Post-Game Talk: Whats a "Special Teams" Jets lose 5-3

garret9

AKA#VitoCorrelationi
Mar 31, 2012
21,740
4,385
Vancouver
www.hockey-graphs.com
I don't think anyone believes that save % is a reliable indicator of goaltender performance for a sample size of one game. Reliability increases with sample size, that's basic stats. The more number-savvy guys around here can probably come up with a number of games where shot quality begins to show a somewhat normal distribution.

Pavelec has been pretty solid on this road trip, let's hope it continues.

The number is 3,000 shots ;)
Where quality of shots disappears.
There were years of work done to figure that out haha.


For me, that chart that garret made showing Pavelec's save percentage over his career vs. league average was very telling. Of course quality of shots against has to figure in at some point but when you've only been at or above league average for a very small percentage of your career, that says a lot.

Chart to be shown here soon...


EDIT:

These ones:

http://benwendorf.tumblr.com/post/43447276610/ondrej-pavelec-winnipeg-jets-save-percentage-1
tumblr_mig0q7aYWP1rlkq4mo1_500.jpg

Taking the last 3,000 shots (using this post as my reference; SV% at all strengths) as an indicator of Pavelec’s most recent demonstrated talent, I’ve shown how those 3,000 shots have progressively stabilized to his observed SV% up to the most recent game. The red line denotes stabilization of league-average save percentage over this same period. I might be questioned on using all-strengths save percentage, but when you are using roughly 100-150 games to assess talent and future performance, overall SV% performs virtually identical to even-strength save percentage
http://benwendorf.tumblr.com/post/43449566967/ondrej-pavelec-winnipeg-jets-save-percentage-2
tumblr_mig212dW7P1rlkq4mo1_500.jpg

Looking over the same spread of 3,000 shots as this post, except now we’re looking at 10-game rolling average SV%. As you can see, Pavelec’s performance only briefly moved above league average, on a memorable spell in the early- to mid-season of 2011-12. Otherwise, he’s been pretty consistently below average. As with the previous post on Pavelec, this is using basic SV%.
 

Ad

Upcoming events

Ad

Ad