Thursday, December 30, 2010

Lies, Damn Lies, and ESPN's Statistic

I'm really not sure the last time I wrote about anything besides hockey on this blog. It just happened at some point. Originally I thought I needed to be broader, writing about the happenings in a wide variety of sports. Somehow I eventually settled into hockey. Which retrospectively makes a good deal of sense since it is both a niche sport and the sport I am most fond of. However, if my obsession with hockey is superseded by anything, it would be my sad devotion to numbers.


I quantify everything, from the number of steps up a stairwell to the exact amount of coffee grinds I use for a single cup. Numbers and I are BFFs. Which is precisely why I get so upset when they are woefully misused. Numbers are my little bear cubs, and you don't get between a bear and his cubs.

And this is why I feel the need to rail against, of all things, the statistic that ESPN is using to underscore the NFL's fining of "Gun Slinger", and general ass-hat, Brett Favre.

Since the $50,000 fine was announced earlier this week, I have repeatedly heard and read a startling statistic recited, one that goes something along the lines of "the fine is the equivalent of fining the average person $1.20".

It didn't seem right, so I took to tracking the original source of the statistic, only to discover it's appearance in the sidebar of an article on the fine at the Mothership, the Worldwide Leader, herself.

Now this is a great statistic for making a point because it shockingly underlines how much money a professional football player pockets, to the extent that a fine that would possibly bankrupt you or I amounts to nothing more than pocket change for the leagues career interceptions leader. And that's precisely why ESPN concocted it.

However, this is a horrible statistic because it is so blatantly and purposefully untruthful that it goes beyond the extent of bad math and settles clearly in the territory of bad journalism.

From the first time I heard it, it sounded off. Quickly pulling up the calc app, I calculated the fine to be 00.3125% of Favre's $16M salary this year. Estimating the average person's salary to be somewhere in the range of $45K, this equates to a .... $125 fine. Surely small enough to still make ESPN's original point, and yet drastically different from their numbers.

So how did they come to the $1.20 statistic? Just bad math? Maybe a misplaced decimal? Something surely we have all fallen to at one point or another. Sadly, no, they were decidedly dishonest because, well hey, it makes a more "shocking" statistic.

See, ESPN decided that instead of looking at it as a percentage of his income, clearly the easiest and most logical way to go about it, they would frame the statistic in the context of the amount of time that Brett Favre "worked". And how should they go about deciding time worked? Easy - minutes in a game. Which is something akin to deciding how long a writer has "worked" on an article by determining the amount of time it takes a person to read it.

Farve's $16M contract easily breaks down to $1M per game, with 60 minutes of play per game (even though - as even my mother, who hasn't watched a football game since the Cowboys defeated the Steelers in Super Bowl XXX, knows - a football game actually lasts somewhere around 180 minutes). Now dividing this $1M by the "60 minutes of game play", they reached a completely asinine and misleading statistic that Brett Favre makes $17,000 per minute of "work".

And thus, since it would only take Favre just under 3 minutes of work to make $50K, it is the equivalent of $1.20 for an Average Joe like us (another quick calculation determines that I am well below status of "Average Joe" when it comes to pay grades).

Not only does ESPN ignore the actual length of a game, they pretend somehow that Favre shows up at 1 PM on Sunday, clocks in, plays the game (in just 60 minutes), and clocks out. Even for Favre, who knows practically no limits to avoiding practice, this assumption that he only works 16 hours a year is beyond preposterous.

The most upsetting aspect of the situation is how their concocted statistic undermines an essentially valid point: Favre is being slapped on the wrist.

Whether you use their bogus of $1.20 or a much more accurate $125, the point remains the same, after his bigger-than-the-game shenanigans, refusing to cooperate with the NFL's investigation of his ... umm ... "dissemination of  unorthodox self-portraits", Favre got off easy. 

It's a shame that, not completely unlike their "Gun Slinger" in a big moment, EPSN improvised, went outside their journalism playbook, and made a idiotically pointless and self-inflicted error.

Monday, December 20, 2010

Shots Taken per Goal

In a battle along the half-boards a winger kicks the puck back to the blue-line. The defenseman throws it across the ice to his partner, who settles the puck down, takes a stride towards the net, raises his stick and sends a shot .... three feet wide and a foot too high, right off the windshield.

Missed shots tend to drive me crazy.  I've been known to fire off a quick frustrated tweet when someone (usually named Kris Letang) can't hit the net. And when I'm at work reminded myself how poor of a hockey player I am, there is nothing more likely to throw me into a Bruce Bodreau-esque tirade than not being able to put the rubber on net.

This isn't something the NHL takes into consideration with it's shooting percentage numbers. They simply divide a players Goals by Shots on Goal. Something Alexander Ovechkin, with a league leading 66 Missed Shots, might be alright with, but something surely Daniel Briere, who has a mere 27 Missed Shots, would feel is misleading.

And so here I've thrown together the On Goal % and Shots Taken per Goal statistics.

I have added Shots on Goal and Missed Shots to create Total Shots Taken. Based on that number I divided a players Shots on Goal by Total Shots Taken for the On Goal %. Then I divide Total Shots Taken by Goals to determine how many shots on average a player takes for a goal.

For example, Steven Stamkos scores a goal for every 6.6 shots taken. 74.4% of those shots taken are on goal.


My Thoughts:

- It comes as no surprise that Claude Giroux tops this list. As a Penguins fan I've seen more than enough of Giroux as he develops into an all-star caliber player

- Crosby and Stamkos. After cooling off for a bit, Stammer is back at it. Just maybe, with them nipping at each others heels, they can push each other further, like Magic and Bird. 50 in 50 looks out of reach, but someone, or both, might score 70 goals for the first time in 14 years.

- Even not accounting for the bevy of frozen rubber he's thrown wide of the net, Ovechkin's shooting percentage has taken a big hit so far this season.Taking all those missed shots into account, it isn't pretty for Ovechkin. You've gotta scroll for a while down that chart before you find his name.

- Same goes for Evgeni Malkin.

- As an Eastern Conference fan, I don't get to see nearly enough games from teams out west. This means I'm usually cautious of pontificating on guys that I haven't seen a ton of. That said, let's agree that Chris Stewart needs to take more shots. He is right behind Crosby and Stamkos at 6.9 Shots Taken per Goal, and he's put an insane 86.6% of his shots on net.

Complete stats available in a Google Spreadsheet