It makes sense that the beginning of the NHL’s analytics era has coincided almost exactly with the arrival of a league salary cap. With every dollar now precious, and with big-budget teams just as cost-conscious as small-market clubs, any advantage is worth pursuing.
The extent of the advantage offered by having a department dedicated to the statistical study of the game is still arguable.
At the very least, a long look at the numbers can prevent teams from making costly mistakes. That becomes undeniable when we look at some of the worst contracts around the league—long-term, big-money pacts which have become albatrosses in their early years.
Consider three of the most prominent examples and the warnings written in advance based on statistical trends. Each of these players is around 30 years old, past his expected statistical peak but still at a place where we would not generally see a steep decline in performance.
Andrew MacDonald’s contract is widely considered one of the worst in the NHL. The 29-year-old defenceman is in the second year of a six-year, $30 million pact which he signed with Philadelphia after the Flyers parted with multiple draft picks to acquire him from the Islanders.
He’s spent the majority of this season in the AHL, dressing for just a single major league game with Philly.
Even before MacDonald signed the contract, the warning signs were there.
Charlie O’Connor of the statistically inclined Flyers blog Broad Street Hockey wrote this shortly after Philadelphia acquired the player:
It's undeniable that Andrew MacDonald's possession statistics this season have been terrible. His Corsi-for percentage (which tracks how much a player controls the puck) ranks 110th out of 115 NHL defenseman that have received at least 750 5v5 minutes this season. His statistics relative to his Islanders teammates are pretty bad as well, which tells us that it's not simply a case of a "he's on a bad team" syndrome.
This wasn’t a player who went into a steep decline the moment he signed in Philadelphia. By the numbers, he was already a low-end NHL defenceman. It just took a change in organizations before conventional wisdom synced with that viewpoint.
Dave Bolland is a player whose NHL career hasn’t collapsed quite so rapidly, but his contract will still be widely recognized as a brutal one.
Two seasons into his five-year, $27.5 million deal, Bolland has a single goal and five points in 25 games for the Florida Panthers and has been healthy-scratched by his coach on several occasions.
Bolland, who was part of two Stanley Cup wins with the Chicago Blackhawks, played for the Toronto Maple Leafs before landing in Miami. As his contract neared its end, there were heated discussions about his value to Toronto and the desirability of re-signing him.
The Toronto Sun’s Steve Simmons wrote a piece with the headline, “Re-Signing Bolland Should Be Leafs' Top Offseason Priority,” and after making various arguments, he concluded that “His value cannot be statistically quantified. But his value should be unquestioned."
Jeff Veillette of TheLeafsNation.com felt otherwise, and in his response to Simmons’ piece, he hit on a number of statistical points, one of which I’ll quote here:
UFA centres with a higher pts/60 than Bolland in the past three years include Paul Stastny, David Legwand, Olli Jokinen, Derek Roy, Mikhail Grabovski, Saku Koivu and Marcel Goc.
Players who produce slightly less but would come a lot cheaper include Vernon Fiddler, Dominic Moore, Brian Boyle and Michal Handzus (who made Bolland expendable in Chicago).
Toronto managed to avoid making the mistake of signing Bolland long-term, but it was less fortunate when it came to winger David Clarkson. The Leafs inked the forward to a seven-year, $36.75 million contract.
As if the term and money weren’t restrictive enough, James Mirtle of the Globe and Mail called the deal “indestructible” because more than three-quarters of the salary in the deal came in the form of signing bonuses, which could not be bought out.
The Leafs snatched Clarkson away from the Edmonton Oilers, much to the disappointment of then-general manager Craig MacTavish.
Writing for the Edmonton Journal in the immediate aftermath of that deal, I argued that it might eventually be looked back at as a blessing:
A 45-point forward who doesn’t drive the play isn’t worth the money and still carries that risk; unless this season represented a massive step forward, that’s likely what he is.
Time will tell, but if [that] scenario plays out, then Craig MacTavish and Oilers fans will ultimately be glad that Clarkson preferred playing in Toronto to whatever contract Edmonton offered. If his body starts wearing down somewhere along the way, they might be very glad.
Clarkson's deal was so bad that in just his second season in Toronto, the Maple Leafs traded him away for Nathan Horton—a player never expected to play another NHL game because of a back injury.
That’s a lot of wasted money that would have been easily avoided by teams with even a cursory interest in analytics.
There is a great deal of room for argument about the nuances of statistical evaluation, but these weren’t players hanging out in nuanced territory. In each case, there were flashing red warning signs that should have been heeded by the clubs in question.
Some NHL teams may still be hesitant about embracing a statistical approach, and not without reason. But to ignore such methods entirely is to welcome disaster.
Jonathan Willis covers the NHL for Bleacher Report. Follow him on Twitter for more of his work.