[For those of you coming here from
Connor Conor McKenna's link - I'm not sure why he says that I'm an anonymous blogger. My bio is, and always has been, on the left-hand side of this page.]
Roy MacGregor's "piece" in the Globe and Mail this weekend isn't just an example of what happens when a lazy journalist is pressed for ideas at deadline time. It's what happens when a journalist is too lazy to even keep up with the state of the art in the sport he covers. Let's take a look:
"the incessant baseball-ization of the game...has become even more absurd than fans dressed as hockey pucks bouncing off each other in a humiliating race around the rink. This movement began in the 1990s, when the NHL determined it might be able to move up in American sports coverage by providing more fodder for analysis. The league developed a game sheet that has, by last count, 22 separate mathematical measures for each player on the ice..."
"Mathematical measures", huh? Nothing but a bunch of eggheads out there, using math to count up the number of goals and assists players get and their plus-minus and the number of penalties that they take. Oh, wait? That's not math? What's mathematical, then? Roy takes exception to the mere notion of winning and losing faceoffs:
"Not all faceoffs are won or lost. Anyone who has played centre knows that once in a while, you will tip off your linemates that you intend to go forward rather than backward, a play that will then count against you on today's scoresheet. Perhaps half of faceoffs are neither won nor lost, though no calculation factors this in."
Roy has clearly never bothered to find out the definition of a faceoff "win". A faceoff is won by the team that gets possession of the puck following the draw. It is not simply the center who gets his stick on the puck first. So if you win the faceoff forward - which centers almost never do after age 9 or so - and you control the puck, you get a faceoff win. Straw Man 1, MacGregor 0.
Are you ready for round 2?
"Perhaps the least helpful statistic...is shots, as they can vary as much as those coming from an AK-47 to a BB gun. Coaches tend to prefer counting scoring chances, but even this is a dubious measure as the NHL is loaded with losing coaches who will claim they "outchanced the opposition." As with virtually all hockey stats, the scoring chance defies pure definition."
Point taken: it's tough to figure out what an individual shot is worth. But isn't the same true of individual goals? Do teams never score lucky goals? Fortunately, we have a little thing called "the law of large numbers" - essentially, if you count up enough shots (or goals), the type of gun they came from tends to not matter.
As for scoring chances, I've compared the totals collected by NHL teams to those collected by our gang of volunteer scorers at SBN and compared the two in turn to Corsi totals (Fenwick, actually) and guess what? The ratios are almost always the same at the team-level, regardless of how exactly each statistic is defined - "purity" is not necessary. There is massive utility in Corsi numbers and scoring chances - dismiss them at your peril.
Roy misses another key point here: the losing team, because they spend more of the game trailing, will almost always outchance the winning team. It's just the nature of the game.
Straw Man 2, Roy 0.
Ok, now the next round is a bit tough to parse:
"Scoring chances, however, are the prime evidence backing a story that flew around the Internet this past week: "Early stats show Kovalchuk not worth it."...surely it is too early to conclude the $100-million (U.S.) was a waste. Using a measure called CORSI...the argument went that, a mere handful of games into the season, Kovalchuk was a disaster while two other Devils players – Travis Zajac and Dainius Zubrus (the pure definition of "journeyman") – were far more valuable to the team."
Now, note MacGregor's argument - it doesn't matter that the Devils got outshot when Kovalchuk was on the ice but not when Zajac was on, nor does it matter that he's a -4 while Zajac is a +2. What matters is that MacGregor believes Travis Zajac, a former 1st round pick who routinely matches up against opponents' top lines and was New Jersey's second-leading scorer last season, is a "journeyman." He just doesn't feel like Zajac is any good, and that - not Zajac's performance - is what determines whether Zajac could put up good numbers.
Now the point about it being "a mere handful of games into the season" is a valid one, but let's look at the original article written by ESPN writer Mike Hume, who's careful to remind us that he's an English Major and not a statistician:
"We’re still only a 10th of the way through the season, so it’s not time to panic yet..."
Mike was looking for an early season story and by using this year's stats - as opposed to Kovalchuk's poor career stats - to make his point, he stretched his argument a little too thin. Kovalchuk's numbers will almost certainly be better over the rest of the season, yet MacGregor attacks Hume even though he agrees with him.
Maybe MacGregor's making sense...Let's see where he goes from here:
"This, we humbly suggest, makes the coach's "we outchanced them" seem virtually contempt-of-court by comparison. Kovalchuk had the third worst CORSI rating on the team, just ahead of Jason Arnott and Jamie Langenbrunner, two of the team's better players. Another way of looking at this, of course, might be that the Devils were off to a horrendous start because its best players were stumbling out of the blocks, whereas journeymen were largely performing as journeymen are expected."
I don't understand that first sentence (and see no evidence of humility), and the second and third ones seem to reiterate MacGregor's belief that his perception of a player's ability is all that matters. He's heard of Jamie Langenbrunner and Jason Arnott, those crafty veterans with the gravitas necessary to lead a team. No statistic should ever make them look bad!
Straw Man 3, Roy 0 (and it's turning into quite a pummeling)
If you're like me, you don't usually walk away from a train wreck of this magnitude - you want to see what ultimately happens to that drunk in the bar who just told off the bouncer.
"It is, however, likely only a matter of time before the league adds this moot measure to its nightly game sheets, one more "baseball" stat to appease those who fail to comprehend that the best explanation for the game of hockey lies in a slight variation of the phrase "scat happens.""
This makes about as much sense as mouthing off to the angry 250-lb bouncer. What's next?
"This is not to say that the game does not break down into convenient numbers. But they are not found in a computer program or determined by slide rule. They're right up there for all to see. On the scoreboard."
Ah, the mythical "slide rule." Back when Roy was a boy, those math nerds used them to compute logarithms lickety-split. And today they use their new-fangled computers to pull the wool over our eyes. The only problem? A real live person sits in an NHL arena and counts up the number of shots when a player's on the ice. No computer is ever involved in coming up with a player's Corsi number.
I apologize for burying the lede, but here's where Roy's laziness comes in. I have spoken to numerous NHL executives about advanced statistics and one of their biggest complaints is that the scoreboard is deceptive. When they win despite playing badly - which happens all the time - they have limited means to convince their players of their poor play. The people around the game whose view of it is more nuanced than the beer leaguer who tells you to "look at the scoreboard" don't agree with Roy for one second.
He even admits it - he pokes fun at coaches who claim they outchanced their opponents. He knows that NHL insiders - people who've been close to the game for decades; Stanley Cup winning coaches and executives; Hall of Famers - disagree with him and have a much more sophisticated understanding of the game. You need only look at the stats sheets from the 1972 Summit Series to see that four decades ago, the best coaches in the world invented advanced statistics to evaluate the best players in the world.
I'm not really sure why Roy MacGregor went on this rant. Every team counts scoring chances. Most have capologists. Many are trying to get an edge with advanced statistics. Lou Lamoriello signed Travis Zajac to an RFA contract with a cap hit of almost $4M a year, indicating that he sees his 25-year-old center as more than a journeyman. Hockey long ago passed Roy by, and it had nothing to do with computer programs or slide rules or anything even remotely resembling math.
My advice to him: use his access to the inner workings of the NHL to educate himself about how the people who actually put their careers on the line running teams do their jobs. It'll be more fun than getting beaten up by a straw man, and he just might learn a thing or two.