2015 Redraft - Provorov, Werenski or Hanifin?

[flyercanuck]

8 hours ago, hobie said:

 

I still believe in +/- as a simple guide to how well a player is doing on any team. It's probably no more reliable than any other stat.

 

Ya, because playing on the best team or the worst team in the league, regardless of how good YOU actually are, has no bearing on your +/-. 

 

Wait a minute....it has a lot to do with it.

 

In other words, the worst player on a team that scores 70 more goals than it allows could have a better +/- than the best player on the worst GF/GA team. Doesn't make him a better player.

[OccamsRazor]

10 hours ago, hobie said:

Who has the time to figure all this out?

 

I do...

 

..corsi, +/- i don't care what you use to me is the best Ivan is the best defenseman the Flyers have ever drafted. Period.

[elmatus]

11 hours ago, hobie said:

Perhaps but if I'm a NYR fan then my Rangers might be skewed better.

 

 

Bias is a thing for sure. That said, this board isn't exactly a gathering of the blind. You may notice we quite regularly hang players out to dry around here. We're not shy with the critique when its warranted. Some may even accuse us of being overly critical at times...

 

@brelic offered some good stats, if that's your thing. Others here will provide you with the eyeball test. Many of us do watch other teams besides the Flyers. I regularly watch whatever happens to be on if the O&B aren't playing. Having watched Columbus a good number of times this season, I would liken Provo's play to Jones more so than to Werenski. Similarly, Ghost and Werenski are better comparables imo.

 

Take that info for what it is. I do believe Provo is the best defenseman to suit up for the Flyers in some time. I also believe he is more rounded and capable overall than Werenski on the defensive side of the game. The stats @brelic mentioned provide basis for that as well. Does that eliminate bias completely? Of course not, but it does go a long way in answering what I feel was your original question.

 

[ruxpin]

If I had to do that draft over, it's   Provorov then Werenski then Hanifin.  OR Werenski then Provorov then Hanifin.   I really like Werenski.

 

Here's the thing.  As others have mentioned, Provorov so far is better defensively than Werenski in both eye test and stats.   It's admittedly harder to score defensive stats because there's more involved than "was he one of the last three players to touch the puck before it went into the opponent's net" on offense.

 

Provorov will likely improve on offense and is already doing well.   But Werenski is already doing well on defense, so who's to say he won't continue to improve there?    The Flyers are what they are on offense so there's some validity to the argument that his offensive numbers would go up on a stronger offensive team; however, while Werenski's slipped this year, you have to argue that's the result of the Jackets' slipping on offense all the way around this year.  I mean, Wennberg, Atkinson, and Panarin were not the strong line that Wennberg, Atkinson and Saad were.   And many of the forwards spent significant time on and off the IR.  So, I think that has to be kept in mind when viewing Werenski's numbers (Jones' numbers are up slightly, so maybe not the great argument I thought it was).

 

If I'm at that draft today, if I'm the Flyers I'm going Provorov because I already have Ghost.   Conversely, if I'm the Jackets, I'm taking Werenski because I already have Jones.   But if I'm team X and don't have a Ghost or a Jones?   I don't know, it's probably a coin toss or whichever I think goes with what I have.   Just on its face, I go Provorov because the offense is there, if not overwhelming, and the defense is better.   If I'm looking for offensive potential, I'm probably going Werenski.

[flyercanuck]

2 hours ago, ruxpin said:

If I had to do that draft over, it's   Provorov then Werenski then Hanifin.  OR Werenski then Provorov then Hanifin.   I really like Werenski.

 

Here's the thing.  As others have mentioned, Provorov so far is better defensively than Werenski in both eye test and stats.   It's admittedly harder to score defensive stats because there's more involved than "was he one of the last three players to touch the puck before it went into the opponent's net" on offense.

 

Provorov will likely improve on offense and is already doing well.   But Werenski is already doing well on defense, so who's to say he won't continue to improve there?    The Flyers are what they are on offense so there's some validity to the argument that his offensive numbers would go up on a stronger offensive team; however, while Werenski's slipped this year, you have to argue that's the result of the Jackets' slipping on offense all the way around this year.  I mean, Wennberg, Atkinson, and Panarin were not the strong line that Wennberg, Atkinson and Saad were.   And many of the forwards spent significant time on and off the IR.  So, I think that has to be kept in mind when viewing Werenski's numbers (Jones' numbers are up slightly, so maybe not the great argument I thought it was).

 

If I'm at that draft today, if I'm the Flyers I'm going Provorov because I already have Ghost.   Conversely, if I'm the Jackets, I'm taking Werenski because I already have Jones.   But if I'm team X and don't have a Ghost or a Jones?   I don't know, it's probably a coin toss or whichever I think goes with what I have.   Just on its face, I go Provorov because the offense is there, if not overwhelming, and the defense is better.   If I'm looking for offensive potential, I'm probably going Werenski.

 

 

Stop making sense.

[ruxpin]

15 minutes ago, flyercanuck said:

 

 

Stop making sense.

 

Sorry, I was trying something new.

[hobie]

10 hours ago, flyercanuck said:

 

Ya, because playing on the best team or the worst team in the league, regardless of how good YOU actually are, has no bearing on your +/-. 

 

Wait a minute....it has a lot to do with it.

 

In other words, the worst player on a team that scores 70 more goals than it allows could have a better +/- than the best player on the worst GF/GA team. Doesn't make him a better player.

 

And also the corsi and every other stat improves as you play with better players so then you compare results with other similarly positioned players on your team but can you compare a 3rd pairing d-man realistically with a 1st pairing d-man. 

 

I guarantee that if d-men are playing mostly with Boston's Bergeron line that any conceivable metric would be good to exceptional. Take those same d-men with TO's Bozak line with Marner mostly and every metric would tell a far worse story.

[flyercanuck]

55 minutes ago, hobie said:

 

And also the corsi and every other stat improves as you play with better players so then you compare results with other similarly positioned players on your team but can you compare a 3rd pairing d-man realistically with a 1st pairing d-man. 

 

I guarantee that if d-men are playing mostly with Boston's Bergeron line that any conceivable metric would be good to exceptional. Take those same d-men with TO's Bozak line with Marner mostly and every metric would tell a far worse story.

 

Right....which is why the eye test works.

 

If you want highlight plays, take Werenski. But there's a reason he doesn't play the PK, and Provorov is Phillys go to guy...he's better defensively.

[AJgoal]

If I could find an easily accessible source of with-or-without-you (WOWY) data, it would be simple to break it down. Just look at how other players fare when they are on the ice with each, vs without. Eliminates most of the noise by telling you if a particular player generally raises or lowers the effectiveness of others.

[hobie]

I would have taken Provorov over Werenski, I would also take Provorov over Marner or Werenski even now. I simply, sort of, wanted to know why people were saying that Provorov is better than Werenski. You feel Provorov is better defensively and he probably is I don't know that for sure but what I like is that Provorov is supposedly more physical.

 

However at 6' is being more physical an impediment to longevity? They're both 6'. 

[hobie]

12 minutes ago, AJgoal said:

If I could find an easily accessible source of with-or-without-you (WOWY) data, it would be simple to break it down. Just look at how other players fare when they are on the ice with each, vs without. Eliminates most of the noise by telling you if a particular player generally raises or lowers the effectiveness of others.

 

 I read an interview of the corsi originator which was at the beginning of the fancy stat movement. He said at the beginning he thought he was onto something but by the end he felt that what he started was too limited, essentially  a different version of +/- which he wanted to replace.

 

I can't find the interview, I think it was on Hot Stove.

 

 

[flyercanuck]

20 minutes ago, hobie said:

I would have taken Provorov over Werenski, I would also take Provorov over Marner or Werenski even now. I simply, sort of, wanted to know why people were saying that Provorov is better than Werenski. You feel Provorov is better defensively and he probably is I don't know that for sure but what I like is that Provorov is supposedly more physical.

 

However at 6' is being more physical an impediment to longevity? They're both 6'. 

 

Provorov isn't a beast in front of his net...but he's smart and will use his body. His style of play should give him a long career.

[AJgoal]

4 minutes ago, hobie said:

 

 I read an interview of the corsi originator which was at the beginning of the fancy stat movement. He said at the beginning he thought he was onto something but by the end he felt that what he started was too limited, essentially  a different version of +/- which he wanted to replace.

 

I can't find the interview, I think it was on Hot Stove.

 

 

 

It is, because it doesn't measure shot quality. But there are other ways to break that stuff out. There are now measurements for "high danger," "medium danger," and "low danger" chances based on where the shots are taken. There's expected goals for. And you can take all of these things and combine them to make a picture. I think a lot of people look at one stat (often Corsi) and think that that alone paints the picture, but I don't think it's ever good to look at it from one perspective only.

[ruxpin]

15 minutes ago, hobie said:

I would have taken Provorov over Werenski, I would also take Provorov over Marner or Werenski even now. I simply, sort of, wanted to know why people were saying that Provorov is better than Werenski. You feel Provorov is better defensively and he probably is I don't know that for sure but what I like is that Provorov is supposedly more physical.

 

However at 6' is being more physical an impediment to longevity? They're both 6'. 

 

I would have been extremely happy with either Provorov or Werenski.   I really like Werenski's game.  I think it was you that said Werenski was kind of the odd-man out at the time of the draft between the three of them.  I think I agree with that, but I really like where Werenski has taken his game.   I'm not sure why he hasn't produced as much this year (do you get much eyeball on him?) but I want to blame it on the injury situation in Columbus.   Maybe Tortorella (not a huge fan)?   

 

I'm not sure I would classify Provorov or Werenski as being particularly physical.   Provorov, I guess, in the sense that he positions himself well and gets "in the way" but not really in the Gudas mode or traditional "Bullies" mode.   He's kind of a smooth skater and positions himself well.   When he does the young player thing and gets himself out of position, he has a hard time physically forcing his way back in (if that makes sense).    I think the two players are fairly similar in that regard.    Provorov is actually a very fluid skater although at first blush he looks a little odd doing it.   He hold his arm on his top hand oddly high.  He almost looks like one of the players on one of the old Bobby Hull hockey games the way he holds his top shoulder and elbow.   But, it's effective for him and doesn't seem to hinder him in any way, so who's to criticize?   It just strikes you odd the first couple times you see him.

 

Others seemed to want to compare Werenski to Ghost instead.   I'm not sure I like that comparison.  Ultimately, I think Werenski will have better offensive numbers than Provorov but probably not to Ghost level.   All that to say that I think he and Provorov are probably the better comparison (although I get the Jones/Provorov comparison better than the Ghost/Werenski).   I think Werenski is one heck of a player who's offensive production could have been hampered this year due to lineup/injury issues and possible a coach who was trying to reign in a bit.   I'm not sure; I don't see enough of Columbus.

 

Werenski may end up being the better full package unless Provorov has another offensive gear.   I think Werenski is good but not great on the back end but could still improve.  I think Provorov actually can improve yet on the backend, but on the front end I'm not convinced he'll be as good as Werenski's ceiling (thus, I think, the desire to compare Werenski to Ghost instead).   

 

All this to say that at this point Werenski > Provorov on offense and Provorov > Werenski on the back end.   They're both young so the difference on both ends may change as I don't think either one has hit their ceiling.   I also think both happen to be exactly what they're drafting teams needed.

 

How's that for a long post that really didn't draw any conclusions?

[ruxpin]

7 minutes ago, AJgoal said:

 

It is, because it doesn't measure shot quality. But there are other ways to break that stuff out. There are now measurements for "high danger," "medium danger," and "low danger" chances based on where the shots are taken. There's expected goals for. And you can take all of these things and combine them to make a picture. I think a lot of people look at one stat (often Corsi) and think that that alone paints the picture, but I don't think it's ever good to look at it from one perspective only.

 

I think I agree with that.   But you've always been better with secondary stats (Corsi included) than me.

[hobie]

2 minutes ago, AJgoal said:

 

It is, because it doesn't measure shot quality. But there are other ways to break that stuff out. There are now measurements for "high danger," "medium danger," and "low danger" chances based on where the shots are taken. There's expected goals for. And you can take all of these things and combine them to make a picture. I think a lot of people look at one stat (often Corsi) and think that that alone paints the picture, but I don't think it's ever good to look at it from one perspective only.

 

So what the fancy stat movement is ending up with is so many variables needing to be covered that anyone wanting to draw an unassailable conclusion may need to utilize what 20, 30 or more source references.

 

+/- isn't perfect but it's simple, now, and pretty well tamper proof.  

[AJgoal]

Note also, that I wasn't necessarily talking about "fancy stats." 

 

There are with-or-without-you charts for:

 

-Goals For

-Goals against

 

otherwise known as the two halves of +/-.

 

Say you want to figure out if Player A's stats are the cause of his +/-, or the result of Playing with Player B. You look at how their stats fare when playing with Player C. 

 

Do both Player B and Player C's stats go up with Player A? then it's a good bet he makes the players around him better. Do Player A's stats go up with Player C, but down with Player B? Then there's a chance that Player B is pulling him down - but it also could be Player C picking him up. You make the comparison across multiple partners to come to a reasonable conclusion.

 

The big thing about +/- in relation to Corsi is that +/- doesn't adjust for quality of goaltending. Corsi does, since goals aren't an actual factor.

[ruxpin]

2 minutes ago, AJgoal said:

The big thing about +/- in relation to Corsi is that +/- doesn't adjust for quality of goaltending. Corsi does, since goals aren't an actual factor.

 

Could you do a Cliff's Notes explanation of this?    From the brain dead bear comes a really stupid question:   if goals aren't an actual factor and the game is about scoring and defending against goals, what is Corsi measuring in that regard?   (Clearly I know whatever is just below "nothing" about Corsi!).  

[brelic]

6 minutes ago, ruxpin said:

 

Could you do a Cliff's Notes explanation of this?    From the brain dead bear comes a really stupid question:   if goals aren't an actual factor and the game is about scoring and defending against goals, what is Corsi measuring in that regard?   (Clearly I know whatever is just below "nothing" about Corsi!).  

 

It measures the event intrinsically linked to scoring - shooting the puck. 

 

From wikipedia - Corsi is an advanced statistic used in the National Hockey League to measure shot attempt differential while at even strength play. This includes shots on goal, missed shots on goal, and blocked shot attempts towards the opposition's net minus the same shot attempts directed at your own team's net.

 

So, the theory is that if you have a positive Corsi (above 50), there’s a good chance your team is controlling play when you are on the ice because you are directing more shots at net than the other team... and more likely to score goals and win humongous games. 

[hobie]

4 minutes ago, AJgoal said:

Note also, that I wasn't necessarily talking about "fancy stats." 

 

There are with-or-without-you charts for:

 

-Goals For

-Goals against

 

otherwise known as the two halves of +/-.

 

Say you want to figure out if Player A's stats are the cause of his +/-, or the result of Playing with Player B. You look at how their stats fare when playing with Player C. 

 

Do both Player B and Player C's stats go up with Player A? then it's a good bet he makes the players around him better. Do Player A's stats go up with Player C, but down with Player B? Then there's a chance that Player B is pulling him down - but it also could be Player C picking him up. You make the comparison across multiple partners to come to a reasonable conclusion.

 

The big thing about +/- in relation to Corsi is that +/- doesn't adjust for quality of goaltending. Corsi does, since goals aren't an actual factor.

 

I know very little about any variation outside of +/-  but I do know a little about WOWY  and when I did the little research I did I found it lacking because as I think that all six players need to be considered in every equation. Which RW is supporting the RD, which center, every team has 2 goalies, the variables are endless.

 

I discounted corsi right away as a true predictor of players/teams results. Many teams have shown that volumes of shot attempts and even shots for and against aren't totally reliable. TO last year and this constantly gets out shot by some stunning amounts yet continues to win, talent trumps any measurable metric.

 

TO uses Hyman on it's 1st line with Matthews and Nylander, most if not all measurable results show that that line would get superior results if they used someone of better talent on that line. The eye test supports that yet Babs constantly praises Hyman and continues to use him on the 1st line.

 

I'm not a Philly fan so I'll tell you the story of Gardiner and Rielly last year. Gardiner thru the 1st half was a +25 to 30 while playing on the second pairing. Rielly was -25 or there abouts playing on the 1st pairing. Gardiner will not pass anyone's eye test but Rielly looks good even when he's bad. Babs finally switched them and both of their results evened out where both were helping to provide TO better than acceptable results. 

 

For me +/- means you provide value or not when on the ice in countless situations.  If you're doing exceptionally well then maybe you can shoulder more responsibility, if you're not doing so well a demotion might be in order.

[AJgoal]

2 hours ago, ruxpin said:

 

Could you do a Cliff's Notes explanation of this?    From the brain dead bear comes a really stupid question:   if goals aren't an actual factor and the game is about scoring and defending against goals, what is Corsi measuring in that regard?   (Clearly I know whatever is just below "nothing" about Corsi!).  

 

+/- is a measure of goals, those scored for and against. Corsi is a measure of shot attempts, for and against (all shot attempts, including blocks and misses which don't show up on the score sheet). Since it measures all attempts, regardless of outcome, goaltending (and shooting efficiency) doesn't factor in.  

 

Now, it also doesn't correct for strength of attempt. So a shot attempt made from a bad angle that is blocked counts the same as a goal. 

 

The bold is a fair question, and the answer is: it doesn't, really. What it is most useful as a measure of is puck possession. If a player's Corsi percentage is above 50 (ie shot attempts for exceeds shot attempts against), that is generally interpreted to mean that they're possessing the puck more than their opponent, and therefore "driving play." So even if you're not scoring, you're minimizing your opponent's ability to score. The flip side is also generally true.

 

Again, in the end, these are all tools. No one statistic ever tells the whole story in a complex system - we don't do that in science, and it certainly can't be done in an environment as fluid as a hockey game. 

[ruxpin]

7 minutes ago, AJgoal said:

The bold is a fair question, and the answer is: it doesn't, really. What it is most useful as a measure of is puck possession. If a player's Corsi percentage is above 50 (ie shot attempts for exceeds shot attempts against), that is generally interpreted to mean that they're possessing the puck more than their opponent, and therefore "driving play." So even if you're not scoring, you're minimizing your opponent's ability to score. The flip side is also generally true.

 

Thank you for this.

 

8 minutes ago, AJgoal said:

Again, in the end, these are all tools. No one statistic ever tells the whole story in a complex system - we don't do that in science, and it certainly can't be done in an environment as fluid as a hockey game.

 

Quite true!   It's certainly not as easy as baseball (a lot more moving parts).  I realize stats don't tell the whole picture in baseball either, but it's a much (much!) bigger part of the picture.

[AJgoal]

5 minutes ago, ruxpin said:

 

Thank you for this.

 

 

You're welcome. I hope I'm doing this justice... I'm actually not a huge stat nerd, but I do like to see if the numbers back up what I think I'm seeing on the ice. If you want a more in-depth look at things, check this out:

 

https://theathletic.com/121980/2017/10/09/an-advanced-stat-primer-understanding-basic-hockey-metrics/

 

 

[OccamsRazor]

What a season by this kid only his 2nd. Just turned 21 in january.

 

Leads the NHL in goal (17) for defensemen. 15 at even strength. Helluva year!!!

 

Not bad for a two way defensemen.

 

 

 

And then to play like the stud he is in the biggest game of the year!!! Nice defensive play too.

 

 

Kid is only scratching the surface.

 

I think he has to be in the Norris running!!! 41 points as the #1 Dman going against the other teams best.

 

24 minutes a night and still finished +17!!!!!!!

 

I take this kid every time in a redraft!!!!!!!