The curious case of Rasmus Ristolainen: ?

[Doohickie]

Ahem, from the Coller article: But if you don’t know when to disregard bad stats, you end up thinking plus-minus is a great measure of defense or batting average is the best baseball stat there is.

 

Example on the batting average thing:  Joey Gallo of the Texas Rangers, in 122 games so far this season, has a .211 batting average (pretty bad).  But he has 37 home runs (4th best in the league) and 72 RBIs (3rd best on the team).

 

Edited September 8, 2017 by Doohickie

[ubkev]

Joe DiMaggio hit .357 with 30 dingers, 125 RBI's and 13 strikeouts in 1941.

 

His OPS was 1.083

 

 

But enough baseball. Rasmus is good, I think. I think he continues to improve this year. I think he and Scandella will do well together.

Edited September 8, 2017 by ubkev

[Thorny]

^ Great stat.

[TrueBlueGED]

Hockey analytics don’t tell you if a player is good, they tell you if a situation is working. - best quote I've read in some time.

So you feel the same way for goals and assists? If not, why?

Absolutely, and from one of the pioneers in the field.

I think Coller is sorely missed at WGR, but let's not get carried away here. He's not even close to a pioneer in the field.

[dudacek]

So you feel the same way for goals and assists? If not, why?

.

I absolutely do.

 

I think generally hockey fans understand that goals and assists are useful but not a be-all-end-all measure for ability by the time they graduate middle school.

 

But there is a segment of the internet that hasn't learned the same about advanced numbers.

[TrueBlueGED]

I absolutely do.

 

I think generally hockey fans understand that goals and assists are useful but not a be-all-end-all measure for ability by the time they graduate middle school.

 

But there is a segment of the internet that hasn't learned the same about advanced numbers.

 

Way to blow up your "Eichel isn't worth $x because points" argument :nana: :p

 

Some writers have developed contextual methods which do a better job of accounting for it than many around here are aware of or would want to admit to. There's still a missing piece*, as there's a reason some analytics darlings can't make the jump to a tougher role (hi Pysyk) while others can (hi Leddy), but it's not like no attempts or progress have been made on this front. And let's be honest with ourselves, too: as Sabres fans we're going to be inherently defensive about the notion that Risto might only be an okay middle-pairing player, because such a prognosis leaves our blue line situation quite a bit more dire.

 

*My theory is most writers are making an underlying assumption of linearity--the "newest" thing is regression analyses, which are relying on ordinary least squares models and R^2 statistics, but I suspect hockey data violates the underlying assumptions of OLS in about 300 different ways. My best guess is there are significant non-linear interactive effects between skill/role/stats that have yet to be uncovered, or perhaps even investigated properly.

 

For instance, I think it has been satisfactorily shown that because of the fluid nature of hockey, matchup differential isn't as large empirically as the conventional wisdom holds. But I don't think the conclusion that "because this variance is small, competition doesn't matter much" is not one that should be so hastily drawn. Perhaps playing against tougher competition is so much more demanding that even smaller increases in those minutes cause an amplified decrease in performance against other, lesser players. Stuff like that hasn't been explored.

[dudacek]

You have what a player is supposed to do, plus what a player attempts to do, plus what a player actually does, plus what a player anticipates his teammate to do, plus what he anticipates his opponent to do. Times 12.

 

It's like analyzing the weather. You can, and there seems to be some really interesting work out there being misinterpreted by better minds than mine. But there are so many variables, good luck being consistently right, let alone controlling the outcomes.

[TrueBlueGED]

You have what a player is supposed to do, plus what a player attempts to do, plus what a player actually does, plus what a player anticipates his teammate to do, plus what he anticipates his opponent to do. Times 12.

 

It's like analyzing the weather. You can, and there seems to be some really interesting work out there being misinterpreted by better minds than mine. But there are so many variables, good luck being consistently right , let alone controlling the outcomes.

My biggest problem with pushback against analytics is this standard is never applied to the eye test. You, and plenty of others, approach the data with mindset inherently allowing you to dismiss things you disagree with.

 

Bill James had a great quote about statistical analysis: If you have a metric that never matches up with the eye test, it's probably wrong. And if it never surprises you, it's probably useless. But if four out of five time sit tells you what you know, and one out of five it surprises you, you might have something.

 

With hockey analytics, when a number falls in that 20% range, the tendency is to toss the baby out with the bath water. Meanwhile, the best GMs in the world are right far less than 80% of the time, and get hailed as geniuses. Huge double standard, especially when it comes to something like Ristolainen, where we and Sabres fans everywhere have a (quite understandable) vested emotional interest in the numbers being wrong.

[3putt]

I am trying to view analytics more neutrally, but I have yet to see where the are predictive. I often see a disdain for counting stats but ultimately that is how teams are judged, I.e.do you have more or less goals than the other team at the end of specified time. And as far as individuals in a team game, there appears to be "program" bias. Everyone raved about Babcock's usage of the young Leaves. But that discounts the fact that he essentially threw defense out the window. With that one would expect them all to have better possession and offensive driver numbers but the eye could see they weren't even trying to defend most times. So I think context is very important in viewing pure analytical data. The giving up the blue line in Risto's case was what I would call the organizational approach. But he also looked lost at times in his own end and the eye didn't see the creative puck movement of a Werenski or McAvoy.

[Randall Flagg]

I am trying to view analytics more neutrally, but I have yet to see where the are predictive. I often see a disdain for counting stats but ultimately that is how teams are judged, I.e.do you have more or less goals than the other team at the end of specified time. And as far as individuals in a team game, there appears to be "program" bias. Everyone raved about Babcock's usage of the young Leaves. But that discounts the fact that he essentially threw defense out the window. With that one would expect them all to have better possession and offensive driver numbers but the eye could see they weren't even trying to defend most times. So I think context is very important in viewing pure analytical data. The giving up the blue line in Risto's case was what I would call the organizational approach. But he also looked lost at times in his own end and the eye didn't see the creative puck movement of a Werenski or McAvoy.

The reason they are used is that they have a better successful prediction rate than any counting stat on the planet. If this wasn't the case, then people who are literally "stat geeks" would certainly not prefer them, that wouldn't make any sense at all. 

 

-----------------------------------------------------------------------------------------------------

 

There was a neat video posted here from a good article a while ago about Ristolainen giving up the blue line, and it gave clips of Nashville players doing the opposite. I'm interested to see if we implement this and how good Ristolainen becomes at it. For the amount of minutes he plays, Ristolainen did a surprisingly little amount of chasing from what I saw. He was often slow to read the play and predict where it was going next, and was occasionally caught out of position because of that, which in my eyes is his biggest obstacle in his path of becoming a legitimate #1D on a contender, but he rarely found himself chasing after a guy after losing the edge (c'mon, McDavid doesn't count).

 

This is because his gap control is very solid. Even as his role was to give up the line to protect the net he did a good job keeping his stick and shins in shooting and passing lanes (except on 2-on-1s). The plays that looked bad in this case almost always resulted, not in poor gap control, but inopportune receding towards the crease area, and even then the goals that happened were low percentage shots that should have been saved (think Tkachuk when we visited Calgary last year). So if Ristolainen is now coached up to be an Ekholm/Pietrangelo style YOU SHALL NOT PASS defenseman, I predict him becoming quite good at it. He's aggressive, he's a competitor, he wants to better himself always, and his gap intuition is much better naturally than it could be. The combination of these things make me think he'd flourish. 

 

So assuming this happens, we now have a defenseman who may not be the best down low at picking out where the play is going to go, but who is very good at snuffing out rushes and keeping these situations limited in the first place because of it, and who will hopefully have more support when acquiring the puck along the boards in a battle (which he's also at least average at) both from forwards in systemic improvements and from a smarter and more well-positioned defense partner in Scandella (who will be brought up in a second). He also has an average to above average breakout pass when it's not a two line pass, and he's also one of the best offensive point men in the game. Remember, he was a top 15 scorer among defensemen already at that age in that system (ahead of Doughty, Weber, OEL, Subban etc) with a career low shooting percentage. He will be elite offensively. His creativity in the offensive zone will hopefully no longer be for nothing. His straight-line skating is average but his edgework and lateral fluidity are above average. We will have a good skater who is elite offensively, good at snuffing out rushes, good at winning board battles, and a little slow to read long-development plays, but with a burning desire to win and to get better that is unteachable. I'm more than fine with this. 

 

I don't think his fancy stats will ever be great, but he is not the possession black hole that he appears to be, as Coller nicely summarized. 

 

Back to the blue line press, Nashville's top 4 D are all incredible skaters. Ristolainen will get burned occasionally by guys like McDavid, or closer to home, Kucherov, Nylander, Drouin. But Scandella has some beautiful long skating strides and (i've read) does a good job reading things defensively, and will in general be able to help Ristolainen recover when this happens. I hope we see this kind of aggressiveness because I really think it's an undeveloped secret stored deep inside Ristolainen's potential. Things like it are why I was so desperate to get a different coach mid-season. I didn't want any more time being wasted in his development doing things way over his head with ice time he should not have been getting. I was afraid that another season of that would have wrecked any chance of getting a top 10 defenseman in the game.

 

The chance is still there, and I guess Phil Housley, with his position and the organization which he spent his time in having developed Weber, Suter, Ellis, Ekholm, Josi etc. is as good a choice as any to work to get this out of him. Buffalo's history is rich with top-tier flashy forwards. It is very shallow in defensemen considered to be among the best in the world in their time. It is hard work, both on the part of the player and the organization, to get to that level. Unless you're a godly stud like Karlsson or a physical freak like Hedman, it is all about the right "do anything to get better and never be satisfied" attitude coupling with a development system for the player that is careful, meticulous, and geared towards making him elite. Getting those two things together is nearly impossible. Ristolainen is ready to go. He has brought everything he needs to to the table. He's held up his end. Hopefully we've now got the guy that can help him get to that level. 

[dudacek]

Good take on Risto re: protecting the blueline.

 

There is a player on the Sabres who was a disaster staying home and reading the play down low and protecting the house, who my feeble memory recalls being able to excel at stepping up at the blueline and closing off the rush at the boards. This guy sucks when waiting for the game to come to him, but has made some impressive plays moving forward and attacking the puck carrier.

 

I remember one of the best puck carriers in the league, some guy named Kane talking about how he hating playing against this guy.

I remember him staying with and harassing Sidney Crosby for a full turn of the D zone before finally knocking him off the puck.

 

Could this system change also mean redemption for Zach Bogosian?

Edited September 12, 2017 by dudacek

[qwksndmonster]

That game against Crosby in his first season here was heroic by Bogo.

[That Aud Smell]

I searched for this thread to revive it.

 

What the hell's going on with that guy?

 

I want him to settle down and be good, if only to stick Hamilton's words about him back down his jowled pie hole.

[darksabre]

He plays too casual. I think he got used to sitting back and logging 30 minutes a night playing that way, and now it's happening again because our defensive depth is hurting. He played 29 minutes last night. 

That's too much. He needs to play less and be more engaged when he is playing. 

 

[French Collection]

New coach with a new system, he will come around.

 

I think he will not be the stud we hoped for but a solid #2 or #3 D-man.

Scandella will help him feel less pressure, and help in the key situations.

Watching McCabe turn it around this early in the season shows what a jolt of confidence can do.

 

Risto needs to play his game, physical, move the puck quickly and use some creativity on offence.

Things will fall into place.

[LGR4GM]

He looks like he is thinking to much and not reacting fast enough. I hope this is a symptom of adjusting to the new style and him trying to mature as a player. I'll give him another 10 games. 

[That Aud Smell]

He plays too casual. I think he got used to sitting back and logging 30 minutes a night playing that way, and now it's happening again because our defensive depth is hurting. He played 29 minutes last night. 

 

That's too much. He needs to play less and be more engaged when he is playing. 

 

Holy Christ - hadn't realized that.

 

I'm not sure whether it's casual or indecisive, but the net effect is the same.

[bob_sauve28]

I remember from pre season Risto was saying he wanted to play with Scandella, which I thought was strange. Why would he announce that? But last night really showed how well those two play together. Scandella is just getting better and better. 

[dudacek]

His first pass has been off all year.

Don’t remember that being a huge issue before.

 

Hope it’s just an adjustment to when and where Phil wants him to move it.

 

 

Also, I expected to see him leading and joining the rush more than we have.

That actually goes for all of the D.

Edited October 25, 2017 by dudacek

[Eleven]

He plays too casual. I think he got used to sitting back and logging 30 minutes a night playing that way, and now it's happening again because our defensive depth is hurting. He played 29 minutes last night. 

 

That's too much. He needs to play less and be more engaged when he is playing. 

 

 

 

 

And 26 minutes the game before.  Too much indeed.

[sabills]

We need Bogo back. Him with McCabe (I think those sides line up, but I can never keep them straight) to push out Tennyson and give another solid line would help out that Top pair a lot.

[Randall Flagg]

His first pass has been off all year.

Don’t remember that being a huge issue before.

 

Hope it’s just an adjustment to when and where Phil wants him to move it.

 

 

Also, I expected to see him leading and joining the rush more than we have.

That actually goes for all of the D.

As well as his decision-making. He's pumping it off the boards no matter what, whether his team is tired and needs to change or is waiting for a rush together. 

 

Also, in the d-zone, he doesn't read too well. He's fast enough and smart enough that he isn't constantly out of position, but he will never dictate what happens with his brain the way an Erik Karlsson, or less loftily, a TJ Brodie does. He's reactive, he doesn't drive things.

[GASabresIUFAN]

Nice article & excellent illustration.

 

But it highlights why Bylsma was so darn frustrating & why he's currently out of a job. He KNEW what he needed to do to put his players in positiins to succeed (he talked a great game) but at the slightest sign of adversity stopped listening to his head & instead went w/ what that tightened sphincter suggested. "No, no, don't give Ristolainen reasonable ice time & role - make him into Gorges for 27 minutes / night. Don't give O'Reilly reasonable minutes working off a leg injury - skate him down to the cement. It may be easier on your players to control the puck, but that's also easier on the opposing D that are being chased down by speedsters like Moulson - dump it & make the D turn it back up before Matty reaches the face off dot."

 

:angry:

It’s funny that people can read the same article and come to a different conclusion. I read this as less an inditement of DD, but more about the incompetence of TM. DD had to use Risto as he did because he had no one else. The inditement of TM was his line about the trade for Kulikov. TM knew he needed someone like Scandella to play with Risto, but traded for the low IQ low skill Kulikov instead. This left DD to put Risto with Gorges again.

[nfreeman]

We need Bogo back. Him with McCabe (I think those sides line up, but I can never keep them straight) to push out Tennyson and give another solid line would help out that Top pair a lot.

 

...said no one on SabreSpace, ever.

 

(but I'd like to see him back and coached by Howie too.)

[sabills]

...said no one on SabreSpace, ever.

 

(but I'd like to see him back and coached by Howie too.)

Believe me, I surprised myself.