3

Daryl Morey and Quantifying Decision Making

Apr 30, 2014; Houston, TX, USA; Houston Rockets guard Jeremy Lin (7) celebrates after making a three-pointer during the fourth quarter against the Portland Trail Blazers in game five of the first round of the 2014 NBA Playoffs at Toyota Center. Mandatory Credit: Andrew Richardson-USA TODAY Sports

Daryl Morey, GM of the Houston Rockets, has been getting a lot of vitriol lately about how he bungled this offseason. On the face of it, too, he did butcher his chance to improve on an already-fringe-contender. He lost both of his vital bench pieces in Jeremy Lin and Omer Asik, plus his 3rd best scorer in Chandler Parsons, only to get Trevor Ariza in return.

That’s not a great series of transactions, to be sure, and the argument that he treats his players too much like financial assets too publicly is also fair, and probably worth taking very seriously as time progresses and we start to think differently about players and how they should be treated.

That all taken into consideration and put aside for now, there’s some really interesting questions to be asked about Morey’s decision making, and — independent of results — whether he made the right decisions when he made them this offseason.

The first decision that Morey made was to pursue a superstar. That’s fine. Morey and his acolyte Sam Hinkie are famously aware that you need superstars to win, period, end of story, and that’s been Morey’s endgame all along, it seems. He could see the finish line, and tried desperately to get there.

To pursue a superstar, was probably always the right decision, because a superstar is, eventually, the key to winning. Given the way that Morey’s built the team, that would eventually have to be his final move anyway.

The alternative — biding his time to gather assets and trade for a superstar, even if the two ended up being related — might have been better, given the Houston front office’s ability to squeeze excellent value out of late draft picks, but it wouldn’t have been objectively, unquestionably better, at any rate.

The next decision didn’t occur until Chris Bosh expressed sufficient interest in Houston as a destination. At that point, Morey faced a bit of a prisoner’s dilemma.

The prisoner’s dilemma, for the unacquainted, is a thought experiment in which two prisoners can either testify against each other or stay silent. If one testifies and the other stays silent, the testifier is set free, and the other is given 3 year in jail. If neither testify, both are given one year in prison. If both testify, they’re both given 2 years.

The prisoner’s dilemma is designed to illustrate that an outcome that’s ostensibly the best for both parties isn’t actually the best decision for both parties. In this case, both staying silent is the best outcome, but if either individual stays silent, they run the risk of getting 3 years if the other testifies.

The point is that the best outcome is actually the outcome in which neither party would change their mind after making a decision, independent of what the other does.

So, testifying is the best decision for both parties.

In the case of Morey and Bosh, Morey had already traded Asik — who he’d been shopping for a year, so that can be held independent of the decision to chase a superstar — for cap space, but to really fit Bosh with a max (or near Max) he would also have to trade Jeremy Lin at the expense of the team’s depth.

On the other hand, Morey had the option to keep Lin, and trade smaller pieces (Terrence Jones, Donatas Montejunas, Robert Covington and Isaiah Canaan all come to mind) to open up about $14.5 (without trading Jones) to $16 million (with Jones) in cap room for Bosh, hoping that Bosh would take a major discount to play on an immediate title contender and possible finals lock1.

Presumably — given that Morey could never offer Bosh more money than Miami and for a stretch Houston was reportedly confident that Bosh was a “done deal” — Morey was banking on the hope that winning a championship would be Bosh’s main priority in choosing a suitor, not money, even if money was a major factor.

On the other side of the fence, Bosh had the option to sign with either Houston or Miami, depending on whether or not Houston dealt away their players.

So, from Morey’s point of view, this prisoner’s dilemma can be diagrammed like this:

Morey Game Theory Tree

Where the first number is the impact of the scenario for Morey, and the second is the impact for Bosh, and where -2 and -1 are really bad and bad situations, respectively, 0 is the same situation as the prior season, 1 is a slight improvement, 2 is a big improvement, and 3 is a massive improvement.

In Morey’s estimation, if Bosh values winning more than money, then Miami is always a “1,” because he’s getting more money, whereas Houston is also always at least a “1” because he’d do more winning in Houston than in Miami. If Houston were to trade Lin, though, he’d get a “2,” because he’d got a lot of money and more winning.

This is the remarkable thing: Morey misunderstood Bosh’s values — Bosh did indeed value money and a good role on a team more than Morey expected — but even if Morey had Bosh’s expectations correct, Morey still made the wrong decision.

Remember, in these kinds of decisions, the “right decision,” the one where both parties are best off, is the one where both parties can make their decision and not ever want to switch their thinking depending on what the other party does. So where the “Trade Lin, get Bosh,” scenario may look like the best for both parties, consider that if Morey were to trade Lin, and then Bosh re-signed in Miami (what actually happened) Morey would want to change his strategy, meaning that it isn’t the best scenario.

In actuality, keeping Lin was always a better decision — the scenario of “keep Lin, get Bosh” had a better return for Houston than Bosh without Lin, and the scenario of keeping Lin and losing Bosh was also better than losing both. Similarly, Houston, in this scenario, was always the better decision for Bosh. As a result, “keep Lin, Bosh to Houston,” is the dominant strategy equilibrium, in game theory terms.

As it turns out, the misestimation of Bosh’s values made this chart inaccurate, but even if that was the working assumption, it was always a better idea for Morey to keep Lin. In that sense, trading Lin was the first — and by far the biggest — mistake of the offseason.

Losing Lin and Bosh meant that Houston didn’t take any steps forward and lost a draft pick and most of their real depth. It was the worst case scenario bred from a misinterpretation of the original situation.

Morey did his best to compensate by signing Trevor Ariza to an excellent deal (about $8.6 million next season and decreasing over time, where most people assumed he would get at least $10 million per year), which would have been a good idea regardless of whether or not Houston lost Lin or retained Chandler Parsons or anything else. The Rockets badly needed his wing defense in whatever form they could get it.

The best thing Morey did, in the end, was sign Ariza and let Chandler Parsons go, a decision that possibly made up for the mistake in trading Lin, long term.

If Morey had signed Ariza and re-signed Parsons, the Rockets would be capped out for the next two seasons on a core of Beverly, Harden, Parsons, Jones, Howard, and Ariza with no notable backup for Beverly, Howard, or Harden (at least, beyond having Ariza to back up him and Parsons).

Houston fundamentally needed to upgrade on defense this offseason, and as painful as it was to let Parsons go, I’m sure, Ariza is a much better get for them on that end, and so is probably a better fit long term:

Parsons V Ariza

Essentially, Morey had the decision to keep Parsons and run with this gutted supporting cast for the next two seasons, or to start gearing up to fill in the players around Howard and Harden. He decided both that Ariza was better for this team long term, and that it would be better to start filling out the supplementary players as soon as possible, and, to that end, he made the right decisions once he made his initial mistake.

Ariza will be a much better fit to this squad than Parsons was, ultimately, and Houston will probably both be better than people expect next season and a massive threat next offseason, when they’ll have a bit more cap space to fill in the supporting cast around the new starting lineup.

Daryl Morey may have made a mistake that messed up the Rockets’ immediate future, but it’s worth keeping in mind that he recovered from that by setting up the Rockets to be even better than they have been in only one or two more seasons.


* Points Allowed Per Possession and Points Allowed Per Possession on Wing Plays were per Synergy Sports. Wing Plays are defined as plays that a wing player would be defending when “in position,” or: Pick and Roll Ball Handler, Isolations, Spot Ups, Pick and Roll Screen Setter (but weighted to have less impact), Off-Screen Plays, and Hand-Off Plays.

** The total DRPM was calculated by adding the total DRPM’s of each player in Houston’s projected starting lineup, with the caveat that Harden, Howard, Beverly, and Jone’s values were weighted by a regression analysis calculating the relative impact on a player of playing in a better overall defense. Because Ariza is a better defender, everyone else’s values were increased accordingly.

  1. By the way, to those who might argue that losing Jones, D-Mo, Covington, and Canaan opens a bigger hole than losing Lin, consider that Lin’s WAR (wins above replacement) was 3.19 last season, and the WAR of the other four combined was 2.18. Part of that is minutes based, but both Covington and Canaan were actually replacement level players (could be filled by anyone), and Montejunas and Jones were much closer to replacement level than Lin and could probably be replaced on the minimum without losing much, if any contribution.

Hal Brown

  • Philippe de Rouffignac

    Quick couple of comments,
    1. Ff you do a combined statistical approach that includes WP/48 you might come to a different conclusion as to whether losing Jones vs Lin is as good as you suggest. Jones = 0.203 WP/48, Lin = 0.081 WP/48
    2. You’re prisoners dilemma scenario is based on the assumption that Bosh would consider 16M/year in the first place. We don’t know what really happened, but another scenario that Morey may have been facing is who to get rid of (of the 4 you mentioned above) in addition to Lin in order to reach 18-21M/year that most pundits expected Bosh to sign for in Houston. So, you can’t get to 18-21 and keep Lin no matter what else you do, so for Bosh to be an option, Lin has to go no matter what. Thus no prisoners dilemma. Given what was reported, Bosh offered north of 80M over 4 years by Houston, the likelihood that Morey was faced with the dilemma you outlined is slim to none.

    • hagrin

      You nailed it on #2. That entire part of the prisoner’s dilemma is almost definitely fiction.

  • Kumar

    Where does the decision to renounce Chandlers option factor in?