The Houston Rockets and the Case of the Disappearing Wins

Newsflash: NBA coaches are hired to be fired. They get fired when their teams lose. They get fired when they win. They even get fired—as George Karl learned in 2013—after being named Coach of the Year. Only three coaches—Gregg Popovich, Erick Spoelstra, and Rick Carlisle—have held their current positions for more than five years.

So when the Houston Rockets started 4-7 this year, it was not entirely surprising that the franchise would tell Kevin McHale to get lost. Question is, will dumping McHale really make a difference?

The Rockets have finished with a winning record every season since 2006-07. No NBA team outside of the state of Texas has a longer streak of winning seasons. Inside Texas, though, it's a different story. The Dallas Mavericks are in the midst of a streak that is 15 seasons long, and the San Antonio Spurs have had a winning record in every season since 1997-98.

While both the Mavericks and the Spurs have won NBA titles since their respective streaks began, the Rockets have only made it out of the first round of the playoffs twice since 2006, and have yet to advance to the Finals. Considering the quality of Houston's teams, this shouldn't be at all surprising.

Team quality is best assessed by looking at efficiency differential, or the difference between the points a team scores and the points it surrenders per 100 possessions. In 2006-07, the Rockets posted a franchise-best differential of 5.2, which is consistent with a team that would win about 54 regular season games. Thirty-seven different NBA teams have bested that mark in the years since, four of them last season. In sum, Houston has been good for years—but not that good.

Last season was no exception to the trend. Yes, the Rockets won 56 games, but the team's efficiency differential of 3.4 is consistent with a team that wins 50 games. So again, good but not great. When McHale was fired, however, the Rockets' efficiency differential was -7.5. That is consistent with a team that would win about 22 games in a complete season. So what happened to all the wins we saw last year?

To figure that out, we can look at other stats to determine which players are primarily responsible for the difference between where the Rockets expected to be this year and where they were when McHale was fired.

When we move from efficiency differential to Wins Produced, we see that the Rockets' wins in 2014-15 were first and foremost about the play of James Harden, who finished the season with a MVP-caliber 19.6 wins. The remainder of the roster only produced 30.4 wins. And just as Harden was the key to what success Houston had last season, it turns out he's key to understanding Houston's current problems.

Last year, Harden produced 0.316 wins per 48 minutes played—more than three times better than an average player in 2014-15. (Since an average team produces 0.500 wins per 48 minutes played, an average player produces 0.100.) After 11 games this year, his Wins Produced per 48 minutes (WP48) was only 0.050. Not only was Harden no longer a star; he wasn't even an average NBA player. Harden's decline by itself reduced the Rockets' projected final season record by nearly 16 wins.

Harden is not the only player whose productivity has declined. Corey Brewer was not a big part of his team's success last season—he only posted a 0.052 WP48. After 11 games this year, though, he went from "not really helping" to "really hurting." His WP48 when McHale was fired was -0.163. Yes, that is well below zero. This cost the Rockets another eight wins in the team's projected final record.

Finally, there is Ty Lawson. Lawson spent his entire career with the Denver Nuggets prior to coming to Houston in 2015. In Denver, he was always above average, producing 48.1 wins over six seasons (with a 0.178 career WP48). With the Rockets, though, Lawson has been awful. After 11 games, his WP48 was -0.001, costing the Rockets about another twelve wins.

Had Harden, Brewer, and Lawson not declined, the Rockets' projected final record would be 47 wins, not a measly 22, and McHale would probably still be in Houston.

A major reason for the decline in their productivity is shooting efficiency. At the time of McHale's firing, the combined effective field goal percentage of this trio was 0.394. An average NBA player is almost at 50 percent in effective field goal percentage. Harden, Brewer, and Lawson literally shot McHale out of town.

Maybe McHale deserves some of the blame here; maybe another coach could have fixed this situation. Obviously teams fire coaches because they think coaches matter, but a study I published—with Michael Leeds, Eva Marikova Leeds, and Michael Mondello—suggests otherwise.

A few years ago, we looked at how player performance changed with the introduction of a new coach. In a study of 62 NBA coaches, only 14 were found to have a statistically significant impact on player performance. In other words, most coaches had no impact on player performance. That doesn't mean coaching doesn't matter; it means that most coaches are not really different from each other.

In Houston, Harden, Brewer, and Lawson started the year not hitting their shots. We can assume all three want their shots to fall, but when it doesn't happen—well, a coach isn't going to be able to do much to fix that.

So is a new coach going to help the Rockets? If shots start falling, it certainly will look that way, but unless the new coach literally has magic words to make that happen, any change in Houston's performance will be more a matter of chance than personnel change. In other words, there probably wasn't a good reason to tell McHale to get lost.