The Predictive Value of CHL Save Percentage

One player that I haven't written about who's often ranked inside the top thirty, and is currently sitting at 50th in the NHL Numbers consensus rankings is goaltender Zach Fucale. I haven't written much precisely because of his position. NHL teams aren't as good at evaluating teenagers and identifying future NHL goaltenders as they are at evaluating teenagers and identifying future defensemen and goaltenders. At least, that's what I found looking at data from 1997-2005 when I did a study on finding quality players in the draft a couple of years ago. An inspection of more recent data doesn't suggest that anything has changed.

One of the reasons that drafting goaltenders is a bad move early in the draft is that it's very difficult for goaltenders to translate CHL success as measured by save percentage to the NHL level. In the chart below, I've looked at every goaltender who was drafted out of the CHL in his first year of eligibility with one of the first one hundred picks in the draft from 2000 to 2007, and presented each goaltender's regular season save percentage in his draft year as well as his career regular season save percentage in the AHL and NHL.

I grant that it mostly looks like a big wall of numbers, but the upshot is that there's very little correlation between a goaltender's save percentage in his draft year and his save percentage in the NHL. There is a positive correlation between the two variables, but at 0.26 (i.e. Pearson's r = 0.26), it's not particularly strong. The problem is that the scouts only add a little bit of value since the correlation between between draft position and NHL save percentage is just 0.34. If we limit the scope to goalies who have played at least 100 games, those correlations both fall dropping from 0.26 to 0.23 and plummeting from 0.34 to 0.10. Part of the problem in that case might be a very small sample (just eleven players), but these aren't numbers that provide a great deal of confidence.

This is especially true when we compare that level of success to the success scouts have with forwards. In the even bigger wall of numbers below, I've looked at every forward who was drafted out of the CHL with one of the first one hundred picks in the draft from 2000 to 2007 (I didn't filter out over-agers because of time constraints, but that would have been a good thing to do), and presented each forward's career regular season points per game in the NHL.

The scouts do much better here. The correlation between draft position and points per game is a much better 0.55. As with goaltenders, the correlation dips when we limit the results to forwards with at least 100 games, but it's still a relatively respectable 0.47.

This alone suggests to me that teams are better off avoiding goaltenders. Scouts just aren't as good at identifying NHL-caliber talent at the position, so you're going to get more value by focusing on the other positions, at least early on.

But there's still another reason to avoid goalies early on: they simply don't contribute much during the first six seasons after being drafted, and with a system that requires getting value for money, it doesn't make sense to be using first and second-round picks on the players least likely to provide value on their entry-level and RFA contracts. In the chart below, you'll see that just five of the thirty-seven goalies played in at least 100 games in the first six years after being drafted.

Even the goaltenders who eventually become starters aren't generally able to do so quickly, and many who do play earlier aren't really helping. Furthermore, there's a lot more risk of outright disaster playing a young goaltender to find out what you've got only to realize he's not very good than there is if you have a similar situation with a player at another position because players at other positions can be sheltered much more effectively. Taking a goaltender early is needlessly risky, and is one reason that I just won't have any near the top of my list.

Next up this afternoon: Mirco Mueller

Previously:

Introduction to Comparables
The Best Offensive Juniors Drafted 11-30
CHL Forwards and Relative Plus Minus
Bob McKenzie's Final Draft Rankings

My Draft List:

1 - Nathan MacKinnon (Comparables)
2 - Jonathan Drouin (Comparables)
3 - Seth Jones (Comparables)
4 - Aleksander Barkov (Comparables)
5 - Elias Lindholm (Comparables)
6 - Valeri Nichushkin (Comparables)
7 - Sean Monahan (Comparables)
8 - Rasmus Ristolainen (Comparables)
9 - Darnell Nurse (Comparables)
10 - Ryan Pulock (Comparables)
11 - Anthony Mantha (Comparables)
12 - Max Domi (Comparables)
13 - Artturi Lehkonen (Comparables)
14 - Alexander Wennberg (Comparables)
15 - Hunter Shinkaruk (Comparables)
16 - Kerby Rychel (Comparables)
17 - Josh Morrissey (Comparables)
18 - Nikita Zadorov (Comparables)
19 - Samuel Morin (Comparables)
20 - Frederik Gauthier (Comparables)
21 - Nicolas Petan (Comparables)
22 - Morgan Klimchuk (Comparables)
23 - Valentin Zykov (Comparables)
24 - Bo Horvat (Comparables)
25 - Adam Erne (Comparables)
26 - Curtis Lazar (Comparables)
27 - Andre Burakowsky (Comparables)
28 - Robert Hagg (Comparables)