Luck, Skill and Jim Harbaugh

As I have noted before, the outcomes in many activities of life involve elements of both skill and luck. Investing is one of these.  So is football.  Understanding the relative contributions of luck and skill can help us assess past results and, more importantly, anticipate future results in many such areas.  Doing so is crucial to the investment management process and to understanding football too.

Luck (randomness) is a huge factor in investment returns, irrespective of manager.  “Most of the annual variation in [one’s investment] performance is due to luck, not skill,” according to California Institute of Technology professor Bradford Cornell in a view shared by all experts (Nobel Prize winner Daniel Kahneman talks about it in this video, for example).  Even more troublesome is our perfectly human tendency to attribute poor results to bad luck and good results to skill.  In investing, we generally underestimate luck.

Luck is really important in football too.  There is no substantive difference from a performance standpoint when a throw is tipped by a receiver and lands harmlessly on the turf, when it hits a defender who drops it, when it hits a defender who intercepts it and when it hits another receiver for a touchdown.  In those instances, luck is very important to the outcome of the game.  Similarly, while fumbles relate to skill from a defensive standpoint (stripping the ball effectively, for example), recovering fumbles is largely a matter of luck (for example, where and how the ball bounces).

It is the element of luck that has had many football geeks forecasting difficulty this season for the San Francisco 49ers. 

When Jim Harbaugh was hired to coach the 49ers before last season, the team had just suffered through yet another dreadful (6-10) season.  Harbaugh’s new players were soon locked out and pre-season training camps were dramatically curtailed, thereby limiting the players’ ability to absorb his system and precluding the new coaching staff’s ability to get to know and evaluate the players.  Perhaps most problematic in a quarterback-dominated league, Harbaugh inherited a QB in Alex Smith who was widely regarded as insufficiently talented and a total bust.

Based upon those realities and upon the related fact that a lack of turnovers and good field position correlate strongly with winning, Harbaugh established an almost painfully conservative offense designed (in a phrase that coaches love) to limit mistakes.  He also developed an aggressive and opportunistic defense as well as outstanding special teams units. Harbaugh’s focus and approach resulted in the 49ers leading the league in turnover differential (a remarkable +28), defensive starting field position (76.1 – on average, opponents started at their 24-yard line) and offensive starting field position (66.6 – on average, the 49ers started at their 33-yard line).  Largely as a consequence, the 49ers improved from 6-10 to an astonishing 13-3 and advanced to the NFC Championship Game. Harbaugh was named 2011 NFL Coach of the Year.

Most experts think that Harbaugh is a fabulous motivator and tactician, but many (and especially stat geeks) have suggested that the 49ers will slip a lot this season due to regression to the mean.  Much of what the 49ers did well last year is thought not to be repeatable.  As my mother used to say, “these things average out.” For example, the 2012 Football Outsiders Almanac projects 7.2 Niner wins this season.  That’s obviously a major fall-off from last season’s 13. 

To begin with, teams that make big jumps to the upside generally fail to sustain it.  Perhaps worse, this season’s Niner schedule seems much tougher (although the schedule for the entire NFC West is tougher).  Moreover, the 21 other teams since 1978 with a turnover margin of plus-20 or better averaged a +3 the next season – a huge drop – and won 2.5 fewer games – another big drop; the luck factor (regression to the mean) at work.  Each of the QBs with a crazy-low interception rate in 2010 (like Brady, Roethlisberger and Cassell) suffered interceptions at a rate of at least twice that in 2011; experts expect Alex Smith similarly to regress toward the league average. And, with a full year in the system under and substantial upgrades at receiver, it’s hard to expect the 49ers to remain so ultra-conservative this year, which should also mean more turnovers (Hall of Famer Jerry Rice describes this as taking the diapers off).  An increase in turnovers would also hurt San Francisco’s outstanding field position averages last season, which were (as noted above) the best in the NFL both offensively and defensively.

One of the other reasons that the Niner defense was so effective: health. Last year, San Francisco’s 11 starters on defense combined to miss a total of just eight games due to injury. That level of health – best in the NFL – is very unlikely to repeat.

There is also something to the idea of a slump for coaches and teams after a really big jump forward (Football Outsiders calls this the “plexiglass principle“).  For example, Bill Belichick followed his 5-11 debut season with the Patriots by leading New England to a stunning 11-win season and a Super Bowl win. It’s mean regression at work.  Teams that get “the breaks” in one season are not likely to get them (in an overall sense) the next.

I don’t disagree with this mean regression analysis.  But I don’t expect Harbaugh and the 49ers to underperform by much (if at all) either.  Like advocates of the efficient markets hypothesis, the stat geeks (and I use the term affectionately) significantly undervalue what skill can bring to the table – largely because real skill (and especially coaching skill) can be so hard to find.

Analysts pay lip service to the idea that Harbaugh is a terrific coach (consider the jobs he did at the University of San Diego and Stanford, for Heaven’s sake), but he’s even better than they recognize.  For example, Harbaugh made the passing game easier for his quarterback last year, particularly when it came to beating the blitz. The typical NFL offense requires that receivers break-off their routes to counter a blitz and thus demands that receivers and the quarterback consistently are “on the same page” as to what the defense is bringing and what ought to be done about it.  These are the so-called “hot routes” based upon “sight adjustments.”  Instead, Harbaugh installed a scheme with built-in hot routes by design. Thus Alex Smith knows that if the blitz comes, he throws to a particular route (which is part of the play call even if there isn’t a blitz). 

In football lingo, if Smith reads blitz, he goes through his progressions shallow-to-deep. He can throw the three-step route to the area vacated by the blitz or to the designated hot route.  If his protection holds, he can move on with his progressions for a deeper throw. The key is that Smith and the receivers don’t have to read the same thing in the same way at the line of scrimmage.  This approach gives the QB the confidence of knowing where the ball should go without having to audible. It also means he knows what to do and doesn’t have to figure it out on-the-fly.  For a beleagured quarterback lacking confidence, it was a brilliant coaching move.

Together with his tactical and motivational gifts, Harbaugh brings a lot to the table.  With a full off-season and a full training camp, I expect Harbaugh’s coaching to be even more impactful this season.

That said, players make the biggest difference in any NFL game.  This season’s 49ers return 21 of 22 starters from last year and upgraded their weakest position along with adding quality depth elsewhere. I expect quarterback Alex Smith to improve too (despite many nay-sayers), particularly because he will be in the same system for two consecutive years for the first time as a pro.  This year’s Niner roster is significantly better than last season’s.  Thus I do not expect a significant performance drop-off.

My analysis is of course supported by what has already happened this season.  Despite many projections of a tough start, the 49ers are one of only three play-off teams from a year ago to have started this season 2-0. San Francisco handily defeated Green Bay – widely touted as the best team in the league – in Green Bay and handled Detroit – also a play-off team last year – in the season’s first two games.

Alex Smith has completed over 70 percent of his passes so far this season (40 of 57) for 437 yards and four touchdowns and a 115.9 passer rating. He’s also working on a franchise-record streak of 216 passes without an interception, which is pretty impressive given some of the QBs who have played in San Francisco.  At least to this point, regression, resmeshion.  Of course, as with investing, the tide could turn in a hurry.

In football as in investing, luck matters a lot, but skill remains undervalued and underappreciated. I wish Jim Harbaugh were back in San Diego, this time with the Chargers.  Skill matters…a lot.

__________

More on luck, skill and football is available here. If you liked this post, you might want to read an article I wrote relating trading to Moneyball (here).  You might also want to read a couple of pieces I wrote on college basketball, fans and behavioral economics:

• Carolina Loss Aversion
• Carolina Crazy