One of the things I am trying to work on this summer is to be more self-critical about the way I treat and act on information. Frederik Andersen’s trade from the Anaheim Ducks to the Toronto Maple Leafs, and his subsequent signing of a five-year, $25 million contract proved to be a good opportunity in that sense.
Hearing Frederik Andersen’s deal with the Leafs is $5.5-million a season. That’s a big commitment for a goalie with 125 games played.
— James Mirtle (@mirtle) June 20, 2016
Initially, I cringed a bit at the term and cap commitment Toronto made to Andersen. Five years is a long time and $5M per year is a big money for a guy who is not guaranteed to play all that well.
But I could be very wrong on that.
While statistically-oriented hockey analysts pride themselves on working hard to uncover objective truths about the game, I think it is still worthwhile to allow the possibility that we are just as vulnerable to biases and heuristics as any other human being. If anything, quantitative data may sometimes lock us into a mindset that lead us to the wrong conclusions.
Just because a goalie who is making close to league minimum salary just backstopped his team to the Stanley Cup, and just because many goalies being paid big money do not perform very well, does not mean that every goalie should be paid less than $2 million, or some other number that sits well with us mentally (this is a point that Carolyn Wilke made at VANHAC earlier this year).
1) We suffer from availability bias when we get fixated on the fantastic save percentage numbers of low-paid goalies (Murray, Hellebucyk, Mrazek, Reimer, Greiss, Neuvirth etc.).
2) We are affected by anchoring effects when we see late-drafted goalies with low salaries out-perform established starters.
3) We experience delight more easily when goalies come out of nowhere to win a starting job than when average starting goalies do enough to justify their minutes.
@MannyElk if you’re an NHL average goalie you’re one of the 20 or so best goalies in the world, or thereabouts. That’s, uh..good.
— Nick Mercadante (@NMercad) June 20, 2016
@NMercad @MannyElk interesting mental exercise to replace ‘avg starting G’ with ‘average firstline center’ when considering salary
— Jack Han (@ml_han) June 20, 2016
As outsiders, it is easier for independent researchers and “fanalysts” to adopt an “outside view” on any given deal. But it does not mean that our perception of the world is not swayed by emotions or by making the wrong mental parallels.
Looking back, even armed with “objective facts” I was unreasonably high on certain players (Mike Condon during 2015-16 season) and unreasonably low on others (Marc-Andre Fleury during the 2015-16 season) for the reasons outlined above.
We do have a large volume of proof that many goalies across the NHL are overpaid for services rendered, and that future goalie performance is subject to much more variance than that of skaters. In Andersen’s case, he might be getting the type of money that a good second-line center may provide. But as it stands now, we have no idea whether we’ll perform as well as a top-line player, or as a below replacement-level one.
The only thing to do then, from a management point of view, is to consider the investment in Andersen (two draft picks, five years’ term and $5M AAV salary) to be a sunk cost, and to keep an open mind if a better, young and cheaper goalie comes along.
As prospect theory indicates, humans are not predisposed to not handle these types of situations optimally – it’s not easy to disregard even unhelpful information when you are working in the confines of an organization (the inside view). So it’ll be up to the Leafs’ coaches and managers to cut bait sooner rather than later if things don’t work out.
But who knows, maybe Andersen could be just what the Leafs need at a good price. Either way, it’s simply too early to tell.
Jack Han is the Video & Analytics Coordinator for the McGill Martlet Hockey team. He also writes occasionally about the NHL for Habs Eyes on the Prize. You can find him on Twitter or on the ice at McConnell Arena.
Hockey Graphs 