|
Post by alphalackey on Feb 13, 2023 15:48:42 GMT -7
I will whip up some shootout numbers based on last season's stats to show you just what they looked like. What I was most proud of was that 'sample size' was perfectly encapsulated in the calculation, by converting "percentage" into "percentile" (i.e. converting to standard deviation for N trials) so yeah, a mutant who went 1 for 1 or 2 for 2 wouldn't be pulling the old Ryan Vandenbussche 6-12 range nonsense.
Definitely agree "game starting goalie must take shootout", even Mike Keenan wouldn't swap in a cold goalie on a shootout.
But the ranges met all the standards for me -- there were no large "buckets", each individual performance level had its own rating (obviously two players that both went 4-for-10 would have the same range), and most importantly, you actually had to roll on your goalie a few times. And I'm not just saying this because Saros has two auto-goals against, I'm just saying, if I wanted that shit, I'd still be playing Strato
|
|
|
Post by alphalackey on Feb 15, 2023 12:17:38 GMT -7
I would start with an average range, say it’s 3-12, and factor all skaters into a max range 4-12 and min 2-12 weighted to average based on High Danger Goals per High Danger Shot. Maybe σ+2 is max and σ-2 is min and the rest fall into place between. I just went with their actual shootout success rate -- when I first looked at these stats, it blew my mind how "actual number of NHL goals scored" was a completely worthless metric. Could add penalty shots to the mix as well. For a baseline I used 3-12 for goalies and ~4-12 for skaters (league average for shootout is almost exactly 1-in-3 so going with the standard paradigm would be to double that and let goalies be 50-50). For what the variance translated to, I took that players percentile performance and gave them the range of a similar percentile performance if one person took all the SO shots that year. It gave a bit of a range but nowhere near high or low enough to have players in the 2-12 range. Basically my plan is to start with this method, see what spreads it gives us, and discuss what ranges we'd like to give to the two end points (i.e. what should a 95th percentile forward / goalie look like; what should a 5th percentile look like) and as long as it's strictly increasing, I'm happy with it (i.e. N-Ath percentile < Nth percentile < N+Bth percentile for all A and B)
|
|
|
Post by QC Mike on Mar 6, 2023 23:49:35 GMT -7
Based on the alternate standings we have our first playoff team confirmed. Congrats Belfast! Now you just have to wait for the regular standings to catch up.
|
|
|
Post by Belfast on Mar 13, 2023 18:18:43 GMT -7
Yer gonna have to pry these trophies out of my cold dead hands gents.
|
|
|
Post by alphalackey on Mar 14, 2023 0:02:02 GMT -7
Yer gonna have to pry these trophies out of my cold dead hands gents. Not the "first Z of the year" trophy, you can bury me with that ![:)](//storage.proboards.com/forum/images/smiley/smiley.png) My top 3 D men are average 10 goals a piece, I got two 30-goal forwards and am on pace for four more in the 20s. And Saros is still Saros. PS we should hammer out what the magic number is for "games Brossoit plays before he retires", when I will stop berating you for taking me to town on an emotional trade choice ![:P](//storage.proboards.com/forum/images/smiley/tongue.png)
|
|
VFS Adam
Power User
Truculence is an art
Posts: 123
|
Post by VFS Adam on Mar 14, 2023 9:18:34 GMT -7
I think I might be "winning" a trophy this year, too!
|
|
|
Post by alphalackey on Mar 16, 2023 20:48:42 GMT -7
I think I might be "winning" a trophy this year, too! "Most draft lottery balls by an expansion team that didn't get first overall?" It'll go nice with "most consecutive first overall picks in the post-COVID expansion era" trophy.
|
|
|
Post by bonavista on Mar 24, 2023 7:57:38 GMT -7
Curious as to why I have a Z beside my name in the standings. It looks to be denoting that I am out of the playoff picture but that is not true from what I can see.
I currently have 20 points with 2 games to go. So theoretically i can reach 24 points. This would put me in 4th place.
|
|
|
Post by QC Mike on Mar 24, 2023 20:44:41 GMT -7
Curious as to why I have a Z beside my name in the standings. It looks to be denoting that I am out of the playoff picture but that is not true from what I can see. I currently have 20 points with 2 games to go. So theoretically i can reach 24 points. This would put me in 4th place. My mistake. You lose the tie breaker with Barry but he's currently only in 3rd.
|
|
|
Post by nash on Mar 25, 2023 8:07:13 GMT -7
We'll review the standings at the end of the season to confirm tie breaks.
|
|
|
Post by amack17 on Mar 25, 2023 11:09:32 GMT -7
I am in fifth because Cole Harbour has the same number of points and more wins. I would not worry too much there Steve, hopefully Bonavista and Cremona will pass us both soon enough!
|
|
|
Post by QC Mike on Apr 13, 2023 23:17:55 GMT -7
It's a good thing these standings are unofficial. Realized the Goal Differentials are off and can't be bothered to go through all the results to see where I went wrong in the last month.
|
|
|
Post by nash on Apr 14, 2023 15:51:20 GMT -7
That's fine Mike. I appreciate you doing the updates. I'll confirm everything end of season and send you the final standings so you can edit the table as appropriate.
|
|
|
Post by bonavista on Apr 15, 2023 21:53:58 GMT -7
I am in fifth because Cole Harbour has the same number of points and more wins. I would not worry too much there Steve, hopefully Bonavista and Cremona will pass us both soon enough! not quite. I passed Oxford but not Cole Harbour.
|
|
|
Post by QC Mike on Apr 21, 2023 20:58:45 GMT -7
Find it interesting that in the alternate standings Cole Harbour makes the playoffs on Goal Differential (current sort order has Regulation Wins ranked higher than OT wins but if we went with ROW thay'ld both have 11).
|
|