I would say that most of the stats crowd do preach sample size and while you can always look at players and teams in a small size but it varies very quickly....
much like sv% lefty
That said if you look at the top possession teams at the end of the season you also usually see the teams with the most success.
There are always outliers still with this but you also see trends occur.
For example last year the top team was LA as far as Corsi went and they missed the playoffs. They led the league in a goal differential 5v5 and possession, the problem? they had bad special teams play which likely played into them missing the playoffs.
Possession stats are what predicted LA would be a strong team prior to them winning their first cup, they also predicted that Colorado would regress heavily after winning the division, that Calgary this year would regress and that Toronto continuing to win games but having terrible possession stats, was unsustainable.
They also pointed to Tanev and Stralman as excellent players even though they were not putting up points. Something that I think many NHL followers would agree with.
Over the past 5 years the top possession teams have been LA, Chi, Bos, SJ and Stl the worst teams have been Buf, Edm, Col, Tor and Cal.
All 5 SC winners as well as Boston's runner up in that list. I think most people would also agree that the the 5 worst teams are pretty legit for the worst in the league as well.
The numbers that are most interesting to me? Goals and points, baby. I'm sticking with that until the NHL announces the introduction of the new Corsi Cup. - Carol Schram
I would say that most of the stats crowd do preach sample size and while you can always look at players and teams in a small size but it varies very quickly....
much like sv% lefty
That said if you look at the top possession teams at the end of the season you also usually see the teams with the most success.
There are always outliers still with this but you also see trends occur.
For example last year the top team was LA as far as Corsi went and they missed the playoffs. They led the league in a goal differential 5v5 and possession, the problem? they had bad special teams play which likely played into them missing the playoffs.
Possession stats are what predicted LA would be a strong team prior to them winning their first cup, they also predicted that Colorado would regress heavily after winning the division, that Calgary this year would regress and that Toronto continuing to win games but having terrible possession stats, was unsustainable.
They also pointed to Tanev and Stralman as excellent players even though they were not putting up points. Something that I think many NHL followers would agree with.
Over the past 5 years the top possession teams have been LA, Chi, Bos, SJ and Stl the worst teams have been Buf, Edm, Col, Tor and Cal.
All 5 SC winners as well as Boston's runner up in that list. I think most people would also agree that the the 5 worst teams are pretty legit for the worst in the league as well. - belcherbd
Thanks. This is all good stuff, and you've explained it well.
Still, I feel like I can reach a lot of the same conclusions through intuition and/or basic stats. Those Top 5 possession teams have generally hovered near the top of the standings, while the Bottom 5 have been in or near the cellar - except for that one crazy year in Colorado. We know they're good & bad without having to look at possession numbers.
Do you know if there are any examples of Corsi darlings who haven't fulfilled their potential like Tanev and Stralman did? I'd bet there's a few of them out there.
These days, I'm frustrated with the guys on 1040 giving Corsi numbers for the Canucks players after each game. That's why I have the bee in my bonnet about sample size: I agree that most of the stats guys who truly analyze these numbers do a better job of putting them into context, and that's what I liked about Jeremy's article.
"Here is just a small sample of what’s new from the reinvented Steamer’s concession. The chefs have introduced 12 new gourmet hot dogs and sandwiches, priced between $5.50 to $7.50. Here is a small sample:"
"Here is just a small sample of what’s new from the reinvented Steamer’s concession. The chefs have introduced 12 new gourmet hot dogs and sandwiches, priced between $5.50 to $7.50. Here is a small sample:" - hillbillydeluxe
Haven't tried it myself, but one of the guys in the press box gave the new Maui FireDog a hearty thumbs-up a couple of weeks ago. He said it cost him $9.
"Here is just a small sample of what’s new from the reinvented Steamer’s concession. The chefs have introduced 12 new gourmet hot dogs and sandwiches, priced between $5.50 to $7.50. Here is a small sample:" - hillbillydeluxe
They look real good in the promo highly doubt they are that good at the game?
When I was at the Mariners games last April the hot dogs in the ball park were so disappointing. And one would think ball park franks would be unreal. All else was great from nachos to fish to pizza and the craft American beer but wow was it an expense to eat at the game. After a few I just bought the $20 seats in the home run zone and went and ate at the pub after after.
Thanks. This is all good stuff, and you've explained it well.
Still, I feel like I can reach a lot of the same conclusions through intuition and/or basic stats. Those Top 5 possession teams have generally hovered near the top of the standings, while the Bottom 5 have been in or near the cellar - except for that one crazy year in Colorado. We know they're good & bad without having to look at possession numbers.
Do you know if there are any examples of Corsi darlings who haven't fulfilled their potential like Tanev and Stralman did? I'd bet there's a few of them out there.
These days, I'm frustrated with the guys on 1040 giving Corsi numbers for the Canucks players after each game. That's why I have the bee in my bonnet about sample size: I agree that most of the stats guys who truly analyze these numbers do a better job of putting them into context, and that's what I liked about Jeremy's article. - Carol Schram
I agree anyone spewing Corsi stats in a small sample size is annoying and 1040 are already annoying as it is.
We know afterwards that they should be there but at the time Col was winning, there was a large debate about whether that possession stats meant anything. Same with LA in 2012 because only a few were predicting dominance.
Looking last years leaders, the only guy who stands out to me in the top 20 is trevor lewis who came in as the 18th best player as far as CF%. He still had a decent season considering his role though.
Semin would be ranked at 25 right after Crosby and before Letang, and Semin only had 15 EV points in 57 games. I would consider him as a guy that should of produced more but didn't.
John Micheal Lies at 40 with 15 EV points in 57 games.
David Booth had good possession numbers in Van I think but was a disappointment statistically. I'm sure there are many more player we could find with good corsi and are poor players.
My point with Stralman and Tanev was that judging them by traditional stats, points, hits, +/-, etc didn't show how good of players they are, while possession stats have shown them to be consistently very impressive.
