If there's one thing that the Wizards' loss to the Pistons last night reinforces it's that the decisions coaches make are often baffling.
Randy Wittman is a perfect example. For much of the season, he's espoused a non-sensical "hot hand" theory of lineup selection in which the coach tried to discern who was "hot" during the course of a game and then sought to have a group of "hot" players on the floor in the fourth quarter.
This "hot hand" theory of lineup selection is bad on several levels, the first of which is how difficult it is to tell whether someone is hot or not.
Second, even if a player is "hot" by whatever criteria the coach has chosen, there's nothing to suggest that player will remain hot. Hot hand research indicates that NBA players, on average, are actually a bit anti-streak -- that is, a player who makes a shot is slightly more likely to miss his next shot.
In effect, Wittman (and the PGs he encourages to feed the "hot hand") are falling victim to a gambler's fallacy. A player is no more likely to continue making shots because he's made X number in a row.
Rather than trying to determine who's "hot," Wittman (and other coaches who think like he does) would be better off determining who his best players are and then maximizing their opportunities. On the court, basketball is primarily a contest of skill. From a coaching perspective, it's a series of probabilities decisions -- similar in many ways to poker or other decision-making games. The way you win long-term in decision-making games based on probabilities is by repeatedly choosing the option that gives you the greatest opportunity to succeed.
In a game like Texas Hold 'Em, it means folding unsuited pocket 7-2 instead of betting on it because it's your "lucky hand." It's not, even if you won with it last week or earlier in the day.
In basketball terms, it means not giving extra minutes to Kevin Seraphin because he's made a couple shots in a row. It means getting the ball to your most skilled offensive players as frequently as you can because they'll maximize your team's chances of scoring.
Anyway, the preceding is kind of a long way of expressing further puzzlement in Wittman's lineup decisions in the fourth quarter against Detroit. Using a probability-based theory of lineup selection, any Wizards lineup in a crucial situation would include Emeka Okafor and Martell Webster. Okafor has been the team's most productive player all season; Webster is the team's best shooter and most important offensive player.
Using a hot hand theory of lineup selection, Okafor and Webster should have been on the floor in the fourth quarter -- Okafor had 20 points and 8 rebounds (on 10-15 shooting) through three quarters; Webster hadn't done much, but had at least hit his lone 3pt attempt. At minimum, if you're using the hot hand theory, Okafor should have been on the floor.
Yet, Okafor didn't play in the fourth quarter, and Webster didn't get on the court until there were just three minutes remaining and the game was already lost. And, it's not like these guys needed rest for a tough back-to-back or a four-in-five stretch. They're about to have five days off for the the All-Star break.
Unless one or both of these guys are injured, I can't make sense of the lineup decisions. I literally have no idea what Wittman was thinking. This does not inspire confidence.
At any rate, below is my weekly stat update for the Wizards using Player Production Average (PPA), an overall rating system I developed that credits players for things they do that contribute to winning and debits them for things that don't. PPA is pace adjusted, accounts for defense, and includes a "degree of difficulty" factor based on the level of competition a player faces during his minutes on the floor. In PPA, 100 = average, higher is better, and 45 = replacement level.
Big improvers this week were Wall and Nene. Webster maintained his solid play, as did Ariza. Singleton has finally edged his season production above replacement level.
On the "drop-off" side was Temple, who should be returning to the bench after the All-Star break.
The big question I'm looking at as I consider the team's turnaround the past month is whether it means anything long-term. I'll save that one for later.