quoipourquoi
Goaltender
I agree! As I wrote in that post:
That also works both ways, e.g. Don Edwards didn't necessarily have a massively overrated season in 1977-78 just because Bob Sauve posted a career-high .911 save percentage in just 479 minutes. The next season Sauve came in at .876 in 1608 minutes. Edwards was +.017 vs league average in '78 and +.016 vs. league average in '79. I don't think it is obvious that he was overrated in 1977-78 and really good in 1978-79, just because of what Sauve did in a small sample.
Backup analysis is hard, there are a lot of variables to consider and the sample size creates some real issues. The best is probably to use an estimate that takes into account multiple seasons and some subjective tweaking based on the team's talent and style of play. My main point, again, is that you shouldn't do this for one or two goalies each season while completely ignoring the context of the rest of the league.
But when you post a table that shows a massive .063 spread between Herron and mostly Wilson that the year before would have been just a .005 gap between the same two goaltenders, I don’t know how well it communicates that specific point.
Across the 1970s, Dunc Wilson had a .882 with Vancouver, a .889 with Toronto, and a .887 with Pittsburgh. Had he stayed with Philadelphia, his original team, should we suppose that he would have matched the numbers Favell and Stephenson had in their primes that we are judging Parent against? Or would he play about the same level he did with his three main teams and make Parent look even better?
You say Parent’s 5th isn’t inappropriate, so which 4 goaltenders should have been considered better goaltenders in 1978? And are they so far above him that 1978 isn’t adequate filler for his career?