Winnipeg Jets Prospects Thread 2016-17 Part III

[Nolan Giesbrecht]

Or that it already taught them their lesson, which I believe is the actual case.

So it was either right before or right after the draft that I heard Chevy say something in an interview that made me wonder this exact thing.

It wasn't blatant, but just something that made me stop and take notice.

But that's all that I remember and I have no idea what was actually said or if their was any validity to my inference.

I am on this site every day, mostly lurking, but I don't remember any one else mentioning this, so it's likely my imagination.

 

[Grind]

The search for the next Milan Lucic!!
The search for the next Zdeno Chara!!

lol

This pretty much sums up every bad draft pick since 2011

 

[Whileee]

This pretty much sums up every bad draft pick since 2011

Well, there were also probably some bad draft picks trying to find the next Theoren Fleury or Ray Whitney, too.

 

[Joe Hallenback]

Well, there were also probably some bad draft picks trying to find the next Theoren Fleury or Ray Whitney, too.

Hey it was all the coaches fault for not using those guys properly

 

[Grind]

Well, there were also probably some bad draft picks trying to find the next Theoren Fleury or Ray Whitney, too.

The difference is teams undervalue the next Fleury and tend to take that guy outside of the first two rounds, and overvalue the next lucic and take him in the top two.

Don't get me wrong, teams shouldn't be taking Nick petan at like 15 overall, but they usually don't. They shouldn't be taking Tom Wilson at 15 overall but they often are.

 

[lomiller1]

The difference is teams undervalue the next Fleury and tend to take that guy outside of the first two rounds, and overvalue the next lucic and take him in the top two.

Don't get me wrong, teams shouldn't be taking Nick petan at like 15 overall, but they usually don't. They shouldn't be taking Tom Wilson at 15 overall but they often are.

How hard it is to get a particular player type matters too. Small 40 point forwards without much of a 2-way game are not a hard thing to acquire in trade or free agency. At the draft It makes sense to Prioritize things that are not so easy to acquire.

 

[Grind]

How hard it is to get a particular player type matters too. Small 40 point forwards without much of a 2-way game are not a hard thing to acquire in trade or free agency. At the draft It makes sense to Prioritize things that are not so easy to acquire.

Not if they don't turn out ever and aren't effective.


The effectiveness of small 40 point scorers with no two way game is vastly underrated, and the opposite is true of "the next lucic".

No matter how you cut it, swinging for "the next lucic" when there's almost always better players who are more likely to pan out I'd always a bad idea.

 

[Whileee]

Not if they don't turn out ever and aren't effective.


The effectiveness of small 40 point scorers with no two way game is vastly underrated, and the opposite is true of "the next lucic".

No matter how you cut it, swinging for "the next lucic" when there's almost always better players who are more likely to pan out I'd always a bad idea.

Sure. Understood.

But if you take size out of your predictive models, does the accuracy and reliability of the models increase or decrease?

Size is overvalued, but it's still a very relevant variable in prospect evaluation and projection.

 

[Grind]

Sure. Understood.

But if you take size out of your predictive models, does the accuracy and reliability of the models increase or decrease?

Size is overvalued, but it's still a very relevant variable in prospect evaluation and projection.

Size is definitely an indicator and the models do worse without it. But like you said, it is over valued

 

[garret9]

Sure. Understood.

But if you take size out of your predictive models, does the accuracy and reliability of the models increase or decrease?

Size is overvalued, but it's still a very relevant variable in prospect evaluation and projection.

Size is definitely an indicator and the models do worse without it. But like you said, it is over valued

Also, halo effect.

Does size predict NHLers because of value or bias inherent in decision makers or both.

 

[Grind]

Also, halo effect.

Does size predict NHLers because of value or bias inherent in decision makers or both.

That was my follow up as well.

Not sure how you'd take it to the next level. Comparing impact of height on a metric like GAR or WAR?

No clue.

 

[garret9]

That was my follow up as well.

Not sure how you'd take it to the next level. Comparing impact of height on a metric like GAR or WAR?

No clue.

I previously showed already that size only a significant variable to making the NHL, not in separating the quality of those that had.

 

[Mortimer Snerd]

Sure. Understood.

But if you take size out of your predictive models, does the accuracy and reliability of the models increase or decrease?

Size is overvalued, but it's still a very relevant variable in prospect evaluation and projection.

The mistake is in thinking that size is an addition to how good a player is. Player X scored a ppg in the CHL - AND he is 6'5 is where it goes wrong. Player X scored a ppg in the CHL BECAUSE he is 6'5.

The models that take height into account only do a little better and that is because being big gives a little better chance of transitioning to the next level where the avg player is a little larger.

Player Y scored a ppg in the CHL in spite of being 5'10.

Now you have to watch him closely to see how he does it and determine if he is likely to transition successfully or not. Skinny 5'10 is less likely to succeed than muscular, built like a brick ****house 5'10. All else being equal. Fast 5'10 is more likely to succeed than slow 5'10. But you need to do a similar evaluation of 6'5. High IQ, good puck skills 6'5 is more likely to succeed than simply overpowers his smaller opponents 6'5.

 

[Whileee]

Also, halo effect.

Does size predict NHLers because of value or bias inherent in decision makers or both.

Probably both. However, I think that the size bias pendulum has been swinging back, and might actually overshoot a bit in the future. I don't foresee a day when size has no impact on performance and results in NHL hockey. Moreover, I have a notion (completely unsupported by data) that a higher proportion of big players are "late bloomers", due to not mastering skating and some of the skill aspects due to growth spurts and lack of physical coordination at ages 16 and 17.

I still think that there will be a role for "projection" beyond actual production results in the future, regardless of the sophistication of the model. Until someone shows me models with r-squared values much higher than I've seen, and much tighter confidence intervals around statistical parameters, I'll maintain that although models are of high value and some models are demonstrably better than some of the "traditional" scouting approaches, we would do well to be somewhat circumspect about how we take some general findings of models as a way of critiquing individual draft choices.

 

[Whileee]

The mistake is in thinking that size is an addition to how good a player is. Player X scored a ppg in the CHL - AND he is 6'5 is where it goes wrong. Player X scored a ppg in the CHL BECAUSE he is 6'5.

The models that take height into account only do a little better and that is because being big gives a little better chance of transitioning to the next level where the avg player is a little larger.

Player Y scored a ppg in the CHL in spite of being 5'10.

Now you have to watch him closely to see how he does it and determine if he is likely to transition successfully or not. Skinny 5'10 is less likely to succeed than muscular, built like a brick ****house 5'10. All else being equal. Fast 5'10 is more likely to succeed than slow 5'10. But you need to do a similar evaluation of 6'5. High IQ, good puck skills 6'5 is more likely to succeed than simply overpowers his smaller opponents 6'5.

See my post above. I agree to an extent, but disagree on two points.

1) 16 and 17 year olds are often on a steep curve in terms of physical development, coordination and application of skills. That's why there is a lot of "projection" in scouting. For some players, it's not he's scoring "because" he's 6'5", but "despite" being 6'5". For some big players (or skinny players, or small players), you have to project where they will be when they are fully developed and coordinated physically. That's why some big players are "late bloomers". This is not a rule, of course, but it is a factor that I think needs to be considered.

2) I think that there is something to the fact performance from junior to the NHL translates better if there aren't physical constraints. Would Scheifele be as effective in the NHL if he still lacked size/strength as he did 4-5 years ago? I doubt it. He's been able to more fully translate his skill into production because of his physical size and strength. When the Jets drafted him, they had to project how his production might improve and translate when he was stronger and more fully coordinated. If they had strictly looked at his production, they wouldn't have selected him at #7 overall.

Obviously, size bias creates a lot of bad decisions. However, I am pushing back a bit at what I think might be a bit too much dismissal of the role of size / strength. Some players transcend size through skill, but a large majority require size / strength to fully maximize their productivity at the NHL level.

 

[Grind]

I previously showed already that size only a significant variable to making the NHL, not in separating the quality of those that had.

In Which case logic hold that its probably more of an indicator that it's a bias.

Especially if you found the impact of small players was on average/floor/median/whichever method is best, were consistently or at least on trend with having a higher impact then larger players.


...which I feel like is a study that's been done and I've read somewhere

 

[Grind]

I think whilee you may be mistaking the echo chamber that is this board or your social (maybe online) circles being representative of a shift in actual draft tendancys.

I haven't seen anything yet that looks like the needle has moved in aggregate across the league much, if at all, and certainly not to the point where the little guys are being overvalued.

Rudimentary models like ours might, but I don't think the league/scouting as a whole has adopted that mindset.


Edit: please no one mistake ny use of the term echo chamber as a jab or something derogatory, it's just that I have personally fell prey to the thinking that what's often agreed upon as truths HERE does not transfer to a couple of beers with the boys

 

[Whileee]

I think whilee you may be mistaking the echo chamber that is this board or your social (maybe online) circles being representative of a shift in actual draft tendancys.

I haven't seen anything yet that looks like the needle has moved in aggregate across the league much, if at all, and certainly not to the point where the little guys are being overvalued.

Rudimentary models like ours might, but I don't think the league/scouting as a whole has adopted that mindset.


Edit: please no one mistake ny use of the term echo chamber as a jab or something derogatory, it's just that I have personally fell prey to the thinking that what's often agreed upon as truths HERE does not transfer to a couple of beers with the boys

Don't know, Grind. Even a few of the TV commentators mentioned how many smaller players were being drafted early. There were some notable small players taken quite early, like Brannstrom, with a big and pretty talented guy like Hague taken in the second. I think that's a signal of some change, though I haven't analyzed the data.

 

[Whileee]

Not if they don't turn out ever and aren't effective.


The effectiveness of small 40 point scorers with no two way game is vastly underrated, and the opposite is true of "the next lucic".

No matter how you cut it, swinging for "the next lucic" when there's almost always better players who are more likely to pan out I'd always a bad idea.

There are very few players like Lucic. There are also very few players like Fleury. Going for either type will fall in the large majority of instances. NHL teams are more likely to try for a Lucic, which reduces the observed success rate.

 

[garret9]

My "The search for the next..." had more to laugh at teams actively looking for exceptionally unique situations in stead of looking for the best possible player.

They are search for a snowflake like the one they saw before.

It wasn't about size, per se.

 

[Whileee]

My "The search for the next..." had more to laugh at teams actively looking for exceptionally unique situations in stead of looking for the best possible player.

They are search for a snowflake like the one they saw before.

It wasn't about size, per se.

I'm with you on this...

 

[Mortimer Snerd]

See my post above. I agree to an extent, but disagree on two points.

1) 16 and 17 year olds are often on a steep curve in terms of physical development, coordination and application of skills. That's why there is a lot of "projection" in scouting. For some players, it's not he's scoring "because" he's 6'5", but "despite" being 6'5". For some big players (or skinny players, or small players), you have to project where they will be when they are fully developed and coordinated physically. That's why some big players are "late bloomers". This is not a rule, of course, but it is a factor that I think needs to be considered.

2) I think that there is something to the fact performance from junior to the NHL translates better if there aren't physical constraints. Would Scheifele be as effective in the NHL if he still lacked size/strength as he did 4-5 years ago? I doubt it. He's been able to more fully translate his skill into production because of his physical size and strength. When the Jets drafted him, they had to project how his production might improve and translate when he was stronger and more fully coordinated. If they had strictly looked at his production, they wouldn't have selected him at #7 overall.

Obviously, size bias creates a lot of bad decisions. However, I am pushing back a bit at what I think might be a bit too much dismissal of the role of size / strength. Some players transcend size through skill, but a large majority require size / strength to fully maximize their productivity at the NHL level.

Don't disagree Whileee. There is a risk of overreacting to the size bias if we reject players because they are big. All these factors need to be observed and weighed. A late growth spurt might be more likely to lead to the late bloomer. That information should be available. OTOH we shouldn't assume a player will grow into his body just because he is big and awkward.

 

[Whileee]

Don't disagree Whileee. There is a risk of overreacting to the size bias if we reject players because they are big. All these factors need to be observed and weighed. A late growth spurt might be more likely to lead to the late bloomer. That information should be available. OTOH we shouldn't assume a player will grow into his body just because he is big and awkward.

I kind of like how the Jets have picked some over-agers in the draft. I think that's one way to perhaps pick up some later developing players, and there might still be an under-appreciation of that group in general across the NHL.

But increasingly I think the difference between teams in scouting is likely to diminish. It will be important for teams to find and exploit marginal inefficiencies, but perhaps more important to understand the actual value of picks vs. players vs. prospects. I still believe that picks are over-valued near the draft and undervalued around the trade deadline. A team that understands that and acts consistently on that basis could perhaps gain some relative advantage.

 

[Mortimer Snerd]

I kind of like how the Jets have picked some over-agers in the draft. I think that's one way to perhaps pick up some later developing players, and there might still be an under-appreciation of that group in general across the NHL.

But increasingly I think the difference between teams in scouting is likely to diminish. It will be important for teams to find and exploit marginal inefficiencies, but perhaps more important to understand the actual value of picks vs. players vs. prospects. I still believe that picks are over-valued near the draft and undervalued around the trade deadline. A team that understands that and acts consistently on that basis could perhaps gain some relative advantage.

I'm not sure about the overage players. So far none have developed much beyond where they were in the year they were finally drafted. I believe Poolman will be the first - when/if he makes it.

I'm not wild about taking Kovacevic in the 3rd rd. If we are going to keep trying these overage players I think it should be limited to the later rounds. His stat line looks pretty good compared to this year's draft class but remember that it was already his D+2 year. It should look good.

I agree with the rest. Quite strongly. It will become increasingly difficult to outdraft the competition except by picking earlier. That is a naturally occurring, evolutionary process. Even the variations in valuing these assets will level out.

A variation I see that is actually a little humourous is that picks seem to have a lesser value the day before the draft than the day after. As soon as teams put a name to a pick they seem to become more possessive of it. That overvaluation seems to persist until the player has proved out, one way or another. Picks are traded all the time. Even at the draft. Prospects are hoarded like diamonds or gold.

 

[garret9]

Copp was an overager.