesota can't play like it does in the first half over 48 minutes because the level of shooting is unsustainable. Even with Andrew Wiggins' development as a 3-point shooter, the Timberwolves aren't actually the league's second-best 3-point shooting team, as they've been in the first half.
Think of the Timberwolves' shooting as like a dice player rolling a pair of dice. In the first half, Minnesota is rolling a 12 and shooting as well as reasonably possible given the shots the team is getting. In the second half, the Timberwolves are rolling a two and shooting as poorly as realistically possible.
When you combine the two dice rolls, Minnesota is probably coming out about right -- slightly better than league average from 3-point range. That makes sense: The more dice you roll, the likelier you are to end up with a result around average.
At the same time, as sabermetrics pioneer Bill James noted in his essay "Underestimating the fog," when you compare two variable outcomes (like dice rolls) instead of adding them, you increase the variability instead of reducing it. That helps explain why the Timberwolves can look so much different from one half to the next.
Over time, however, Minnesota will keep rolling the dice more times, making it less likely the outcomes in the first and second half continue to be so dramatically different. The odds are the Timberwolves' first and second halves will look a lot more similar in the future, but not perhaps in quite the way Minnesota fans are hoping.