Do Coaches Matter?

In a nutshell, he proposed that given the choice between trading Jordan, Bledsoe, and Butler for Kevin Garnett and Paul Pierce, and trading two picks (2015 first and second rounders) for Doc Rivers, did the Clippers make the right choice in choosing to go with Rivers? He concedes that if Chris Paul wanted Rivers, the choice was very easy, but that the Clippers (and Paul) may have regrets later on, overrating the Doc-Rivers-effect.

But while this writer seems to think the Clippers overvalue the effect of coaches, I think Haskins is undervluing them. He gives a number of examples, including Rivers, Vinny Del Negro, and Larry Brown, demonstrating that coaches with good teams have good records, and coaches with bad teams have bad records. That seems to make sense, but what I think he's completely ignoring are the coaches with good teams that have bad records, and vice versa.

What is a Good Coach?

To me, a good coach is one that takes talent and performs better than expected, and a bad coach is one that takes talent and underperforms. Using ESPN's stat, Expected Wins, basically relies on point differential and correlates well with actual standings, the Clippers under Vinny Del Negro should have had 61 wins but ended with 56. Considering the fact that they had one of the best clutch players in the league in Chris Paul, you'd really expect them to overperform (as point differential doesn't really account for the ability to win tight games). Underperforming by 5 wins seems like the mark of a bad coach, doesn't it? And it was actually the 4th worst underperformance in the league.

But before we get ahead of ourselves and start Vinny-bashing, this really isn't enough to answer the question, is it? This is just one season, and one example, and could be explained by any number of reasons. What we need are more examples and more seasons before we can determine if we can actually identify "good" coaches and whether Doc Rivers is one of them, thus making the trade worthwhile.

What are Expected Wins?

First, I'll explain what Expected Wins are. Expected Wins is a formula used by ESPN's former analyst John Hollinger, and it's based on the Pythagorean Expectation created by Bill James for baseball. Noted statisticians Daryl Morey (also Houston's GM) and Dean Oliver use similar methods. All of these methods revolve around points scored and points allowed, and generally correlate very well to actual wins.

The inherent flaw with using Expected Wins (or any point differential measure) is that blowouts tend to skew the extremes ends of the spectrum. That is, one 20-point victory can offset 20 one-point losses. This makes strong teams that can score a lot look like they are underperforming, and weak defensive teams look like they are overperforming.

So when we look at the 2012-13 season to see if there were any other interesting things to discover, as expected, we see some surprising teams at the top and the bottom of the pack. Three of the top offensive teams, the Thunder, Clippers, and Spurs are huge underperformers, and two of the bottom defensive teams, the Bobcats and Suns are huge overperformers. So maybe if we can pick out some of these abnormalities, say by adjusting for the number of blowouts, it will normalize the Expected Wins a bit.

There are better ways to do this, but I figured the easiest way would be to take the "bite" out of any blowouts. For the purposes of this exercise, any victory (or loss) greater than 15 points will be considered a blowout. And so in order to lessen the effect, I will decrease any blowout to a 15 point victory (or loss). No team can win by more than 15, or lose by more than 15. And I think this will provide a more level basis on which to evaluate the different coaches.

Getting rid of some of the blowout effect, it appears to have corrected much of the problem. Bad teams like Phoenix and Charlotte performed mostly as expected, and good teams like the Clippers and Thunder were exactly what their point differential indicated: good teams.

Who are the Good Coaches?

So now we can get back to our original issue: Who are the good coaches? Or more specifically, which coaches are getting better execution out of their players, winning tight games and properly utilizing the talent?

It's not just a matter of where the teams fall on this list, but rather where they fall on this list given the talent on the roster. The Clippers and Thunder have Chris Paul and Kevin Durant respectively, and yet they've failed to win enough close games to overachieve beyond their point differential. Both teams are sub-.400 in games decided by 3 points or less. That's absurd. Miami and Golden State, on the other hand, have great records in close games (both above .600), and they have successfully overachieved beyond their point differential. They, too, have stars capable of getting clutch buckets in LeBron James and Stephen Curry, and yet it appears that their coaches (or assistant coaches coughGoldenStatecough) have been able to get their teams to execute down the stretch to win these close games. And it's really not at all surprising to see well-known coaches such as Gregg Popovich and Tom Thibodeau overachieving.

Unfortunately, determining who the good coaches are cannot be done based on one year's evidence. We see some signs of evidence of teams overachieving/underachieving with respect to point differential, but it can't really be quantified how much of that is due to the coach. This is where historical data comes into play. For now, given that these take a lot of time to throw together, I'll just go back 2 more years, to 2010-11, the beginning of the Vinny Del Negro era.

Well, it looks like Vinny's at least been consistently solid. We all knew he wasn't a bad coach, but he doesn't seem to be doing what Popovich is doing, where he always has his team outperforming their point differential without relying on blowouts to do so. Thibodeau seems to always be in the top echelon as well. Of course, "top echelon" just means getting your team a couple extra wins, and a couple wins based on point differential is really the difference between a couple buzzer beaters falling or not. It's not really an exact science. That said, a couple wins are also the difference between going to the finals and getting bounced in the first round, so perhaps it's at least worth looking at.

Conclusion

So have we successfully addressed the issue of whether coaching matters? Can we even identify good coaches? There's really no perfect way to measure the asset that is a coach. It's almost intangible, and yet we know it exists. Well, we think we know it exists. We see terrible substitution patterns, and we see botched possessions out of time-outs. And the grass always looks greener on the other side, doesn't it? But I'm sure even the great Gregg Popovich frustrated fans when he insisted on playing an ice cold Manu Ginobili during the playoffs, or decided to completely freeze out the always-productive DeJuan Blair.

Can we ever truly know if Justin Haskins is right, and that the Clippers should have traded for Garnett/Pierce rather than for Rivers/Redick/Dudley? No. Not really, but they are very different options. The Rivers/Redick/Dudley move looks to the future much more than a Garnett/Pierce move, and with 5 years of Paul/Griffin, there will be more opportunities down the line, and Redick/Dudley will be needed.

Can we determine if the Clippers are overrating their coach? Again, not really. But my analysis at least shows that for the last 3 years, Rivers has always gotten his teams to at least play to their potential or exceed it, which is never a bad thing. We must also remember that the alternatives were Lionel Hollins, Brian Shaw, Byron Scott, and Alvin Gentry, the last of whom is now an assistant coach. Could the Clippers have gotten Gentry to be an assistant coach to a head coach other than Rivers? Maybe not, and that's an opportunity cost worth considering as well.

I honestly don't think the Clippers will ever regret their moves this summer. They made a bold choice, and other than a few media outlets here and there, a universally-praised one. Rivers had the reputation as being the best coach on the market, and they got him. One day, maybe we will come up with a better way to rate coaches, and perhaps then we will be able to determine if our coach is overrated. For now, all we can do is set high expectations, and hope that Rivers and the rest of the Clippers can meet them.