If you have played fantasy hockey for any length of time, you have been frustrated by goalies. Particularly in recent years the sting of spending a lot of draft capital on a sure thing goalie (be it draft pick or auction dollars) only to have that goalie blow up in your face has become all too common. How are goalies returning investment in 2019-20? Anecdotally, not great. If you drafted Sergei Bobrovsky you probably aren’t all that happy at the moment. Andrei Vasilevskiy didn’t have the best start either, though Tampa as a whole has been better as of late.

This week we are going to take a look at goalie performance (using another helpful Frozen Tools Report) and throw in a little bit of info on average draft position just to make it interesting.

First up I pulled the Quality Starts report on the Frozen Tools Reports page. Quality Starts is a handy stat to look at when thinking about goalie performance. It enables us to capture the goalie starts that would be useful for fantasy purposes. The report also gives us info on Really Bad Starts, which as the name of the stat implies are not at all desirable and tend to blow up your fantasy week.

Currently, the top page of our report looks like this:
 

Darcy Kuemper leads the way for goalies that have actually played a reasonable number of games, followed by James Reimer and Thatcher Demko? Those are definitely not names that were popping up a lot in pre-draft rankings. In this view, we do need to account for games played, and they are both around 20. That low number of games played is certainly less valuable in many formats than Carey Price’s 44 games but still demonstrates that when they see ice time you likely want them on your team.

If we sort by Really Bad Starts instead we see that Brian Elliott and Martin Jones, in particular, are crushing anyone who is misfortunate enough to own them. They are nine and ten on this list but are the only goalies in the top ten with more than ten games played.
 

The question I wanted to answer wasn’t just who are the goalies doing well (or poorly). I wanted to know if the goalies that are going well were drafted accordingly. Or put another way, are highly drafted goalies performing better than goalies who are taken later in the draft. To do this I pulled out some draft data from Yahoo to compare to our QS rankings. Once we add that data to the table and rule out any goalie who has played less than 20 games, here is what our top ten looks like. 
 

(The Draft % column is the number of leagues where the goalie was drafted, and ADP is that player’s average draft position – assuming they were drafted.)

We have six goalies who were drafted in 100% of Yahoo leagues, and their draft positions are all over the place. Vasilevskiy was drafted 12th overall and Darcy Kuemper 134th.

One thing to note here though is these data don’t do a good job of telling a full picture. We can see that Tristan Jarry has an average draft position of 107.5, but that was only in 3% of leagues. In every other league, he wasn’t drafted. In other words, his ADP isn’t really comparable to the other goalies who were drafted in 100% of leagues.

Sometimes it is easier to judge these things with a visual, so if we compare ADP and QS% we get a chart that looks like this.
 

I see dots all over the place. There is a slight trend toward higher QS% with better draft picks but not dramatically so. Those dots at the top of the chart just represent goalies who were not drafted which helps give us an idea as to how they are distributed along the continuum.

But what does a “slight” trend mean in practical purposes? Was it worth it for people to invest picks in goalies?

To figure that out I wanted to know how likely you were to get a good goalie based on where you drafted the goalie. I started with the goalies who were drafted in 100% of leagues and looked at their ADP and quality starts.

The goalies that were drafted in the top 50 picks had a 50% chance to have an elite 60%+ QS%. They had a 25% chance at being an above-average goalie, in the 50-60% range and 25% chance to be worse than average. There were no terrible goalies (40% or less) in that group.
 

As the table shows the number of goalies in the bottom two tiers increases as we fall down the draft. That means that by the time we get to the later drafting rounds (100-150) there is only a 12.5% chance that a drafted goalie ends up in the elite category. It is definitely some evidence that we collectively as fantasy managers tend to draft goalies in an approximate order of their value.

The wrench is when we start talking about the other goalies. The ones that were not consistently drafted. If we take a look at goalies that were originally drafted in less than 30% of leagues (so were likely available to pick up in free agency), and have met the 20-game threshold we see four elite goalies, eight above average goalies, eight below average goalies, and four terrible goalies.  The table below compares the percentage of goalies that fell into the elite, above average, below average, and terrible categories.
 

The results aren’t encouraging. They indicate that a drafted goalie has almost exactly the same chance to perform well (or poorly) as a goalie picked up off the waiver wire during the season. That’s not great. We would think that we draft goalies because they have a better shot at being good than a free agent goalie, and therefore would expect to see more of them in the top two tiers.

The chart does say that drafted goalies have an 8.33% higher chance of being elite (which isn’t great in and of itself), but we do need to keep in mind that we would need to grab a goalie in the top 50 picks to have a 50/50 shot at landing an elite goalie.

So what do we do with all of this? Well, there are a couple of takeaways. In general, when we draft goalies, we seem to be draft them in a reasonable order. Unfortunately, that “reasonable” order still at best gives us a 50/50 shot at a return on investment (i.e. an elite performance at an elite draft position). It also tells us that goalies who were not drafted this year but have played at least 20 games, have almost exactly the same spread in their performance as drafted goalies meaning that our collective wisdom is not really better at deciding which goalies to draft and which goalies to leave in free agency at the start of the season.

The one caveat to this kind of analysis is it doesn’t reward heavily for games played or saves numbers. Many of the focused on during drafts were anticipated starters, while many of the undrafted players were backups or 1Bs who have taken on bigger roles (i.e. Tristan Jarry, James Reimer etc.).

Thanks for reading.

Until next week.