The Sabres as a possession team

[dudacek]

Some individual player conversations and the weight that some people put on possession numbers inspired me to to take a deeper look at how the Sabres stack up.

First of all, some context: shots for and against charts generally follow team and individual success, but they don't necessarily dictate it.

Nobody will be surprised to learn Carolina, Florida, Edmonton and Dallas were top 5 in this area last year.

They may be surprised to learn LA was 4th, New Jersey 7th and Philadelphia 8th.

Buffalo was right in the middle of the pack, finishing 16th at 50.7 per cent. Teams right around them include Toronto, Tampa, the Rangers the Senators, the Jets and the Flames.

Detroit and Montreal were bottom five and Boston, Washington and the Islanders weren't much better.

https://www.nhl.com/stats/teams?report=percentages&reportType=season&seasonFrom=20232024&seasonTo=20232024&gameType=2&sort=satPct&page=0&pageSize=50

[dudacek]

The New Jersey Devils under Lindy Ruff were an excellent puck possession team:

They ranked 5th last year at the time of his firing, and 4th overall the year before that. Even the 28th-place team of three years ago finished above 50% and 14th overall

Edited August 24 by dudacek

[dudacek]

The gold standard for possession teams is Carolina: adjusting for players who played at least 70 games, every one of the top 9 possession players in the league were Hurricanes.

Jesper Fast and Jordan Staal led the way with just over 63 per cent.

Only 11 NHL players hit 60% and only 43 hit 56%

50%, obviously, is the mark of a middling possession player

Edited August 24 by dudacek

[dudacek]

The best possession player last year currently on the Sabres is Ryan McLeod. He ranked 34th overall at 56.7

He put up 53.9 the previous year in 57 games as a sophomore.

[dudacek]

The other players the team acquired recently aren't nearly as strong.

• Bo Byram was 50.2% in Colorado, but dropped to 46.5 in Buffalo
• Jason Zucker was 47.7 in the desert, 49.1 in Nashville
• Sam Lafferty was 48.4 in Vancouver
• Dennis Gilbert was 49.2 in Calgary
• Nick Aube-Kubel was 43.2 in Washington
• Beck Malenstyn was 40.7 in Washington

The board has talked elsewhere about how the latter pair's numbers were coloured by historically difficult defensive deployments. Which is a good spot to point out that possession numbers are certainly influenced by who you play with, when and where. I did not dive into the deployments of the other new acquisitions

[dudacek]

On the flip side, the players the Sabres walked away from included their three worst possession performers last year.

• 21 Tyson Jost 43.6
• 20 Erik Johnson 46.6
• 19 Victor Olofson 47.0

Other guys who moved on:

• 15 Zemgus Girgensons 49.7
• 12 Casey Mittelstadt 51.0
• 11 Eric Robinson 51.2
• 10 Jeff Skinner 51.2
• 9 Kyle Okposo 51.3

I did not dive into their deployments either

Edited August 24 by dudacek

[dudacek]

The Sabres best possession player last year was their best player.

Rasmus Dahlin recorded a 53.7% total, tied for 84th in the NHL with Nikita Kucherov, Adam Fox and Vladislav Gavrikov out of players who played at least 70 games.

What I thought was an interesting bit of context:

• Of the NHL's most-used defencemen (more than 18 minutes ES per night), Dahlin ranked 9th overall.
• In that same category Owen ("trainwreck in his own zone") Power ranked 12th with 51.5%

Edited August 24 by dudacek

[dudacek]

The Sabres 3 worst returning possession players >40 games played are

• 18 Jordan Greenway 47.3 (tied with Patrick Kane and Erik Cernak at 437th out of 611 skaters)
• 17 Henri Jokiharju 48.7 (366)
• 16 Connor Clifton 49.1 (335, tied with Brad Marchand)

The lowest ranked NHL player with more than 40 games played was Boston's John Beecher at 36.9%

[dudacek]

After Dahlin, the Sabres' best returning possession players >40 games played are

• 2 Tage Thompson 53.2
• 3 Zach Benson 53.0
• 4 JJ Peterka 52.8
• 5 Ryan Johnson 52.5
• 6 Peyton Krebs 51.9
• 7 Mattias Samuelsson 51.5

All stats from NHL.com

https://www.nhl.com/stats/skaters?report=puckPossessions&reportType=season&seasonFrom=20232024&seasonTo=20232024&gameType=2&playerPlayedFor=franchise.19&filter=gamesPlayed,gte,40&sort=satPct&page=0&pageSize=50

[Pimlach]

meh 

[dudacek]

8 minutes ago, Pimlach said:

meh 

To what? Possession numbers as a meaningful stat, or the numbers themselves?

[Pimlach]

1 hour ago, dudacek said:

To what? Possession numbers as a meaningful stat, or the numbers themselves?

Yeah, kind of all of it.  Seems very subjective to collect meaningful data for one.   Some of this depends to some extent by how a player is used and who they play with.  

But thanks for providing, it is something else to ponder until we get the season started.  

Edited August 24 by Pimlach

[dudacek]

1 hour ago, Pimlach said:

Yeah, kind of all of it.  Seems very subjective to collect meaningful data for one.   Some of this depends to some extent by how a player is used and who they play with.  

But thanks for providing, it is something else to ponder until we get the season started.  

I really think possession numbers aren't much different than +/-: they're an attempt to apply a team stat to an individual player and are moderately useful in context.

While possession (SAT) makes obvious gains by expanding the sample size it also loses by tracking something far less important when it comes to the outcome of a game.

Something else I consider when comparing a 52% player to 48% player is how that translates to being outshot 26 to 24 in a game. I know @thorny is going to tell me every inch matters, and he's not wrong. But I can't help but think that when comparing two players it's not something too dramatic;

[Thorny]

12 minutes ago, dudacek said:

I really think possession numbers aren't much different than +/-: they're an attempt to apply a team stat to an individual player and are moderately useful in context.

While possession (SAT) makes obvious gains by expanding the sample size it also loses by tracking something far less important when it comes to the outcome of a game.

Something else I consider when comparing a 52% player to 48% player is how that translates to being outshot 26 to 24 in a game. I know @thorny is going to tell me every inch matters, and he's not wrong. But I can't help but think that when comparing two players it's not something too dramatic;

Personally imo it’s far more useful than plus minus, but not to the extent traditional counting stats I DO care about lose value 

I see them as predictive: but while I might be cautious about the guy with 42 goals and surprisingly poor metrics being able to replicate that total going forward (see: predictive) they do *nothing* to tarnish what IS. Ie, that 42 goals is full value that season regardless of underlyings 

Edited August 25 by Thorny

[PerreaultForever]

Problem with possession stats is if you are a perimeter team your stats look better than your record. Without dirty net front presence possession means very little. 

I don't know what Ruff plans to do in terms of a team system but I am hoping we become a forecheck team more than a possession team. Creating errors on the other team is, in my opinion, far more conducive to winning than possession, especially if it is primarily perimeter and neutral zone possession. 

Florida won the cup. The league is going to keep shifting to greasy and mean and Sabres need to get with the program. 

[Thorny]

2 minutes ago, PerreaultForever said:

Problem with possession stats is if you are a perimeter team your stats look better than your record. Without dirty net front presence possession means very little. 

I don't know what Ruff plans to do in terms of a team system but I am hoping we become a forecheck team more than a possession team. Creating errors on the other team is, in my opinion, far more conducive to winning than possession, especially if it is primarily perimeter and neutral zone possession. 

Florida won the cup. The league is going to keep shifting to greasy and mean and Sabres need to get with the program. 

This was LA until Pittsburgh did it with speed and pace of play. If they sabres try to copy, the time it’ll take to catch up will leave then hilariously behind the trend anyways.
 

Just build a balanced and talented team 

[PerreaultForever]

Just now, Thorny said:

This was LA until Pittsburgh did it with speed and pace of play. If they sabres try to copy, the time it’ll take to catch up will leave then hilariously behind the trend anyways.
 

Just build a balanced and talented team 

We don't have Crosby. 

[Thorny]

Just now, PerreaultForever said:

We don't have Crosby. 

Well they played Nashville in the final and we have a Saros 

Or was it Rinne? 

We also have a Rinne

[PerreaultForever]

Just now, Thorny said:

Well they played Nashville in the final and we have a Saros 

But they got to the final and that wasn't Saros, wasn't it Pekka Rinne? 

Anyway, do we have a Saros? 

[Pimlach]

7 minutes ago, Thorny said:

This was LA until Pittsburgh did it with speed and pace of play. If they sabres try to copy, the time it’ll take to catch up will leave then hilariously behind the trend anyways.
 

Just build a balanced and talented team 

This about sums it up.   And a great coach will put that balanced and talented team in a position to best utilize their skills.     That is real coaching, not copy the style of the latest Cup winner with the wrong horses to do it.  

[Brawndo]

8 hours ago, dudacek said:

The New Jersey Devils under Lindy Ruff were an excellent puck possession team:

They ranked 5th last year at the time of his firing, and 4th overall the year before that. Even the 28th-place team of three years ago finished above 50% and 14th overall

For those who like analytics, they will be happy to hear he had a great relationship with the Devils Analytics Department, he frequently met with them and incorporated their suggestions into his game planning 

[LGR4GM]

12 hours ago, dudacek said:

I really think possession numbers aren't much different than +/-: they're an attempt to apply a team stat to an individual player and are moderately useful in context.

While possession (SAT) makes obvious gains by expanding the sample size it also loses by tracking something far less important when it comes to the outcome of a game.

Something else I consider when comparing a 52% player to 48% player is how that translates to being outshot 26 to 24 in a game. I know @thorny is going to tell me every inch matters, and he's not wrong. But I can't help but think that when comparing two players it's not something too dramatic;

This is why teams use xgf% because it contextualizes shots.

10 hours ago, Brawndo said:

For those who like analytics, they will be happy to hear he had a great relationship with the Devils Analytics Department, he frequently met with them and incorporated their suggestions into his game planning 

The Devils got worse analytically after Ruff left too.

I like Lindy a lot and I'm hoping for the best.