Neil Paine: Building Hockey’s Advanced Stats Dream Team

[Bear of Bad News]

http://fivethirtyeight.com/datalab/building-hockeys-advanced-stats-dream-team/

An interesting exploration of an interesting exercise.

Given the criteria described in his post, who would your "dream team" be?

Click to read more...

 

[Bear of Bad News]

I'd go: Hasek ($4), Lemieux ($4), Orr ($5), Chelios ($1), Ovechkin ($3) and Kurri ($1).

 

[Bear of Bad News]

Dryden's an interesting pick at $2, given his metrics relative to the league:

http://hockeygoalies.org/bio/drydenk.html

 

[Hasbro]

Shanahan ($2) Lemeuix ($4) Jagr ($3)

Orr ($5) Chara ($1)

Brodeur ($3)

My basketball team was

Russel ($4)
Duncan ($3)
Bird ($4)
Drexler ($2)
Stockton ($2)

 

[Bear of Bad News]

Jagr's a nice grab at $3. If I did it again, I'd probably grab Dryden at $2, and replace Hasek with Jagr.

Dryden was 500.6 goals above replacement in his career, which is pretty high on the all-time list.

Then you look at it on a per-60 basis, and it's nuts - Dryden was 1.28 goals better than replacement level, on average, over his career. Granted, he never faced the inevitable decline phase that all other goalies face, but here are the other four on the list:

Hasek: 711.8 goals above replacement, 1.00 per 60 minutes
Roy: 867.0 goals above replacement, 0.86 per 60 minutes
Brodeur: 664.5 goals above replacement, 0.54 per 60 minutes
Belfour: 537.0 goals above replacement, 0.58 per 60 minutes

(Caveat: note that this just takes into account save percentage, and not anything that goalies do to prevent shots - or to create their own team's shots).

This is also regular season. For playoffs:
Roy: 212.2 goals above replacement, 0.84 per 60
Hasek: 90.5 goals above replacement, 0.74 per 60
Brodeur: 119.2 goals above replacement, 0.56 per 60
Dryden: 103.7 goals above replacement, 0.91 per 60
Belfour: 121.0 goals above replacement, 0.73 per 60

Where Dryden looks pretty nice. There may be strength of schedule considerations involved (I've got schedule strength for all of these five guys except for Dryden).

 

[Bear of Bad News]

Here's the player menu, for what it's worth:

 

[Hasbro]

You mention per 6, but Brodeur usually played 70, which was always an advantage in fantasy.

 

[Bear of Bad News]

You mention per 6, but Brodeur usually played 70, which was always an advantage in fantasy.

I meant per 60 minutes (so essentially, per game).

Agreed that Brodeur's durability is an asset - that's a main reason I prefer "vs. replacement" metrics instead of "vs. average" metrics. There's a definite value to being average over a long season (and even more value to being a Brodeur talent who is also durable).

(Although then taking "vs. replacement" metrics and expressing them on a per-game basis is then a bit weird, I'll admit).