mossey3535
Registered User
- Feb 7, 2011
- 13,571
- 10,246
Ya, no I don't agree with that.
Here's something to chew on:
2008-2013 drafts:
League average GP per pick: 87.1
St.dev: 149.9
N: 1126 picks
Canucks average GP per pick: 41.4
St.dev: 90.2
N: 37 picks
Mean difference has a P value of .0654 or nearly significant at 95% CI, significant at 90% CI. The effect size is .369 (cohen's d), which can be interpreted as about halfway between a small and medium effect size. The numbers bear out the Canucks were either extremely unlucky or extremely bad at drafting. I view it as the former, you clearly view it as the latter. Not sure we are gonna resolve this anytime soon.
Edit: Also for further context, here's how NHL teams stack up for average GP/pick from 2008-2013:
I'll concede that Vancouver's regular season success put them at a disadvantage, particularly in the first round but still, those are some grizzly results.
In general good work but when you say "nearly significant at 95%", it doesn't convince me. 95% CI is the bare minimum I think is credible. If it's not significant at that level it's probably random variance, as Melvin asserts (although I disagree with some of his reasons for asserting that).