Are these weighted by ice time?
Posts by Aaron K
Letang is suspiciously absent!
Like a balloon and then something bad happens!
Was it a space cow?
Always respect the chain o’ command
The good ship… Planet Express Ship
I think this is missing PHI/PIT
excellent @rwohan.bsky.social on the obvious issue at Tottenham over many seasons -- you can't build a good football team without good passers, and it's weird they tried www.espn.com/soccer/story...
In the strongest possible terms: I approve of your continued existence
This seems related to the changes in rush offense as well (at the NHL level, at least)
Ah yes, just like 1976 when the bicentennial celebrations carried the incumbent party to victory
More convenience than value add, but my GPS also being a portable Wikipedia and an ebook reader is nice
Amarillo, TX
So, here is that fundamental change - a new subdomain has been set up SPECIFICALLY for data scraping, bots and any other forms of automated access.
(and intended to be the only way fairly soon, but more on that a bit later in the thread)
Circa 2018: The Stars should trade their top D prospect (0 GP) for Erik Karlsson
Amazing how many people thought you were describing yourself
Kind of appropriate that 5v5 third periods look like a frown
Can’t believe anyone disputes this
First pass at estimating the origin coordinates for blocked shots (top row) based on the distribution of unblocked shots
(hope my titling is clearer this time)
"For a squad that can make the most of its touches but frequently forfeits the territorial battle, his blend of control, facilitation and potency is nothing short of indispensable."
Great film room analysis on Jason Robertson by Vinh Cao.
open.substack.com/pub/virtualh...
I think this is consistent with your imputation process too?
Captioning a work in progress too, lol
Yeah, my idea was to see if can i use the distribution of fenwicks to guess where there shooter is on an unblocked shot. Looks like most blocks on forwards are closely contested, but defensemen are blocked from further away (to be expected, i guess)
Sorry, that wasn't clear. Blue is fenwick locations, red is shot blocker locations
Heat maps of blocked shots (red) overlaying unblocked shots (blue) at 5v4 for the 2020-21 to 2025-26 seasons
Same thing but for power plays
Defensemen/point shots are fronted less than forwards (the blocks are further from the shooter), but the blocked shot prongs line up with the most common blue line shooting locations
Kind of looks to me like the distribution of shooting locations on blocked shots is not that different from unblocked shots, even though the NHL doesn't record the shooter's location
Heat maps of blocked shots (red) overlaying unblocked shots (blue) at 5v5 for the 2020-21 to 2025-26 seasons
work in progress/interesting: kind of ugly chart overlaying blocked shot locations vs unblocked locations for 5v5 slap, wrist, and snap shots (the types with the most noticeable difference in distribution by position) broken out by position
Ah, that would make sense too. Anecdotally, from coach interviews it seems like there's been more emphasis on not springing the other team off failed attempts in recent seasons
The other thing I thought is the proportion of rush shots that are on odd man rushes may be down (more shots on 2-on-2s vs 2-on-1s, for example). For some reason I thought Corey included odd man rush markers in his entry data, but it looks like I was wrong about that
Ah, that makes sense and would explain some of it. Better rush defense makes sense for rush shots becoming less dangerous than they used to be, but for rush chances to have a negative on an xG value still seems weird to me
Ah, I should try looking at the A3Z data to see how that shows up there