The Trouble With Drafting Goalies

(This article was originally published on August 14, 2008 on my now defunct personal blog. It stands as perhaps my most seminal piece so I wanted to re-share it here. To some degree, it explains the goaltender market and why teams have generally moved away from drafting puck stoppers in the first round. I have edited it slightly to remove overly dated references.) 

During my look at Sutter’s draft history as the Flames GM, I commented that I was surprised the club drafted two goalies in a single year (Keetley and Lalande in ’05). That and the preponderance of goaltending prospects the organization (the two above plus McElhinney, Irving and James Spratt) has got me wondering about goalies recently, in particular the draft/develop part. 

I’ve come to the shaky but ever strengthening conclusion that drafting goaltenders is just about a complete waste of time.

Bear with me. 

Utility

Goaltending is a wholly unique position and one that is paradoxical by nature. 

It is arguably the most difficult role in team sports and probably the most important one in the game. On the other hand, no other position is more team dependent since a goalie can only exert influence on one half of the goal differential equation: goals against. In a game of six goalies versus six skaters, the former loses every single time, because the latter can both score and defend. Goalies can only play for the shutout: they can’t check anyone or shoot from the point during the power-play.

Therefore, a goalie’s influence is deep, but narrow. Outside of stopping the puck (which, of course, is a vital function) he is pretty much useless. 

Various issues afflict the development of goaltenders. They have vastly different needs and parameters as they progress relative to skaters. The margin of error for prospective ‘tenders is slimmer than any other position: there is no hiding a goalie on the fourth line, or with better line mates, or with a couple of shifts a game. 

A goalie’s night is always 60+ minutes long (barring the hook) and the consequences of his mistakes are always significant (goal against). As such, goalies don’t get the same opportunity to work their way up through a depth chart like forwards and defenders. There are usually 12 or 13 forwards and 6-8 defensemen on any given team, many with differing roles that are parceled out according to idiosyncratic abilities. 

Guys who can’t score, defend. Guys who can’t defend, score. Some do both. Others fight, draw penalties or pester. A few are only good at special teams. What’s more, skaters can slowly grow into their roles, cushioned by the soft bosom of “depth”, meaning older and better guys are there to carry the mail while the kid gains experience. 

There are no such options for Mr. Sieve: if he can’t reliably stop the puck, he’s finished. Even those goalies relegated to back-up, playing a few contests a year, are mercilessly alone in the crease when the game starts. A single bad mistake – a slap shot from center ice that squeaks through your trapper – can end a night or a career.

I can identify two resultant difficulties in goalie development: 

1.) There are only two roster spots for goaltenders at each level: starter and backup. An organization can’t have a lot of goalies on hand with the hope of culling the few sharp ones from a wide field like they can with skaters. Brent Krahn, a Flames former first rounder into whom the club has sunk a bunch of time and money, is out the door this year because there just isn’t any room for him anymore.

2.) High pressure development. Small number of spots + low margin for error = brutal development environment.

For an up-and-comer, displacing an established starter is very difficult. The nature of the position means teams are rarely willing or able to experiment with a new ‘tender or allow him to “find his legs”. No, a goalie has to be at least as good or better than the guy ahead of him in order to get the time and opportunity to play.  

What all this means is:

  • Excess goalies in a system are basically useless. 
  • Working your way up through a system as a goalie is a long, difficult, pressure-filled process.  
  • There are way more goalies than there are roster spots for goalies.

Value

Another apparent paradox: goalies are worth next to nothing on the trade market. Anything outside of an honest-to-goodness established top five puck-stopper might get you a prospect, a draft pick or even nothing at all. 

Again, despite the relative difficulty and importance of the position, the league-wide need is strictly limited. There are around 700 players in the NHL currently: several hundred forwards, several hundred defensemen and about 60 goalies. And a bunch of those are superfluous backups that could likely be swapped out for a whole host of prospects that are bubbling underneath. 

Maybe half of those are starters in the league. And maybe a small percentage of those guys have results that are worth trading for. And those guys are almost never available, meaning there’s a median class of goaltenders that are good enough to be better than their backups, but virtually interchangeable with each other. Again, a team can always be looking for more scoring or more defense or the like, but there really isn’t such a thing as “more goaltending”. There’s only better goaltending and a new goalie has to be a significant improvement above the current guy to garner any interest at all.

Even then, a team that supplants their starter with someone else is almost always stuck with a player and contract that are suddenly expendable, but is worth jack squat on the trade market.

If NHL-caliber goalies aren’t worth much, goalie prospects are worth even less (as much as that’s possible). If the Flames decide to trade any of their current crop (including Leland Irving) they’d be lucky to find an interested partner, let alone any kind of return. Because, again, spots and opportunities are extremely limited. I’d be willing to wager pretty much every single NHL franchise has their own Keetleys and Spratts and Lalandes and Irvings, with a batch of undrafted back-stops in various junior, minor and European leagues to choose from besides.

Because of the various unique issues afflicting the goaltender position, draft order seems to have less predictive power in terms of future success. I say “seems” because I haven’t done a thorough investigation on the topic, but the conclusion strikes me as intuitive. I did take a look at last year’s results with an interest in how well a player’s draft position matched with his performance. First skaters:

This is fairly simple. I took the top 30 scorers from last year and looked at where each guy got drafted. Turns out, 21 of the top 30 scorers (70%) were former first rounders. Only four of the thirty were chosen outside the first four rounds. 

This suggests that GMs significantly improve the chances of landing a viable scorer earlier the draft. Which, of course, makes perfect sense. Higher pick, better player.

Now goalies:

Much different results. Only eight of the top 30 (27%) SV% guys were picked in the first round. Thirteen of the 30, however, were chosen in the fifth round or below (43%). In fact, Ty Conklin and Tim Thomas, who placed inside the top five, weren’t even drafted at all (I have a “9” beside them because I wanted to stick values. Plus, the ninth round doesn’t even exist anymore, so the effect is the same). Tomas Vokoun and Henrik Lundqvist were two of the best goalies in the league last year: they were ninth round and seventh round picks, respectively.

This is a very basic and cursory look at the predictive value of draft order for goalies, but this first glance suggests it is significantly reduced for goalies versus skaters. 

Altogether…

Goalies are very important but have a low utility relative to skaters. There are only a precious few positions for goalies in an organization and they can’t differentiate themselves from competitors for the spots, outside of simply being better at stopping the puck. 

  • Having lots of goalies in a system is rather pointless since goaltending isn’t additive like scoring. 
  • A goalie’s development path is often difficult and long.
  • Goalie supply far outweighs demand.  
  • Outside of the best in the biz, goalies have very little value as trade assets. Goalie prospects are worth even less by several orders of magnitude. 
  • It is very difficult to predict a goalie prospect’s future success. Current SV% leaders were liberally sprinkled throughout the draft. 

Based on all this I’m forced to conclude that drafting goalies is next to pointless. Trading even a first round pick for, say, a decent NHL goaltender whenever you need one (which isn’t that often) seems to make WAY more sense than drafting a goalie prospect with the same pick. The first option is proven at the NHL level and ready to contribute now. The prospect probably won’t be ready for many years and may not be a viable NHL back-stopper anyhow. Meaning a GM in that position has to address the need immediately anyways (trade, sign, waiver wire) and chances are his first round goalie will turn out to be an untradable dud a few years down the road.

It’s hard to develop goalies and hard to determine whether young goalies will turn into bonafide NHLers. And there’s a whole bunch of them out there, to the degree that there always seems to be quality guys looking for work every year. 

Were I an NHL GM, I would focus on skaters and forbid the drafting of goalies, aside from perhaps the odd perfunctory sixth or seventh rounder every four years or so. 

(Edited to add – my position has softened from my initial conclusion. At this point, I’m okay with drafting puckstoppers starting in the third round.)

  • The GREAT Walter White

    I think Ortio, Gillies, Brossoit and McDonald were all fine draft picks. Whether they pan out or not they all make sense to me, yes even McDonald at 34 OA. Keep a goalie starting at every level in the system:

    1) CHL/NCAA/Europe
    1) ECHL
    1) AHL

    And never put hope on any one goalie.

  • Greg

    First thing I notice is that most of those goalies in the “top 30” aren’t even at an average save percentage… “Top 30” is too wide a swath. It’s almost half the league, so the equivalent for skaters would be taking a sample of the top 300 scorers. I’d guess if you did that, you’d end up with a much lower percentage of 1st rounders than 70%.

    If you take a smaller sample of the actual “top” goalies, then yes, there’s some outliers, but the ratio of 1st round picks goes up a lot. Top 5 or 10 is 40%, and top 15 is over 50%. So it seems like, even though you can get good goalies later in the draft, if you want a top one, you’ve still got better odds of drafting one in a higher round.

    That said, it is still a lot of voodoo. I’d bet the top 15 goalies changes so much every year you’d get wildly different results on this each year. I think Kent’s right that trading for a good goalie from another team with a 1st round pick when you really need one does make way more sense. Like if Bishop shakes loose from TB’s cap structure, I’d be willing to trade a 1st for him, assuming it isn’t a lottery pick and he’s willing to sign an extension that fits our cap structure.

    The strategy I’d employ as a GM: pick a huge (6’4″+) goalie with a 4th or less every 2 out of 3 years (maybe an occasional 3rd if you think you’re getting one of the 5 best draft eligible goalies) and hope that 1 pans out every 6-8 years; if you ever have a complete lack of hopeful prospects, gamble a 2nd on a trade for an up and comer (ala talbot, Jones, etc and hope it works ala bishop or kipper… But never do a 1st for a non-sure fire guy); if you have neither capable starter or promising back up, and there’s no reasonable options on the UFA market, pony up a 1st for a legit top-20 starter, assuming it’s not alottery pick and the guy you are getting will buy you at least 5-6 years of time to hopefully develop your own stable again.

  • The GREAT Walter White

    Wow, you were really working the thesaurus to fill this article back in the day! It’s hard to believe teams actually used the number one pick to grab a goalie in the past, I can’t see that ever happening again unless Hasek’s offspring produces a kid with Jedi goalie powers. I guess teams play the odds with drafting goalies high with the hope that “this one” is the “chosen one” for the next 5 to 10 years, but unfortunately Kidd, Krahn, and Irving have proved the theory of walking away from the casino empty handed. Shame!……Shame!

  • Scary Gary

    My idea for drafting goalies is to pick a kid from Europe every year using a late draft pick (round 4-7).

    The European team will develop them and if they turn into something you can bring them over, if they end up doing nothing then it’s not big loss.

  • Scary Gary

    I like the idea of using a pick in the 5-7 rnds range on a tender. The odds of any player panning out in those rounds is so low why not grab the odd goalie here or there. Some great ones have been drafted late like Lundqvist. Goalies take so long to develop it doesnt make sense to use a 2nd or 1st unless you really like the raw talent of the guy and even then it takes so many years of development time it still a crapshoot.

    I like the Gillies pick. I like the Ortio pick. If either guy doesnt make it you havent given up much. I know Kent has said that Goalies dont have much trade value but it is still difficult to get a really good one from the UFA market as teams tend to lock up “thier guy” and it is usually the leftover scraps and players in decline that are available like Millers and Hillers. So I think it is still smart to draft a guy every year or two.

  • The GREAT Walter White

    Do you really trust the same guys who drafted McDonald over Demko to pick our next goalie with a second round pick?

    Do you really trust the same guys who traded Brosoit plus for Smid (because we had too many good goalie prospects…) to pick our next goalie with a second round pick?

    They set up our 2 rivals with goalie depth….

    WW

    • supra steve

      You’re nothing, if not completely predictable, Wally. You’re knowledge of how the future will unfold is beyond reproach.

      That being the case, any good tips for stock picks heading into the new year? Something like “oil sucks, it will NEVER be worth owning”?

      • The GREAT Walter White

        The GREAT WW has said ever since the draft that McDonald was a bad pick.

        I’m not the only one:
        Demko was the consensus #1 goalie in the draft; that means everyone in hockey knew Demko was the better goalie, except BB and BT who thought they were smarter than everyone else…..

        If you are going to go off the board, do it in the later rounds.
        You may discover a gem like Gaudreau or Brodie.

        Going off the board in the early rounds gives you: Baerchi, Backlund, Jankowski.

        Even Poirier is starting to look like a bad pick when Shinkurak was the obvious best pick.

        WW

        • supra steve

          To me, in hockey scouting, “consensus” almost NEVER equals “everyone in hockey”. That’s where you go so terribly wrong.

          On McDavid, I would think the consensus was probably very close to everyone in hockey. But when you get beyond that type of an extreme example, opinions differ. The Flames were willing to spend a higher pick on Gaudreau than any other organization, and that one paid off big. With Janko, it probably won’t, but till he finishes college and spends a few seasons as a pro, who the heck knows what he may become? Did you know what they had in Gaudreau before the later part of last season or the early part of this one? Most people were still unsure how his skills would translate at this level, until he actually got to this level.

          Fact is, no organization makes every pick a great one. The Flames draft record appears to have improved over the last several years, and I’m happy enough with that.

        • The GREAT Walter White

          I believe the love for McDonald comes from his participation with the WHC-17, WJC-18, and now WJC-20 programs, which is prototypical resume to get drafted high. While McDonald has so far walked the path of highly touted goalie there is no guarantee he will work out, nor will Demko if you throw in the Jack Cambell comparison (#11 pick) who was the American wonder boy in his junior years. Campbell is dead last in the AHL in stats and as recently sent down to the ECHL for the second time in his short career, so nothing is guaranteed with goalies, just a plain crap shoot. I agree with another post where you sign or draft a late round guy from Europe, especially from Finland or Sweeden where at the moment they have better training programs for goalies than in North America.