Summer '15 Thread (All Proposals/Blog Rumors in here)

[cslebn]

So post TDL I thought it would be good to use some analytics to take a look at where our team stands. Source for this data is http://ownthepuck.blogspot.ca/.


Notes on the data:
The data is sourced form the 2012-13 season till 2/19/15.
This data since it accounts for multiple years does not account for massive improvement or players such as Cody Goloubef. That can also impact a player such as David Savard or Kevin Connauton (who has a similar Corsi QoC per Behindthenet as Dalton Prout).

See attachment for data (not able to get it loaded to a website from work). Also, I really didn't want to post HERO charts for everyone so sorry about the data miner type xls snap shot.

This data shows that we've got a lot of second liners on the tweam for forwards with our best forwards being Foligno, Hartnell, and Dubi. Joey shows there really is room for improvement. The data also shows where linemates have really hurt Letestu.


Our defense is very poor looking in this, especially JJ and Savard. Savard clocks in as the worst D man on the team. That may be skewed by the years used to pull this data but like JJ you have to suspect this is a tools and toolbox issue now. I like JJ and this seems to be an indicator that he really isn't the player we hope. Instead he's the guy that is always just a little away from scoring or making the stellar play.


The analysis also shows how devoid of truly top end talent we are. That said 6 guys rate as 2nd liners. 5 guys as 3rd liners. 1 as a 4th and 1 didn't rate (sorry Boll). We also have Morin, Wennberg and Dano that don't chart based on GP. And I didn't use Bourque since he's in Springfield. That's good depth. With more time that should be compared with say the Bruins or LA.

This makes the goals this summer in my mind:
- Replace JJ with someone that can handle the minutes or find another D man that can eat some of his minutes; and Savard is not the answer.
- add 4th line depth (or just run better data on the guys who didn't have the GP to make this cut).
- Resign Calvert!
- kick Jarmo in the nuts for trading the best D man on the team per analytics
- look for another RW but I don't think this is anywhere near as important as the D and Calvert.

Click to read more...

 

[Mayor Bee]

Credit for compiling all of this into one spot.

Unfortunately, this is one of the major pitfalls of analytics in sports, and especially in hockey. Sabermetrics in baseball is tremendous for telling us what happened; it allows us to "see" that Darrell Evans was underrated and that Jim Rice was overrated, that the HOF is hideously overrepresented in certain time frames because of a skewed statistical picture due to the conditions of the game, that this pitcher's record declined sharply due to reasons related more to the makeup of the defense behind him than to an actual loss of skill on his part.

But as a predictive model, there's a big breakdown. Just in baseball, I've done well by doing nothing more than using BABIP as a baseline for predicting pitchers' ability to keep the ERA down and win more games, which is about the only thing that matters in fantasy baseball. Hockey is so limited in what is tracked beyond the most basic numbers that getting a picture of what happened is difficult, making any type of predictions of future performance and outcomes damn near impossible.

I love numbers, I love analysis, I love history, but I don't like what currently passes for "advanced stats" in hockey. I recognize that there are inherent limitations in what can be tracked and logged, so I'm not blaming that, but we've gone from drawing water using a thimble to using a pint glass...and it's being drawn from a lake.

 

[EspenK]

I don't fully appreciate (read "don't like") analytics. The way I understand it a guy who shoots 300 times a year but never scores is better than a guy who shoots 100 times and pots 25. I'll take Player B thank you.

Anyway for you analytics guys I offer this up when re-building the team.
Assuming Calvert is re-signed for 2 mill and Letestu walks and Bourque is bought out there is a cap hit of 64.3 mill for a roster made up of 14 signed F's, 7 signed D, Bob and a back-up at 1 mill. About 6 mill or so in added salary to take one if trades are made or if a UFA is signed.

 

[Mayor Bee]

I don't fully appreciate (read "don't like") analytics. The way I understand it a guy who shoots 300 times a year but never scores is better than a guy who shoots 100 times and pots 25. I'll take Player B thank you.

The purpose of the most popular advanced stats (Corsi and Fenwick) is basically to track possession of the puck via shots, attempted shots, etc.

As CBJ fans, we should all be able to pretty quickly spot the flaw in this: we all went through the Scott Arniel experiment. If you look at Corsi from 2010-11, it's staggering...you'd look at Columbus and think that this was a dominant team. Every single player was a positive except for Methot, Dorsett, Pahlsson, Clark, and Moreau. The highest on the team were Huselius and Upshall, which leads us to the next obvious flaw...

On the year, Columbus was below .500, ranked 25th in goals scored, and allowed the 5th-most goals in the entire league. Every single game was seemingly a shooting gallery for the opposing goalie, but it turns out that having the 5th-worst shooting percentage and 2nd-worst save percentage means you won't win many games.

Like I said, I love numbers and sports and analyzing and history. I recognize that Corsi and Fenwick are supposed to be used more for supplemental purposes rather than the be-all-end-all. But here's what it ultimately comes down to: the correlation between shots and shots attempted and actually winning games is not that strong, and people mistake correlation for causation anyway.

 

[Robert]

The purpose of the most popular advanced stats (Corsi and Fenwick) is basically to track possession of the puck via shots, attempted shots, etc.

As CBJ fans, we should all be able to pretty quickly spot the flaw in this: we all went through the Scott Arniel experiment. If you look at Corsi from 2010-11, it's staggering...you'd look at Columbus and think that this was a dominant team. Every single player was a positive except for Methot, Dorsett, Pahlsson, Clark, and Moreau. The highest on the team were Huselius and Upshall, which leads us to the next obvious flaw...

On the year, Columbus was below .500, ranked 25th in goals scored, and allowed the 5th-most goals in the entire league. Every single game was seemingly a shooting gallery for the opposing goalie, but it turns out that having the 5th-worst shooting percentage and 2nd-worst save percentage means you won't win many games.

Like I said, I love numbers and sports and analyzing and history. I recognize that Corsi and Fenwick are supposed to be used more for supplemental purposes rather than the be-all-end-all. But here's what it ultimately comes down to: the correlation between shots and shots attempted and actually winning games is not that strong, and people mistake correlation for causation anyway.

No stat is pure..... blocked shots as well as shots that hit the post do count yet are not recorded in traditional stats..

Advanced stats are additional tools, just as one has several screw drivers in their tool box, it takes several tools to get the job done...

 

[cslebn]

Eh just another way to judge players.

How many of the guys do you really disagree with the way the stats show them to be?

 

[Robert]

Eh just another way to judge players.

How many of the guys do you really disagree with the way the stats show them to be?

This is the problem... the NHL as well as the NBA focus on "player" as much or more than "team".... Corsi and advanced stats focus on one thing... TEAM...

 

[JacketsDavid]

I'll admit I'm a nerd, but pay little attention to hockey stats.
But I follow baseball and the metrics have really changed things. Again it can't measure everything (will to win, determination, etc) but it's trying to break down what measurably makes a successful player.

Quite frankly I'm amazed we have 6 top 6 forwards. I would have assumed about 4, with a bunch of 3rd line forwards and another bunch of 4th line forwards.

That being said if I was rebuilding the team:
1. We have the right goaltender. That is where I think we start. But at some point need to solidify the backup because Bob (as much as I love him) always gets knicked up and just the normal rest a goalie needs. So need a guy we're comfortable with for 25-30 regular season starts. mc may be that guy but we'll see if the team pays him.
2. Build from the net out. Figure out a way to solidify the blue line - likely means trading some valuable forwards for d-men. Focus on preventing quality shots/goals.
3. Hope the young forwards mature and progress. For now need to largely focus on short-term contracts with solid 2 way players, with whatever talent we have (and again 'm assuming to fix the blue line we'd have to give up some assets).

Assuming the young forwards progress the scoring would hopefully come (and let's be honest we've never been an offensive juggernaut - closest in my mind was Nash rookie's year) but it will be a few (2-3 seasons) down the road.

But that would be my plan.

 

[major major]

Quite frankly I'm amazed we have 6 top 6 forwards. I would have assumed about 4, with a bunch of 3rd line forwards and another bunch of 4th line forwards.

I've been saying it for the longest time: We don't have a shortage of top-6 forwards, we have a shortage of top-line forwards. And now with the emergence of Foligno and Johansen we're only short one top line forward, but with the way Hartnell is playing it's not an immediate problem.

The much bigger issue is top-pair D. JJ and Savard would make a fine second-pair, but they can't cut it as a top-pair. On the left-side, perhaps Murray recovers and develops, or perhaps we draft a guy like Hanifin who can play top-pair soon. On the right-side it's hard to see how there's any alternative to making a trade.

 

[Viqsi]

Credit for compiling all of this into one spot.

Unfortunately, this is one of the major pitfalls of analytics in sports, and especially in hockey. Sabremetrics in baseball is tremendous for telling us what happened; it allows us to "see" that Darrell Evans was underrated and that Jim Rice was overrated, that the HOF is hideously overrepresented in certain time frames because of a skewed statistical picture due to the conditions of the game, that this pitcher's record declined sharply due to reasons related more to the makeup of the defense behind him than to an actual loss of skill on his part.

But as a predictive model, there's a big breakdown. Just in baseball, I've done well by doing nothing more than using BABIP as a baseline for predicting pitchers' ability to keep the ERA down and win more games, which is about the only thing that matters in fantasy baseball. Hockey is so limited in what is tracked beyond the most basic numbers that getting a picture of what happened is difficult, making any type of predictions of future performance and outcomes damn near impossible.

I love numbers, I love analysis, I love history, but I don't like what currently passes for "advanced stats" in hockey. I recognize that there are inherent limitations in what can be tracked and logged, so I'm not blaming that, but we've gone from drawing water using a thimble to using a pint glass...and it's being drawn from a lake.

I want to reply with the usual "well, more information is always a good thing"... except that my PI's pet project/life's work is about developing an objective measurement of evidence and its value, and that's brought about by the fact that current methods for approximating same (the apparent abuse of P-values) have so many flaws that in many cases adding more data doesn't help.

 

[Double-Shift Lasse]

http://bluejacketsxtra.dispatch.com/content/stories/2015/03/06/blue-jackets-feature.html

But Kekalainen did agree that are some difficult and delicate decisions to be made in the months ahead. The status quo cannot be maintained, if only because so many players are pushing upward from within the organization. The Blue Jackets listened to trade offers for center Artem Anisimov and right wing Cam Atkinson, but none of the offers were deemed worthy of deal. Yet.

The Blue Jackets will keep listening to offers, and they likely are to trade a forward or two this summer to ease a glut. With Anisimov, Atkinson, Matt Calvert, David Clarkson, Brandon Dubinsky, Nick Foligno, Scott Hartnell, Boone Jenner and Ryan Johansen, the Blue Jackets’ already have their top three lines set. (Calvert is a restricted free agent.)

That doesn’t include rookie Alexander Wennberg, who has been thrust into an NHL role this season. Three other fourth-line types — Jared Boll, Corey Tropp and Jeremy Morin — have contracts that extend beyond this season.

That’s 13 forwards.

And that’s not to mention rookies Josh Anderson, Marko Dano and Kerby Rychel, all of whom are on the cusp of NHL-readiness. Dano is with the club now, and the other two did not look out place in Columbus earlier this season.

It also doesn’t include the two forwards the Jackets received in the Wisniewski trade: prospect center William Karlsson, whom Kekalainen raves about, and NHL veteran Rene Bourque, whose career has unraveled after he scored 75 goals in a three-year stretch from 2008 to ’11.

There also are unrestricted free agents Mark Letestu, Jack Skille and Brian Gibbons, and restricted free agent Michael Chaput. And it leaves no room for top prospects Oliver Bjorkstrand and Sonny Milano, both of whom are expected to turn pro next season. Bjorkstrand has 48 goals in 49 games in the Western Hockey League this season.

 

[BluejacketNut]

There shouldnt be any room for guys like Bjorkstrand, Milano and such, let them fill in for injuries and develop in the AHL for a couple years and then put them in the lineup when they're NHL ready.

 

[ndd17]

If we are going to play next season with current roster help us make PO can only one decision from JD and JK. Fire Richards as quickly as possible from the beginning of the season. Otherwise, in the November/December will be too late, another lost season. And God forbid us from Hartsburg, more better see Larsen as HC. But my dream is to see Bylsma.

 

[Milltec]

There shouldnt be any room for guys like Bjorkstrand, Milano and such, let them fill in for injuries and develop in the AHL for a couple years and then put them in the lineup when they're NHL ready.

agreed

If we are going to play next season with current roster help us make PO can only one decision from JD and JK. Fire Richards as quickly as possible from the beginning of the season. Otherwise, in the November/December will be too late, another lost season. And God forbid us from Hartsburg, more better see Larsen as HC. But my dream is to see Bylsma.

agreed

I am hoping for a complete overhaul on defense with Murray, Goloubef, and Savard being retained and everyone else being expendable (on defense). Also please find a way to get rid of Boll, and Tropp.

 

[Cyclones Rock]

Corsi and Fenwick are just proxies for puck possession. The assumptions are that shots require possession of the puck which can't be disputed. However, it gives equal value to all shots/attempted shots, which may or may not be true even over a large sample. They don't actually measure exact time with the puck. I've read that it correlates very strongly, but I'm sure that there are some significant variances. I'd guess that a guy like Adam Oates would have had a poor individual Corsi/Fenwick even though he held the puck a lot because his pass first style didn't generate many shots for himself. I guess it's too much work to actually measure real time.

Many race tracks now place chips in the saddle gear of race horses. Exact distances run are tracked. I'll guess that NHL players will have some sort of tracking system installed in their equipment and in pucks which will give an exact measure of time with puck-which is actual possession.

Another issue that I've always had with Corsi/Fenwick is that the great Russian puck possession teams of the 70s and 80s took very few shots yet actually possessed the puck a great deal of the time....at least by my eyes. These teams were the text book examples of "puck possession" yet I'd guess that they'd fair poorly on Corsi or Fenwick which supposedly measure possession.

Until actual puck time is tracked scientifically, Corsi and Fenwick are of limited value, imo. They may be ok in terms of generalities, but getting too specific about their values is probably a mistake.

 

[DarkandStormy]

I wouldn't be surprised to see a couple forward plus perhaps one of the many picks we have dealt in a package for a top line defenseman. Johnson/Savard just won't cut it as the top pair.

 

[CraigsList]

Guys, I just have a simple question. If you were able to get Seth Jones from the Predators with your 2015 1st, would you do it? It'd give you a RD again after the loss of Wisniewski. Just a trade I thought of in my head.

 

[cslebn]

Guys, I just have a simple question. If you were able to get Seth Jones from the Predators with your 2015 1st, would you do it? It'd give you a RD again after the loss of Wisniewski. Just a trade I thought of in my head.

I'd have to think long and hard and the answer may very well be yes (top 2 protected of course).

 

[EspenK]

I'd have to think long and hard and the answer may very well be yes (top 2 protected of course).

Unless there is some glaring weakness in his game that has been uncovered I would make the deal.With the same caveat above

 

[CraigsList]

I'd have to think long and hard and the answer may very well be yes (top 2 protected of course).

Unless there is some glaring weakness in his game that has been uncovered I would make the deal.With the same caveat above

Imo, Jones will probably never show his real top pairing potential because he'll always be behind Weber. If your pick is 4th/5th/6th-ish, I think it's a good idea. It even benefits the Preds since they are looking for their #1C.

 

[major major]

There shouldnt be any room for guys like Bjorkstrand, Milano and such, let them fill in for injuries and develop in the AHL for a couple years and then put them in the lineup when they're NHL ready.

I'd put it even stronger: Bjorkstrand, Milano, Wennberg, and Rychel are only to get long stints in the lineup when they steal a spot. Dano has already stolen a spot IMO.

 

[EspenK]

I'd put it even stronger: Bjorkstrand, Milano, Wennberg, and Rychel are only to get long stints in the lineup when they steal a spot. Dano has already stolen a spot IMO.

So you think Wennberg hasn't been "rosterized"? I think he is here for the duration unless he really screws up. Not saying he deserves it but it just looks like he is a fixture at this point.

 

[Cowumbus]

I think Anisimov gets traded for top 4 d unless we want to just have 4 solid lines. Something's got to give

 

[Milltec]

Guys, I just have a simple question. If you were able to get Seth Jones from the Predators with your 2015 1st, would you do it? It'd give you a RD again after the loss of Wisniewski. Just a trade I thought of in my head.

Unless we get top 2 pick, absolutely. if we are at 3 or greater im making that trade every time.

 

[CBJx614]

So you think Wennberg hasn't been "rosterized"? I think he is here for the duration unless he really screws up. Not saying he deserves it but it just looks like he is a fixture at this point.

Agreed, they wouldn't have him as the PP QB if he was just up here on an injury stint. He and Dano are on this team next season to start unless they play themselves out of a spot in training camp and preseason. Especially now that he's putting up points