When it comes to goaltending, bet on consistency

Evaluating goaltenders remains one of the most difficult objectives in hockey, whether you are pouring through video or analyzing data.

So much of the position’s performance hinges on the defensive structure and integrity in front of the goaltender, and our eyes frequently lie to us. What appears to be a great save may be a goaltender merely recovering poor positioning; what appears to be an ugly goal may have been a shot impeded by traffic in front.

But we aren’t flying totally blind (pun not intended). We know the core deliverable of a goaltender is to thwart the opposition and minimize goals against. And we know a goaltender does this by, simply put, making lots of saves.

After hundreds of games, we get a sense for which goaltenders are better than others – Tampa Bay’s Andrei Vasilevskiy, New York’s Igor Shesterkin, and Winnipeg’s Connor Hellebuyck have shown themselves to be a cut above the rest, regardless of how you analyze their play.

From a data perspective, we’ve made advancements to better quantify goaltending performance. Raw save percentage never seemed adequate enough a measure because it is comprehensively clouded by team effects.

More recently, we’ve used Goals Saved Above Expectations, which attempts to baseline all goaltenders by measuring saves in the context of the difficulty of shots faced. This measure may have helped us better differentiate between the likes of, say, Dominik Hasek and Martin Brodeur. Both are Hall of Fame goalies, but Brodeur’s workload – no matter how you quantified it – was routinely less burdensome. Such is the luxury of backstopping a Devils team with impeccable structure and a knack for pushing shooters to the least dangerous areas of the ice.

Goaltenders are notoriously volatile, too. You can draw from endless examples. Between the years of 2016-18, Sergei Bobrovsky – then of the Columbus Blue Jackets – had an argument as perhaps the best goaltender in the game. He was stopping 92.6 per cent of shots across all situations, and bringing in shot quality, had erased a whopping 67 goals over a two-year span.

He parlayed that into a mega deal with the Florida Panthers. Since moving south, has stopped just 90.5 per cent of shots and has been 12 goals worse than expected in the process. It’s been a disastrous ordeal.

For many reasons, this volatility – and the fear of volatility – is why general managers have become reticent to bet heavily in net. Platoons are much more common; teams will either invest more in ‘1B’ or backup options, or bargain hunt for goaltenders on short deals to invest at other positions. Front offices, like the rest of us, are still struggling with how to separate most goalies from one another.

I started thinking about this in the context of today’s goalie crop, and what makes the “better” goaltenders stand out from weaker ones. Ultimately, I think what you want to bet on is consistency. Goaltending performance will fluctuate, but what teams really want is a goalie they can rely on to not lose games. Outperformance at the position is, of course, always welcomed.

Let’s look at 32 regular goaltenders over the past five seasons. We are looking at each goalie’s Goals Saved Above Expectations by year on a percentile basis. To the right we have two bins: years in which a goaltender sank his team’s play, and years where a goaltender outperformed.

What does it look like?

Yost1

Directionally, the table looks right – just look at the names in the top eight or so of the table, followed by the names at the bottom eight of the table. If you take an approach of “give me consistency, or better, just don’t crush my hopes,” you see the goaltenders who stand out.

No one has been more impressive on this front than Darcy Kuemper, freshly signed to a five-year deal with the Washington Capitals. Nashville’s Juuse Saros and Winnipeg’s Hellebuyck aren’t far behind; you can’t hang a single season around their necks, and they have outperformed across several seasons, erasing many more goals than the average netminder.

On the other end of the equation, look at the goalies with higher sink rates. Toronto’s Matt Murray and New Jersey’s Mackenzie Blackwood have really hurt their team’s playoff chances in meaningful ways over the years. Columbus’ Joonas Korpisalo hasn’t put together a strong season yet. And, despite being 41 years old and showing woeful (and deteriorating) performance, the Buffalo Sabres decided to allocate a roster spot for another season to Craig Anderson.

This table isn’t designed to give you a perfect pecking order of best goalies to worst, but it does give you a point of reference for the goaltenders you can have confidence counting on in a given season.

Goaltending is extraordinarily difficult to measure, but we must continue to evaluate these players in a meaningful light. The teams who can marry video and data analysis at the position as a basis for cap management and roster decision-making will win more games in the future.

Data via Evolving Hockey, NHL.com, Hockey Reference, Natural Stat Trick