Why NHL goaltending is a buyer’s market

Despite a rocky relationship with Randy Carlyle n 2013-14, he doesn't have any qualms about playing for the Leafs coach.

There are six NHL teams whose starting goalies will become unrestricted free agents in the summer of 2014: Buffalo (Ryan Miller), Edmonton (Devan Dubnyk), Florida (Tim Thomas), New York Islanders (Evgeni Nabokov), New York Rangers (Henrik Lundqvist) and the St. Louis Blues (Jaroslav Halak). In addition to those six teams, Calgary’s goaltending is something of a mystery, with their needs likely dependent on how Karri Ramo performs this year. Philadelphia (Steve Mason) and Colorado (Semyon Varlamov) will both have restricted free agent starting goaltenders to deal with when the summer comes.

It’s plausible that Thomas and Nabokov will retire, which would reduce the number of goalies in search of jobs. However, in addition to the remaining four unrestricted starters, Jonas Hiller, Brian Elliott and Ray Emery will all be available to the highest bidder. There will be at least seven plausible starting goaltenders on the market, but only six teams with openings for a starting goaltender—plus some potential demand from Calgary and Philadelphia.

Some of those six teams will be hoping to fill their starting jobs internally. Florida has been waiting patiently for Jacob Markstrom to become an NHL starter They are presumably hoping that he can take the job from Thomas this year. Jhonas Enroth was a second round pick in 2008 in Buffalo who has played pretty well in the first three years of his apprenticeship to Miller. If they can seize those jobs, there could plausibly be as few as four teams that need to find a starting goalie come the summer, and as many as nine goalies with experience serving as starters chasing those jobs.

Why does this matter right now? Well, it’s one the reasons that it’s exceedingly difficult to trade a goalie for significant value. If you’re the Ottawa Senators, hoping to trade Craig Anderson at some point in order to allocate scarce salary dollars more efficiently, or the Toronto Maple Leafs, needing to trade James Reimer at some point before 2015 or risk seeing him leave as a UFA, the first hurdle that you run into is the fact that there are more goalies with starting pedigree than there are jobs.

The second hurdle is the fact that goalies are very difficult to project with any degree of certainty. This factors into the price that NHL general managers are willing to pay for them, both in terms of trade value and salary, reducing it substantially.

That’s something of a contradiction, because a goalie who stops a lot of pucks is immensely valuable. In Reimer’s rookie season, he played 37 games for the Leafs and posted a .921 save percentage on 1134 shots. The average save percentage for a starting goalie in 2010-11 was around .915. Some arithmetic will show you that Reimer’s save percentage means that he prevented 6.8 goals more than the average starter would have on 1134 shots.

If you were to compare Reimer’s performance to that of the NHL’s backup goalies in 2010-11 (something known as replacement level analysis), the difference is even greater. Backup goalies had an average save percentage of .907 in 2010-11. If you do the arithmetic, you find that Reimer saved 16 goals that a league average backup would have let in.

How much are 16 goals worth to a team? A quick rule of thumb applied in hockey analytics is that three goals are worth one point in the standings. So 16 divided by three equals 5.33 points. That’s how much more Reimer was worth to the Leafs in 2010-11 over a league average backup. Keep in mind—this was accomplished in only 2080 minutes of ice time. The thirty most heavily used goalies in the league averaged 3368 minutes. If you pro-rate Reimer’s work over that much time, you come up with a goalie who would have been worth more than four wins above replacement level with an average starter’s workload.

That’s a very valuable thing. If a team could acquire a goalie and have some degree of certainty that he would post a save percentage of .920 or greater, it would be worth paying a lot for. Unfortunately, one of the things that people who dig into hockey data have learned is that the margins between an excellent goalie and a backup are small. Less than one save per hundred shots separated the average starting goalie from the average backup in 2011-12. Small margins mean that it’s easy for chance to overwhelm the skill differences between goalies. A goalie with a true talent of .920 can appear to be worse than one with a true talent of .910, and for an extended period of time.

A look at the data confirms this. Thirty-one goalies played at least 100 games between 2010-11 and 2012-13. This is a list that is essentially made up of the NHL’s starting goaltenders. Their performance, as measured by save percentage, is extremely volatile. Eleven of those had at least one season (minimum 30 games) with a save percentage of .910 or worse and another season with a save percentage of .920 or better in that time. In a three-year window, the same goalie can frequently produce both an exceptional save percentage season for a starting goalie and a poor one.

If you can’t be certain that a single-season performance is driven by talent rather than chance, spending a lot, in the form of talent or money, to acquire a player isn’t sensible. It takes some humility to admit that you can’t be really certain about a player’s talent without having seen the player face thousands and thousands of shots, but it’s the rational thing to do.

Unfortunately for the Maple Leafs, if teams act this way, it means that there’s unlikely to ever be much of a willingness on the part of another team to give up much for Reimer. Greater supply than demand and high degrees of uncertainty about a goalie’s skill and contribution mean that, even if a player has appeared to be very good at times, the market won’t be there, assuming everyone acts rationally.

People don’t always act rationally—the Bruins were able to trade Andrew Raycroft for Tuukka Rask and the Sharks were able to swap Vesa Toskala for a package that included a first round pick that they ultimately turned into Logan Couture. However, it’s rare enough that it’s not really worth holding out hope for. Assuming Reimer is now the full-time backup in Toronto, if he gives the Leafs a solid goaltending duo for the next two years and then leaves as a free agent, they’ve done about as well as they can reasonably hope.

When submitting content, please abide by our submission guidelines, and avoid posting profanity, personal attacks or harassment. Should you violate our submissions guidelines, we reserve the right to remove your comments and block your account. Sportsnet reserves the right to close a story’s comment section at any time.