clock menu more-arrow no yes mobile

Filed under:

Frequently Asked Questions #13: The False Dichotomy between "Stats" and "Watching the Game"


The visceral reaction that many people have to the notion of statistics in sports has always surprised me.  Since the day my dad taught me to read baseball boxscores, I've been obsessed with numbers in sports.  Some of the greatest sporting achievements are burned on our brains as numbers: Hank Aaron's 755 career home runs; Gordie Howe's 801 regular-season NHL goals; Wilt Chamberlain's 100-point game; Joe Montana throwing for five touchdowns in Super Bowl XXIV, followed by Steve Young throwing for six touchdowns in Super Bowl XXIX - and on and on and on.


So it's a rare sports fan who doesn't value sports statistics (my ex-girlfriend who got into baseball because she thought J.T. Snow was cute notwithstanding.)  And yet...Nothing seems to make people's blood boil more than seeing *different* statistics than the ones they're used to.  How angry did on-base-percentage and wins-above-replacement once make most baseball fans?  Now we see them all the time on ESPN broadcasts and it would be shocking to find out that a front office didn't use them.  At the same time, sports fans have no problem quoting stats they know are utterly worthless.  How often do we hear about quarterback passer rating?  Or that a guy's an 'RBI man'?

To complicate the situation even further, the focus of this site isn't really on statistical analysis of hockey.  If it was, it would be probably be a good idea for me to have a background in statistics.  Hell, it would be nice if I had actually taken a statistics class instead of a half-dozen classes on probability (which is not statistics, in case you were wondering.)  Or maybe we could link to writers with formal backgrounds in statistics...If they existed.

What we do here and advocate for is actually more like event-counting, be it approximate pass charting, detailed event charting, or scoring chance counting.  All of these things are in their infancy, so for the most part, we make use of the NHL's most basic event-counting - shots, goals and individual ice time - and the most complex statistical operation we use is adding all these numbers together over the course of 82 games.

But we still get comments like this - and I've selected this because I think it's the most eloquent and non-combative version of the "stats r dum" we've ever had here:

"My way of studying a game on TV is to turn the sound down, watch everything except the puck and to take notes (words more than numbers). I make no claim that this is superior or even equal to statistical analysis – in fact, I’m sure that I’m missing out on some truths by ignoring one and all of these new tools. It just makes me happier...whereas statistical analysis dessicates the whole thing for me."

To put it another way: taking notes for one game is "watching the game" and enhances your enjoyment of the game, but taking those same notes for all 82 (or 1230) games would be "statistics" and boring.  The irony is that what we've always argued for at this site is for people to take a systematic and rigorous approach to watching the game.  It doesn't matter what you track - as long as you track something - you'll ultimately end up with a useful result. 

So I can't see this as anything other than a false dichotomy.  Unfortunately, two things will need to subside for people to understand that: the native hostility many people have toward numbers and toward new ideas.