Jump to content

GDT: Sabres at Ducks ?7p PT


pi2000

Recommended Posts

No change to the Sabres Line up. 

LW C RW Forwards
Victor Olofsson
VICTOR OLOFSSON

Rating: 74.5#37 LW

Jack Eichel
JACK EICHEL

Rating: 80.2#8 C

Sam Reinhart
SAM REINHART

Rating: 75.5#22 RW

FL1 Rating
Rating: 76.73#14 FL1
Jeff Skinner
JEFF SKINNER

Rating: 77.3#16 LW

Marcus Johansson
MARCUS JOHANSSON

Rating: 73.5#75 C

Vladimir Sobotka
VLADIMIR SOBOTKA

Rating: 68.9#178 RW

FL2 Rating
Rating: 73.22#17 FL2
Jimmy Vesey
JIMMY VESEY

Rating: 70.9#110 LW

Casey Mittelstadt
CASEY MITTELSTADT

Rating: 71.3#132 C

Evan Rodrigues
EVAN RODRIGUES

Rating: 72.3#68 RW

FL3 Rating
Rating: 71.49#22 FL3
Zemgus Girgensons
ZEMGUS GIRGENSONS

Rating: 70.8#147 LW

Johan Larsson
JOHAN LARSSON

Rating: 70.7#150 C

Kyle Okposo
KYLE OKPOSO

Rating: 73.0#54 RW

FL4 Rating
Rating: 71.48#13 FL4
 
DEFENSIVE PAIRINGS Defense
Jake McCabe
JAKE MCCABE

Rating: 71.9#89 LD

Rasmus Ristolainen
RASMUS RISTOLAINEN

Rating: 73.5#46 RD

DL1 Rating
Rating: 72.68#27 DL1
Rasmus Dahlin
RASMUS DAHLIN

Rating: 75.8#13 LD

Colin Miller
COLIN MILLER

Rating: 75.5#22 RD

DL2 Rating
Rating: 75.65#3 DL2
Marco Scandella
MARCO SCANDELLA

Rating: 71.7#95 LD

Henri Jokiharju
HENRI JOKIHARJU

Rating: 72.5#10 RD

DL3 Rating
Rating: 72.10#22 DL3
1ST POWERPLAY UNIT
Victor Olofsson
VICTOR OLOFSSON

Rating: 74.5#37 SK1

Jack Eichel
JACK EICHEL

Rating: 80.2#8 SK2

Jeff Skinner
JEFF SKINNER

Rating: 77.3#16 SK3

Sam Reinhart
SAM REINHART

Rating: 75.5#22 SK4

Rasmus Dahlin
RASMUS DAHLIN

Rating: 75.8#13 SK5

 

LW C RW Forwards
Rickard Rakell
RICKARD RAKELL

Rating: 75.2#37 LW

Adam Henrique
ADAM HENRIQUE

Rating: 73.0#84 C

Jakob Silfverberg
JAKOB SILFVERBERG

Rating: 75.7#26 RW

FL1 Rating
Rating: 74.67#26 FL1
Maxime Comtois
MAXIME COMTOIS
Ryan Getzlaf
RYAN GETZLAF

Rating: 74.7#45 C

Ondrej Kase
ONDREJ KASE

Rating: 77.2#12 RW

FL2 Rating
Rating: 75.98#5 FL2
Nick Ritchie
NICK RITCHIE

Rating: 70.8#116 LW

Sam Steel
SAM STEEL

Rating: 73.7#61 C

Troy Terry
TROY TERRY

Rating: 72.4#66 RW

FL3 Rating
Rating: 72.28#11 FL3
Max Jones
MAX JONES

Rating: 72.9#66 LW

Derek Grant
DEREK GRANT

Rating: 70.0#164 C

Carter Rowney
CARTER ROWNEY

Rating: 69.2#107 RW

FL4 Rating
Rating: 70.69#25 FL4
 
DEFENSIVE PAIRINGS Defense
Hampus Lindholm
HAMPUS LINDHOLM

Rating: 74.7#28 LD

Josh Manson
JOSH MANSON

Rating: 72.3#71 RD

DL1 Rating
Rating: 73.51#21 DL1
Michael Del Zotto
MICHAEL DEL ZOTTO

Rating: 72.1#85 LD

Cam Fowler
CAM FOWLER

Rating: 75.5#17 RD

DL2 Rating
Rating: 73.77#19 DL2
Jacob Larsson
JACOB LARSSON

Rating: 71.4#13 LD

Korbinian Holzer
KORBINIAN HOLZER

Rating: 71.9#79 RD

DL3 Rating
Rating: 71.67#24 DL3

 

Betting line:

sabres.thumb.JPG.2cd13f1bfb35c076808d93c3101bc115.JPG

I'll be there too. 

Link to comment
Share on other sites

3 minutes ago, Mustache of God said:

I don't think i'll ever understand what those line-up numbers mean. Why does Sobotka look like lava?

By some aggregate stat, Boat is #178 in the league for RW. He's lava because generally speaking, if there are 31 teams (and assuming an even distribution of talent), RWs #1 through #31 should be allocated on one per team for the first line, #32 through #62 for the second line, etc. As I understand it, to put it another way your "first line" should be players between 1 and 31, second line between 32 and 62, etc. Having #178 [0] on 2nd line means he's playing on a line well above where his stats say he should (tweener 6th and 7th line). Remember this is small sample size right now, so a couple goals or whatever factors that stat uses probably creates big swings in ratings. If Boat is still 178 in January, there's a question with either his play or how the stat is calculated.

[0] If I'm reading right, there are 5.7 RWs in the league *per team* that are better than Boat.

  • Like (+1) 2
Link to comment
Share on other sites

49 minutes ago, Mustache of God said:

What goes into determining that number? Fancy stats? Ugly stats? Stats that look so-so after a few ICE COLD LABATTS?

That I don't know. I figure there are a lot of smart people that look at it and it probably bears out results and is adjusted over time to get better. 

Link to comment
Share on other sites

Man, the consistency of the lines and pairings is so refreshing. It'll be interesting to see what Krueger does if the team goes on one of those patented Buffalo Sabres scoring droughts.

1 minute ago, MattPie said:

That I don't know. I figure there are a lot of smart people that look at it and it probably bears out results and is adjusted over time to get better. 

I'm pretty sure you just went 0-3 on assumptions.

  • Like (+1) 1
Link to comment
Share on other sites

4 minutes ago, PASabreFan said:

I'm pretty sure you just went 0-3 on assumptions.

Well, looking at the actual list I don't see anything out of place. Are you arguing Svobotka isn't the 178th centre in the league?

Here's the list, with a link to the detailed rating system: https://www.corsicahockey.com/nhl/players/nhl-player-ratings-rankings

Link to comment
Share on other sites

3 minutes ago, MattPie said:

Well, looking at the actual list I don't see anything out of place. Are you arguing Svobotka isn't the 178th centre in the league?

Here's the list, with a link to the detailed rating system: https://www.corsicahockey.com/nhl/players/nhl-player-ratings-rankings

RW you mean? How can he be the 178th-ranked right wing?

Link to comment
Share on other sites

2 minutes ago, PASabreFan said:

RW you mean? How can he be the 178th-ranked right wing?

I think he's listed as a center, so he gets rated that way even if he's playing wing (my original post was probably wrong). There seem to be a lot more centers than other forward positions, as there are 182 centers and 159 LW and 153 RW [0]. Larsson is the second-lowest LW, BTW.

In any case, don't trust me; go look at the numbers and decide for yourself. I don't follow hockey that closely any more, but looking at the first page of forwards I don't see anything out of place by much. We can argue a few places up or down, but I don't have issue with Eichel being #8 or Reinhart being #22 (center). Eichel would be the best center on most teams. If you look at RWs, Reinhart's rating puts him around Kessel and Laine.

[0] makes sense, I think the Sabres only have like 2 guys that are RW, Reinhart and Boat being centers. Makes me wonder if all the forwards are rated the same way or if they tweak the weights for center vs wing for things like defense stats.

Link to comment
Share on other sites

NHL Player Ratings Explained

Motivation

Hockey stats have come a long way since the days of Plus-Minus and PIMs. A plethora of sophisticated evaluative metrics now exist for skaters and goaltenders alike, including Game Score and WAR. These new-age stats each individually represent vast improvements over mainstream norms, but even they are prone to blind spots. Inevitably, metrics developed by different people using different methods will each have their own unique strengths and weaknesses.

We believe this diversity should be celebrated. Moreover, we see a tremendous opportunity for combination or “stacking” of these metrics. This principle is central to our Player Ratings.

Stacking is most effective when the base models are significantly different.”

 

Method

For more reasons than one, combining player metrics into a singular rating is not straightforward. Namely, we encounter the following hurdles:

  1. What “representation” of each statistic should be used?
  2. How can we combine metrics that may not resemble one another?
  3. What relative importance should be given to each metric?

The first challenge can be demonstrated using Game Score as an example. Consider a player having appeared in 200 or more games over the span of 3 seasons. If you were asked to summarize this player’s performance to date using Game Score, what might your approach be? One option is to simply divide the player’s sum Game Score by 200 to obtain an average Game Score per Game (GS/GP). In doing so, we assign the same importance to games played 3 seasons ago as games played in the current or last season, which may not be optimal. Additionally, this approach may unfairly flatter players who are given more minutes.

This is what we mean by using the best “representation” of a given statistic. Our approach to solving this problem was to develop a method of testing the “usefulness” of a particular statistic, then applying this procedure to a large array of different versions – effectively turning this into an optimization problem. Usefulness here was defined as follows: if measures of team roster strength (separated by player position) were taken using a given statistic, how successfully could we predict game outcomes using them in addition to basic predictors like home ice advantage and fatigue? This framework was used to select representations of each player metric. For Game Score, the optimal version was found to be a recency-weighted average of GS/GP over the previous 120 games.

Another important consideration is how to curb results obtained in small samples such that they do not put forth an unfair representation of a player’s ability. Our solution is to regress a player’s metrics towards a known average in cases where they have not yet exceeded an arbitrary Games Played threshold. For a given parameter P and metric M, the mean value of MmM is computed for all players whose cumulative games played are fewer than P. A regularized version of MM’ is then calculated in cases where a player has played fewer games than P.

M’ = (GP * M + (P – GP) * mM)/P

In order to produce a single number from this set of components, we opted to use a linear combination:

pr.jpg

where Mi is the ith regularized metric and wi the corresponding weight. It is clear that this framework becomes totally dependent on the values given to the yet undefined weights w.

Thankfully, we have already defined a method for expressing the “usefulness” of a metric. This method can be recycled in order to optimize the weights such that they produce the best-performing Player Rating.

 

Results

We think we’ve produced the best single-number representation of a player’s quality that currently exists publicly. Visit our Player Ratings page by clicking the button below and see what you think.

So yeah, Sobotka is one of the worst forwards out there. Probably doesn't belong on the 2nd line.

Link to comment
Share on other sites

1 hour ago, PASabreFan said:

Man, the consistency of the lines and pairings is so refreshing. It'll be interesting to see what Krueger does if the team goes on one of those patented Buffalo Sabres scoring droughts.

He will sit in a regal chair. He will then contemplate the situation.

Link to comment
Share on other sites

8 minutes ago, Wyldnwoody44 said:

I hope sobotka scores an ovechtrick just to silence the masses. 

Also, I miss conversations about hockey, just hockey, no F'n science, no stats and no other bs. Alas, those days are gone 

 

I'll take this current version of discourse over the awful stuff we were saying ten years ago on here, like Tim Connolly's brain damage being a punch line.

  • Like (+1) 1
Link to comment
Share on other sites

5 minutes ago, Wyldnwoody44 said:

I hope sobotka scores an ovechtrick just to silence the masses. 

Also, I miss conversations about hockey, just hockey, no F'n science, no stats and no other bs. Alas, those days are gone 

 

I enjoy a sprinkling of it.

Broken Ankle's line up chart with a sprinkle of stats is just a nice way to see how we stack up line vs. line when you don't know 90% of the other teams players because we just never see them. 

  • Like (+1) 1
Link to comment
Share on other sites

This topic is OLD. A NEW topic should be started unless there is a VERY SPECIFIC REASON to revive this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...