A recent post for Psychology Today relates how the human tendency to create narratives from unrelated events can result in irrational conclusions:

Humans constantly connect dots. We don’t see disconnected incidents. We see patterns that reveal what’s really going on. For the most part, this compulsion serves us well, which is why it’s part of our cognitive hardwiring. But it also causes us to see faces on burnt toast, cancer clusters, and countless other patterns that don’t exist. Then we make up stories to explain them. Which can get us in heaps of trouble.

Dan Gardner’s topic is the recent spate of random flocks of birds dropping dead from the sky, but you can see this tendency play out in sports analysis all the time. For example, in his recent 30 thoughts post, Elliote Friedman postulates that the San Jose Sharks’ woes may have something to do with the 2010 Olympics:

Interesting theory on San Jose’s struggles: That Joe Thornton, Patrick Marleau, Dany Heatley and Dan Boyle lack the motivation of last season, when they were competing for a spot on the Olympic Team. Don’t know if it’s entirely fair since they are a huge part of the NHL’s fifth-ranked power play, but when a team is struggling that badly at even strength, the stars take the heat.

Seems like a plausible hypothesis on it’s face I suppose: the San Jose Sharks were good last year, the Olympics were last year, Thornton et al were “auditioning” to make the Canadian Olympic team, the team stinks this year, ipso facto the players aren’t as motivated. However, it falls apart upon further inspection.

For now, let’s ignore the fact that the Thornton et al are established, excellent-to-elite professional hockey players whose motivation to win likely extends well beyond the incentive of making an Olympic squad. Let’s also ignore the fact that one would have to assume an “Olympic motivator” effect localized to the San Jose Sharks, as numerous NHLers who appeared for their various countries in competition last February seem to be having better seasons than they did last year (Sidney Crosby and Ryan Kesler come to mind).

To investigate this claim, we can simply compare the performance of the Sharks’ big guns from each season and see what’s changed. If their play has legitimately dropped across the board, then perhaps there is some kind of post-Olympic malaise gripping the San Jose dressing room.

There’s no question the trio of Dany Heatley, Patrick Marleau, and Joe Thornton were more potent at even strength in 2009-10. They scored 2.20, 2.61 and 2.76 points per 60 minutes of ice, as compared to 1.71, 1.39, and 1.58 so far in 2010-11. To give some context to those rates, your average top six forward in the NHL should usually be producing at a rate of 1.80 ESP/60, all things being equal. The Sharks’ three most expensive guys up front were batting well above that last year and are below this season. So far, so good.

Things begin to break down once we move past the counting stats, however. Here are the possession rates for Heatley, Thornton and Marleau for each season (stats with the score tied to factor out playing-to-score effects and corrected for zone start ratio):


Thornton: 1.92/60

Marleau : 7.01/60

Heatley: -0.81/60


Thornton: 1.78/60

Marleau: 2.74/60

Heatley: -2.57/60

Some modest steps back for Marleau and Heatley, while Thornton is right in line with his previous season. The ranking is the same as well, with Marleau leading the way and Heater bringing up the rear. The Sharks’ “big three” were right in the middle of the pack amongst regular sharks skaters by this metric last year, and that’s about where they currently sit. Some decline, but not too much has changed there.

What has changed though is the rate at which shots are going in when they’re on the ice this year. Thornton et al got the bounces last season, to the tune of above average PDO’s (on-ice SV% + SH%) ranging from 103.1 (Thornton) to 102.2 (Heatley). This season, in contrast, things have regressed back to the mean – Thornton and Marleau are hovering below the NHL average of 100 (97.2 and 96.4 respectively), while Heatley is right in line with it (100.9), but still several percentage points below his rate last year.

Ironically, as Friedman alludes to above, the $21 million trio is scoring at a greater cumulative rate on the power-play so far this season. Thornton (8.64 PPP/60) and Heatley (7.09 PPP/60) are two of the most efficient power-play producers in the league while Marleau is still above average at 5.04/60. Last year, no regular Sharks skater scored more than 6.50 PPP/60, with both Thornton and Heatley coming in just below that mark.

So, are the Sharks’ struggles connected to last year’s Olympics? Ehhh…probably not. The money men were mediocre at possession last season and are again this year. They are even more potent on the PP, however, and the only real deleterious influence relative to last year is the percentages at ES, which tend to vary around the mean without reference to things like international sporting events. To grant some legitimacy to the motivation theory, one would have to assume that the effect (again, apparently localized to Sharks players) is marginally suppressing their possession, causing them to score less frequently on their shots, causing their goalies to stop less shots while they’re on the ice (or, if you prefer, causing them to give up better quality shots), but disappearing once the Sharks have the man advantage.

Seems unlikely to me.