Annals of generic statements

« previous post | next post »

The Philadelphia Eagles have been winning recently, and this led [Inquirer columnist] John Gonzales to pose a generic question yesterday on the philly.com sports blog You Talkin' to Me?:

The Eagles are on a nice little run right now. Someone asked [Eagles head coach Andy] Reid about that the other day, about why his teams seem to come on strong late in the season. (Maybe that's something we just imagine; conventional wisdom and all that.) He just sort of brushed it off without giving an answer. Surprising, I know.

You guys are smart(ish). Why do Reid's teams appear to play better at the end of the year? And who wins today's game?

Another Inquirer sports columnist, Bob Ford, took the question semi-seriously:

The obvious answer – the one that will appeal to you, Gonzoremedial – is that the coaching staff tinkers along the way, adds a piece here, removes a piece there, crafts the schemes to the abilities of who it has, and, presto, the team is operating at peak efficiency at the end of the season.

There's probably some truth there, but 10 or 11 seasons is not a statistically significant sample. Let's say that because of the schedule and injuries, they start strong in five seasons and end strong in five seasons. That's what you would expect. Now if you give them a strong finish as opposed to a strong start in, say, just two of the swing seasons, that makes it look like an overwhelming trend. When, in fact, it's probably just happenstance and the vagaries of the schedule, etc.

Gonzales called a violation:

Lord. And I thought your columns were tiring. My eyes glazed over about halfway through.

Ashley asked Quintin Mikell a good question the other day about Reid's easing up on them in practice as the season goes along so they'll be somewhat fresher. Mikell couldn't pinpoint the exact time when Reid does that each year but conceded that it helps. I'll go with that.

Ashley Fox agreed:

Seriously, Bobby. That was brutal. What exactly were you trying to say? You Page One guys really are losing it.

Don't downplay the mental aspect of things. Many of the players have experienced success down the stretch, and that does matter. It matters that the Cowboys haven't experienced success. Yes, their schedule is brutal, and maybe they're just not good enough, but it's impossible to discount the effect of past December crashes.

Ford tried again:

If the Eagles win two of their three remaining games, just for argument's sake, they will be 6-2 in the second half of this season. In the last five years, that will make them 23-16-1 in the second half of seasons and 22-18 in the first half of those seasons.

Anyway, Eagles roll over the 49ers today. 38-13.

That pretty much ended the conversation, which had lasted long enough to fill the blog entry anyhow. Gonzales:

Way to look stuff up, Bobby. Birds 31, Niners 17.

Fox:

I agree. Eagles beat the fighting Bob Langes today, 35-24.

The actual score was 27-13.  Oh, and Dallas unexpectedly beat New Orleans.

These three sportswriters are half-seriously acting out a conversational pattern that takes place over and over again, one which none of us are very well equipped by nature to deal with.  In talking about sports (or life), people love to speculate about the causes and consequences of generally-accepted generic propositions.  Some of these propositions are true, while others are a compound of stereotype and confirmation bias that don't even rise to the level of sampling error.

I've been reading recently about some relevant intellectual history (Ian Hacking's The emergence of probability; Theodore Porter's The rise of statistical thinking), and these works underline how recently, painfully, and incompletely our species has recognized the reality of random variation and the difficulty of inferring its causes from observation.



10 Comments

  1. Richard Hershberger said,

    December 21, 2009 @ 10:42 am

    What I take away from this is that I should read Bob Ford, as he clearly is the only one of this group who isn't an idiot.

  2. slobone said,

    December 21, 2009 @ 11:47 am

    Hmm, well I still don't think there are very many members of our species who really recognize the reality of random variation. Otherwise Vegas would be out of business…

  3. Peter Taylor said,

    December 21, 2009 @ 12:07 pm

    Gonzales called a violation

    Did it pick up?

    More seriously, I presume that's a sporting metaphor, but it seems rather a mixed one. Has Gonzales switched from being a participant to being a referee? Or are my instincts wrong here, and "calling a violation" is something which participants do? (I would expect them to "claim" one, with the actual call being left to an arbiter. Mind you, I would also expect them to claim a "foul" rather than something which sounds like rape, so I'm not sure my instincts can be trusted here…)

    [(myl) "Call a violation" is a reasonably common sports expression, at least in the U.S. If you're playing a pick-up game, without a referee, then players do it, if anyone does. "Violation" is a broader category than "foul" — in basketball, for example, it includes things like walking with the ball, stepping out of bounds, hanging around too long under the basket, etc. And the expression is often used metaphorically, e.g. "I call a violation of Godwin's law".

    In the case under discussion, the violation seems to have involved introducing questions of sample size and statistical significance into a round of pundits' thumb-sucking.]

  4. fiddler said,

    December 21, 2009 @ 2:05 pm

    In a convoluted way, this reminds me of financial reporters trying to come up with reasons for the vagaries of the Dow.

  5. John Cowan said,

    December 21, 2009 @ 6:22 pm

    Nothing has ever happened in sports, ever, that wasn't equally well predicted by an appropriately biased coin. Except, of course, DiMaggio's streak.

  6. Peter Taylor said,

    December 22, 2009 @ 9:05 am

    "Call a violation" is a reasonably common sports expression, at least in the U.S. If you're playing a pick-up game, without a referee, then players do it, if anyone does. "Violation" is a broader category than "foul" — in basketball, for example, it includes things like walking with the ball, stepping out of bounds, hanging around too long under the basket, etc.

    Thanks. FWIW "foul" has that full range in my dialect (south-eastern BrE).

  7. Forrest said,

    December 22, 2009 @ 2:27 pm

    The writers might as well have answered "Numbers are boring and complicated" then went back to invent a narrative.

  8. Killer said,

    December 24, 2009 @ 5:09 pm

    In college I wanted to write a paper about the Oxford-Cambridge boat race (rowing). Looking over the history of the race, I found two stretches where one of the universities (I forget which) had won for nine consecutive years. So I researched why that happened. As I recall, the answer was, "Well, it just sorta happened."

  9. Lane said,

    December 27, 2009 @ 12:50 am

    Oxford (my master's alma-mater, if there is such a thing) has two 9-win streaks and one 10-win streak. Cambridge has none that long. I wonder if that third streak (1976-85) was after your college research, and whether it would change the findings… Obviously Oxford men are more courageous, virile and whatnot, right?

  10. Lane said,

    December 27, 2009 @ 12:51 am

    Ay, got that wrong – looking back, Cambridge has a 13-win streak from 1924 to 1936. Probably "just happened"…

RSS feed for comments on this post