Hello everyone and welcome in for another article. In this one I will evaluate how goalies as a whole performed for fantasy hockey versus their average draft position (ADP) back at the beginning of the season. I want to take a second to invite you to join the Apples & Ginos Discord server where I would be more than happy to answer any questions you have about fantasy hockey goalies, 1970s era Ford Pintos, and any other unreliable things you can think of.
Goalies being overvalued is a fact I had known intuitively for years just by playing lots of fantasy hockey. But I had never taken the time to actually try and determine just how overrated they are and just how bad a strategy it is to take them anywhere in the first five or six rounds. The approach I’ve taken here is to examine ADP data from Yahoo on each of the goalies and compare it to their season end finishes. Essentially what this will tell us is whether or not goalies taken early returned better value by season end compared to those taken later. (Quick sidenote: I do have ADP data from ESPN and CBS, however I did not have what I considered to be reliable season end finish data from those sources so I chose to stick to Yahoo for the purposes of this exercise.) I compiled season rank data from Yahoo standard cats league settings, as well as the well-known Keeping Karlsson Ultimate Patron Fantasy League (KKUPFL), which is a points format. Points formats will invariably assign more value to goalies who saw lots of work since saves, wins, and shutouts add up to points and you can’t get those points without being on the ice. It turned out that it didn’t matter much which way I sliced it – the results were absolutely nuclear:
Before I get carried away, let’s break down this chart column by column here. First we have the ADP data, which by the relatively high position of Kevin Lankinen I’m assuming was closed well after the season had already started. Next, I showed the rank of each goalie relative to one another by the Yahoo standard cats league season end finish (Phillipp Grubauer #1, Marc-Andre Fleury #2, Semyon Varlamov #3, etc.). The next column is the Yahoo standard cats league season end finish among all players. Here Grubauer as the #1 goalie was ranked #6 among all players including forwards and defensemen. I also showed the percentage difference between the ADP of that goalie and their season end finish. I did the same exercise using the KKUPFL points scoring, with one addendum. The KKUPFL notoriously devalues goaltenders’ point totals, so I took an average of the Yahoo cats league finishes versus the KKUPFL finishes and came up with a multiplier that roughly values the KKUPFL setup at the same level as the Yahoo basic setup. This is the “ADJ” column, with the “ADJ % Diff” column to match. This simply allows me to compare across the two leagues and ensure it’s not just scoring settings that make the goalies appear less valuable than they actually are.
With the explanations out of the way, the results are immediately obvious. With the sole exception of Andrei Vasilevskiy (and one could certainly argue even he was not a “good” value), the top 14 goalies by Yahoo ADP were horrific draft day investments. There are certainly different reasons for each (Rask/Lehner/Andersen/Kuemper all dealt with significant injuries), but on the whole this is a wildly damning indictment on any strategy involving using a pick on a goalie before round 7 at the absolute earliest. Using an early pick on an underperforming goalie is very tough to recover from, especially if you continue to roster them long after you should have moved on. Let’s be honest, how many of you still had Carter Hart on your rosters in March? That was two months of a third round pick actively hurting your roster. The three position players directly after Carter Hart in ADP? Roman Josi, Gabriel Landeskog, and Max Pacioretty. Rostering a net negative and missing out on studs like those three players is a monumental hit to your roster.
But what about a value over replacement argument? Yahoo standard cats league devote six categories to skaters and four to goalies. Perhaps the raw numbers aren’t good relative to the skaters, but aren’t you giving up a positional advantage if you go with the late goalie strategy? Short answer: nope. The average rank of goalies (column “Y! Rank”) 1-14 by ADP was 24.5, while goalies 15-28 in ADP (removing Francouz and Bishop) averaged out at 23.5. For the KKUPFL points-based rank we get goalies 1-14 at an average of 24.9 versus 22.3 for goalies 15-28. These are arbitrary cutoff ranks to be sure, but the fact that these numbers are even remotely similar let alone favoring the late round goalies should be all the proof anyone requires to realize that you are simply not getting any superior positional advantage by drafting goalies in the early rounds. Moving the goalposts to 1-15 and 16-30 to put Grubauer in the first group changes the averages to 22.9 for the first group and 26.6 for the second group; still less then 4 ranking spots’ worth of positional advantage over a seven round difference in average ADP.
Could you make the argument that Vasilevskiy (and Hellebuyck in the points format) are exceptions to this rule as proven elite workhorses? Possibly. But you’re also giving up a positional advantage at another position by drafting them in the first or early second round. The risk-averse play in my mind is to do what I’ve always done: get ahead at forward and defense and pick up goalies that are hot. This is a position where Juuse Saros can look completely useless for a month and then morph into the ghost of Dominik Hasek down the stretch. There is little rhyme or reason or even discernable trends when it comes to goaltender performance and for that reason I choose to stay fluid riding hot hands at the position, choosing to place my high draft picks on players that are less likely to disappoint me over the course of the season. The inevitable conclusion of this analysis cannot be overstated enough: Goalies are simply not good investments in fantasy hockey.
Now, does this mean that we should avoid drafting goalies altogether? Should we aim for a couple in that second tier? I’m not fundamentally opposed to drafting one or two goalies in that 15-28 tier, perhaps in the seventh to tenth round range. But you’re just as likely to get an Elvis Merzlikins in that range as you are to hit on a Marc-Andre Fleury. If there are skaters with great upside that are still looming large on your draft lists in those mid rounds, fading goalie entirely is a very viable strategy. Think about the waiver wire goalies that emerged this year alone: Jake Oettinger, Alex Nedeljkovic, Mike Smith, Jack Campbell, Vitek Vanecek, Chris Driedger, Kevin Lankinen… with a little bit of patience and intestinal fortitude through the first few weeks of the season, a “Zero-G” strategy is very effective in allowing you room to load up at skater positions and chase waiver goalies all season long. There is a psychological benefit here as well: because you didn’t invest anything in your goaltenders, you won’t hesitate to drop them for whoever is hot this week. There will be no Carter Harts hurting your roster while you patiently wait for them to right the ship, and if you’re being diligent on the waiver wire hunt you could easily end up with a relatively dominant Mike Smith/Jack Campbell pairing that you end up riding straight into the playoffs.
“I will not draft an early round goalie”. Write it down, put it on your flash cards, get a tattoo of it on the back of your hand so you will look at it when you’re about to click on Jack Campbell in the fourth round of your 2021/22 drafts, do whatever you need to. But for the love of god, don’t bet on a goalie in the early rounds.
Thanks for reading, you are appreciated!
Title photo credit: Brian Fluharty