When delivering the usual upbeat lines about fitness, team cohesion, and improved health, the Wizards have been heaping praise on the “upgraded” bench. John Wall even called it “great” at Media Day. Training camp optimism aside, have the Wizards improved their bench? And, if it is better, how much of a difference will it make?
It makes sense that having a good bench would help a team win more often. Analysis shows that isn’t the case, however. In the NBA, starters are the key difference makers because they’re better players, they consume more minutes, and they include what separates great from good from average from bad: the superstars. In a league ruled by elite talent, a good bench is more luxury than necessity.
Last season, for example, starter production had a strong correlation with wins while bench production did not. (For those who nerd hard, the starter correlation was 0.9; for the bench it was 0.0.)
The league’s best bench last season belonged to the Mavericks, which managed just 24 wins. Some of the league’s best teams had top-ten benches (Raptors, Cavaliers, and Warriors), but so did middle-of-the-pack squads like the Heat, Nuggets and the Clippers, and bad teams like the Nets and Lakers.
The two Finals teams were both in the top ten in bench quality, but Golden State also had the second-best starting lineup (behind Houston), and Cleveland had the eighth-best starters, LeBron James, and played in the East.
The Rockets had the 20th-ranked bench, but won the most regular season games and were a pulled hamstring from the Finals. The 52-win Sixers ranked 19th. Portland, Indiana, San Antonio, New Orleans, Milwaukee, and Oklahoma City all had worse benches and better records than the Wizards.
A better predictor of team success than bench quality: the difference between starters and the bench. And, higher is better -- teams with a bigger drop-off from starters to bench tended to have better records. The top teams in starter-bench differential were the Rockets, Pelicans, Warriors, Spurs, and Thunder. The only non-playoffs team in the top ten was the Hornets, and they ranked 10th. The Wizards were 12th in starter-bench differential. In other words, having a good bench is fine, but what really matters is the starting group.
This is not to say the bench is worthless. Having competent players in reserve can be important when injuries strike. Tomas Satoransky blossoming into a solid pro was critical to keeping Washington’s season afloat when Wall was sidelined with knee trouble.
Last season, Washington had a decent starting lineup (14th in production using the Player Production Average (PPA) metric), and a slightly below average bench (21st in production, according to PPA). The Wizards’ 2017-18 bench mob of Kelly Oubre Jr., Satoransky, Mike Scott, Ian Mahinmi, Jodie Meeks, and Tim Frazier managed a collective PPA of 73. (In PPA, average is 100 and higher is better.)
Replacing Scott, Meeks, and Frazier with Austin Rivers and Jeff Green would have boosted the bench PPA to 80, and their rank from 21st to 14th. That likely would have added a win or two, which could have pushed the team to the sixth or seventh seed and a first-round matchup with Boston or Philadelphia instead of Toronto.
The real issue, of course, is whether the bench is better in the coming season. My homemade projection system says yes, and predicts the team’s bench could climb from below average to about average (a rank of 12-14). A simpler system that applies average aging effects is less optimistic and suggests the team’s bench is more likely to be about the same quality as last year’s.
My best guess is that the bench will be a little better this season, and that it won’t matter much. Like most teams, whatever success the Wizards earn will be driven by the performance of their top players.