The reason is simple: goalies are highly variable. The difference between strong and weak goaltending is only a goal every 150 shots or so, which means it takes thousands of shots before we can be reasonably confident which category to put someone in. That creates a lot of problems for a team that's trying to figure out what to do in net.
Given the variability, I'd argue that the worst thing a team can do is lock up a goalie to a big contract before he has a sustained track record of success. It's easy to take an optimistic view of a goalie coming off a strong season, to think he's figured things out. But what does one good year really prove?
Nine goalies have played at least half the games in each of the last five seasons. Among those five goalies, the smallest range in save percentages across those years is .01.
|Goalie||Best year in last 5||Worst year in last 5||Range over last 5|
When we have a long track record like this, it's easy enough to see who the best and worst goalies are. But looking at how much goalie save percentages fluctuate, it's awfully hard to be confident that one good year should be reason to make a big investment.
That might lead us towards veteran goalies who've performed well for several years, but there's a problem at the other end, too. Most goalies don't get heavy NHL minutes right away, so by the time we have a healthy sample size on them, they're often already at or past their peaks. So while waiting may eliminate the risk that the goalie wasn't really as good as we thought, it adds risk that he may not last as long as we'd hoped.
And let's be honest: we're never really going to eliminate risk with goalies. They're hard to evaluate precisely and they fluctuate a lot from year to year, and that can't really be avoided. Still, there are two approaches that minimize risk.
One option is to target goalies who have a reasonable track record of success but still aren't regarded particularly highly for whatever reason. Jaroslav Halak is one good example -- his save percentage ranks right up there with the high-end talent, yet he has been traded four times already. His $4.5 million cap hit will likely be just the 16th-highest next year, and would be ranked lower still if RFA rights hadn't suppressed the cap hits of a few guys behind him.
In this year's goalie market, James Reimer is the most obvious example from this category. His record is not without blemishes -- he had a bad season in 2011-12, and while some write it off as just the result of playing through a concussion, we can't really know from the outside how big a factor that was.
Reimer also, of course, receives a lot of blame for what happened in Game 7 against Boston in 2013. Personally, I prefer not to focus on such small samples in general, and in particular find it hard to hang a "not clutch" label on a goalie who had a .923 save percentage for the series and pushed the proceedings to seven games despite getting two goals of support or fewer in four of the first six contests. On the whole, I end up thinking he's a pretty viable option.
Just to underscore that there's always risk involved, I'll acknowledge that I would've put Devan Dubnyk in this under-appreciated category a year ago, and he had a positively dreadful year this year. Still, if a team can find a guy like Halak, who has a strong track record yet still comes reasonably cheaply, they should take advantage of that.
Profiting from risk
There's another way to handle that uncertainty: a team can target relative unknowns who come cheaply. In this case, instead of trying to find a player with a strong-but-under-appreciated history, we're looking for cheap lottery tickets. This is how to make uncertainty work for you in a market for a variable commodity, by making multiple cheap bets and seeing which ones pan out.
We've seen several goalie prospects get traded in recent years. Some, like Semyon Varlamov, have fetched the types of high prices that teams shouldn't generally want to pay for relative unknowns. But some, like Sergei Bobrovsky or Ben Bishop, have been available for quite little.
The team has to plan ahead to take this approach. Sometimes, the guy they acquire will turn out to be Anders Lindback, and they need to make these moves at a time when they can withstand a year or two of trial and error if need be. If the team is already otherwise a contender, they can't afford to risk that opportunity on a goalie who is largely unknown.
And like with everything, the organization -- from ownership down -- needs to understand and accept the risks that are being taken. Sometimes the unknown goalie will turn out to be bad and the team will do worse than they did last year; there won't always be obvious measured progress from one year to the next.
The organization needs to accept this to make the right moves. A person can't make rational decisions about the direction of his franchise if he feels that taking risks puts his job on the line. At some point, as the rest of the franchise develops into a group that can reasonably contend, the team may reach a point where overpaying for veteran goaltending is a necessary evil. But it's far more common for teams to make that decision too early than too late.
The uncertainty of goaltending is unavoidable. But given sufficient foresight and commitment to process, smart teams can make it work in their favor.