Our 6th Overall Pick: Elias Lindholm or Sean Monahan?

[Erick Estrada]

Originally Posted by edslunch

So well it might be that soccer has 10 times the population of potential players due the global popularity, the top 30 will still be relatively as good as the top 30 football players because both populations are relatively large And there is an effective limit to how good a player can be in each of those games.

That's like saying that $100 Canadian is relative to $100 American. It's not as $100 will buy you less in Canada than it will buy you in the US. It's all relative to what you can attain with the picks and in the NFL you can attain much more with 2nd to 5th round picks than you can in the NHL.

[Enoch Root]

Originally Posted by Erick Estrada

Can't compare the NFL and NHL drafts. A 4th round pick in the NFL can be a starter on day one of his first season while a 4th round NHL player faces much steeper odds of being a player and most don't make it period. They relativity only relates to the numerical order. When you trade a 2nd round pick in the NFL you are trading a starting player when you trade a 2nd round pick in the NHL you are trading a player that has less than a 20% chance of ever playing in the NHL. It's simply not the same and can't be compared when it comes to calculating relative values of top 10 picks.

Last week Raiders traded the #3 overall pick to move down to #13 and got only a mid 2nd round pick in return. No team in the NHL is going to trade a top 5 pick for a later 1st round pick to gain just a 2nd round pick. Simply not going to happen.

I don't know how many different ways that I can explain that the details don't matter - it is simply a function of RELATIVE change WITHIN a set of data.

As for the trade, that is one specific trade. Also, statistics create a line (curve in this case) of best fit - that does not mean that every data sample will fall on the line. There is always variance.

The average height of a MHL player is 6'0" (or whatever it is). If you randomly choose one player, he could be 6'7" but that doesn't change the fact that if you choose one player randomly, you should expect him to be 6'0"

[Enoch Root]

Originally Posted by Erick Estrada

That's like saying that $100 Canadian is relative to $100 American. It's not as $100 will buy you less in Canada than it will buy you in the US. It's all relative to what you can attain with the picks and in the NFL you can attain much more with 2nd to 5th round picks than you can in the NHL.

EE: comparing a 2nd in the NFL to a 2nd in the NHL is irrelevant.

All that matters is that a 2nd is less valuable than a 1st. And as quality drops, quantity rises. It is a relative distribution. Period.

[Erick Estrada]

Originally Posted by Enoch Root

EE: comparing a 2nd in the NFL to a 2nd in the NHL is irrelevant.

All that matters is that a 2nd is less valuable than a 1st. And as quality drops, quantity rises. It is a relative distribution. Period.

It's only relative if you are looking at straight numbers and not the variables they represent. Your argument is flawed. If you looked at the curve of the top 200 NFL players drafted compared to the top 200 NHL players drafted over the years you would see that the 2nd round in the NFL produces nearly just as many quality players over the long haul as the 1st round and the curve would be relatively flat starting to decline slowly at the 3rd round to the 4th round. In the NHL the curve would drop steeply after only the top 10 picks of the first round. Therefore you can plainly see that a 2nd round pick in the NFL is not relative to the the worth of an NHL 2nd round pick.

[Enoch Root]

Originally Posted by Erick Estrada

It's only relative if you are looking at straight numbers and not the variables they represent. Your argument is flawed. If you looked at the curve of the top 200 NFL players drafted compared to the top 200 NHL players drafted over the years you would see that the 2nd round in the NFL produces nearly just as many quality players over the long haul as the 1st round and the curve would be relatively flat starting to decline slowly at the 3rd round to the 4th round. In the NHL the curve would drop steeply after only the top 10 picks of the first round. Therefore you can plainly see that a 2nd round pick in the NFL is not relative to the the worth of an NHL 2nd round pick.

<sigh>

I understand what you are trying to say, but that is a qualitative statement, not a statistical relation.

Again, I will use the height distribution example I used before: the differences in height between Philipino men is much less than the differences among European men. But that is irrelevant - standard deviation accounts for that.

I have stated my argument so I am done - I don't want to derail this thread any further.

[GGG]

You refuse to recognize that you are making two fundemental assumptions.

1) the distribution of high end talent is the same in all sports.
2) that there are no discontinuities in the draft.

The second one is the easiest to address. Generational players or even elite players are discontinuities. Their value is significantly higher than any statisical model can suggest because there arent enough crosbys to be significant. The model states that drafting Crosby and Drafting Yakupov has the same value. So if at any point there is a non normal drop off in talent the model doesnt account for it.

To illustrate the first point if we lowered the draft age to 10 the value between the 1st pick and the 10th pick ate equal. At the age of 10 we have very little predictive power so using your table to trade picks would fail.

[ricardodw]

In 1979 (the last year of 20 year old draft and the WHA forcing the issue by signing taking the cream of the 19 year olds) the first 30 picks had career averaging 749 games. picks 31-60 average career was 452 games ... 3 picks in the top 60 never made it to the NHL.... another 6 played less than 80 NHL games.... 21 out of the 30 played more than 80 NHL games!!

Taking the 1990 draft as an example of the 18 year old draft.... average career of top 30 - 566 games 31-60 240 games. 18 of the top 60 never played 1 NHL game.

10 more played less than 80 games..... 32 out of the 60 played more than 80 games (1 season)


Drafting 18 year old kids has made the NHL draft a crap shoot. It is almost impossible to predict how an 18 year old will do in the NHL ...... It was a lot easier evaluating 20 year olds.


1990 is not unique..... in 2000 there were 12 players drafted in the top 60 with no NHL games and 16 more with less than 80 games ......

[opendoor]

Originally Posted by Enoch Root

Just for some perspective and comparison, the NFL draft value chart separates the 4th and 6th picks by 200 points, which would equate to a mid 3rd round pick (78th pick to be precise).

Offering up a 1st round pick or an NHL player like Tanguay to move from 6th to 4th is overpayment.

And in all likelihood, we'll get the player we want at 6th anyway.

Now, if you are offering a player that isn't in your plans anyway, well then carry on. And that may well be the case with Tanguay.

I'm not sure those values translate to the NHL, and if they do trading down is extremely undervalued. Just looking at the most recent draft where there was a lot of trading up and down all through the 1st round (2008) and it's pretty clear that the team moving up paid a premium every single time using the NFL numbers:


5th (1700 pts) was traded for 7th (1500 pts), 37th (530 pts), and 68th (250 pts). So that's 1700 pts traded for 2280.


7th (1500 pts) was traded for 9th (1350 pts) and 40th (500 pts). That's 1500 pts for 1850 pts

The Islanders were responsible for both of those trades which means they were easily able to turn an asset worth 1700 points into ones worth 2630 points which is a 55% increase using the NFL model.


And here are some other ones:

12th (1200 pts) for 17th (950 pts) and 28th (660 pts). That's 1200 traded for 1610.


12th (1200 pts) for 13th (1150 pts) and 74th (220 pts). That's 1200 traded for 1370.


15th (1050 pts) for 18th (900 pts) and 70th (240 pts). That's 1050 traded for 1140.


21st (800 pts) for 23rd (760 pts) and 54th (360 pts). That's 800 traded for 1120.


23rd (760 pts) for 24th (740 pts) and 73rd (220 pts). That's 760 traded for 960.


An argument can be made that the NFL numbers reflect the true value of the picks, but that doesn't seem to be how GMs actually value them. If we are to predict what it'll cost to move up in a draft, I'd expect the team trading up to have to give up about 25% more in points than they're receiving based on past years. In fact, just glancing through the past half decade of drafts I couldn't find a single instance in the 1st round where a team trading down received fewer points than they gave up. Given that, I'd argue the NFL numbers do a poor job of predicting the real cost of trading up in the NHL without some sort of adjustment.

Assuming the 25% figure is roughly correct (and it seems to be looking at other drafts), then a move from 6th to 4th would probably require Calgary to give up assets worth about 2250 points which would be their 6th (1600 pts) and a late 1st.

[Enoch Root]

Originally Posted by GGG

You refuse to recognize that you are making two fundemental assumptions.

1) the distribution of high end talent is the same in all sports.
2) that there are no discontinuities in the draft.

The second one is the easiest to address. Generational players or even elite players are discontinuities. Their value is significantly higher than any statisical model can suggest because there arent enough crosbys to be significant. The model states that drafting Crosby and Drafting Yakupov has the same value. So if at any point there is a non normal drop off in talent the model doesnt account for it.

To illustrate the first point if we lowered the draft age to 10 the value between the 1st pick and the 10th pick ate equal. At the age of 10 we have very little predictive power so using your table to trade picks would fail.

I assume you are referring to me...

1) at no time did I ever say this and in fact, I almost said the opposite: I am saying that the distributions do not have to be the same at all, simply that on a relative statistical basis, the gradient, from elite to average, in any group (regardless of the specifics of the group) will follow the same pattern

2) again, I already addressed this: the valuation chart, like any distribution pattern, is a line of best fit. At no time did I ever suggest that all data points would fall on the line - every year, and every individual is different. However, in the aggregate, patterns are consistent.

With respect to your example of a draft of 10 year olds - you are correct, there would be no value there and the chart would not apply. However, the chart does apply to the draft because the players do in fact have value and that value is assessable (or else they would not bother drafting)

Re the bold: no, the model does not state that they are equal. The model is a model: that means it is a guideline (line of best fit), but does not, at any time, claim to be absolute.

[Enoch Root]

Originally Posted by opendoor

your post...

I am not going to go back and look at all of the examples in the NHL of picks being traded. Assuming that your examples are representative...

Basically what you have demonstrated with this small group of trades is that the NHL GMs add a small premium, relative to the NFL charts.

If so, fine - that changes nothing. All it does it take the same ratios and multiply by a constant.

[automaton 3]

Didn't see this posted yet:

http://blogs.edmontonjournal.com/201...-monahan-have/

[opendoor]

Originally Posted by Enoch Root

I am not going to go back and look at all of the examples in the NHL of picks being traded. Assuming that your examples are representative...

Basically what you have demonstrated with this small group of trades is that the NHL GMs add a small premium, relative to the NFL charts.

If so, fine - that changes nothing. All it does it take the same ratios and multiply by a constant.

My point is that based on NHL evidence, moving up to 4th from 6th is going to cost a lot more than 3rd round pick. I think Calgary's pick and the lower of the PIT/STL pick would be a reasonable expectation of what it might cost. If Calgary still had their 2nd they could probably build something around that, but they don't.

As for the chart, there isn't really a constant. Looking at the last 7-8 drafts, in the aggregate the premium paid by the team moving up seemed to be about 25%, but it was generally higher when you get into the top half of the 1st round and varied from about 5% to 50% overall depending on the trade.

It's a pretty good guide as long as you allow for the buyer's premium that seems to be present in the NHL, but without accounting for that it leads to a pretty large undervaluing of higher picks, especially top 5 or top 10 ones.

[Walter Reed]

This may have been posted elsewhere but ...

Barkov most statistically promising Nordic forward since Peter Forsberg.
http://blogs.edmontonjournal.com/201...eter-forsberg/

[Ashasx]

Craig Button just did a mock draft on TSN with the teams from 1-14, but I missed it.

Did anyone get it?

nm: http://www.tsn.ca/draftcentre/feature/?id=93427

[$ven27]

Originally Posted by Ashasx

Craig Button just did a mock draft on TSN with the teams from 1-14, but I missed it.

Did anyone get it?

nm: http://www.tsn.ca/draftcentre/feature/?id=93427

1: Jones
2: MacKinnon
3: Drouin
4: Barkov
5: Nichkuskin
6: Monahan
7: Lindholm
That's all I got.

[moon]

8. Nichuskin
9. Zadorov
10. Wennberg
11. Morin
12. Shinkaruk
13. Domi
14. Ristolainen

[drPepper1]

Button's mock draft:

Carolina: Nurse
Flames: Monahan
Oilers: Lindholm
Sabres: Nichuskin

[GGG]

Originally Posted by Enoch Root

Just for some perspective and comparison, the NFL draft value chart separates the 4th and 6th picks by 200 points, which would equate to a mid 3rd round pick (78th pick to be precise).
.

This is the statement I disagree with. You are using the chart without recalibrating based on the standard talent distribution curve of the NHL vs the NFL. This is along the lines of the exampe of trying to draft 10 year olds with the same model but not as extreme.

Additionally it ignores that scouts statements that there is a top 4 that is better and a top 7-9 that is better with discrete drop offs between the groups. Since we know these things you cant just say it should only take a 3rd round pick based on this chart.

I will agree with you that in general establishing a drafte value ranking system is possible with calibration based on the curve for that particular sport which can be used to quickly evaluate a winner or overpayment. However your above statement is not supported by the model you are using.

[Hockey_Ninja]

Originally Posted by drPepper1

Button's mock draft:

Carolina: Nurse
Flames: Monahan
Oilers: Lindholm
Sabres: Nichuskin

That's how i see it except with Lindholm and Monahan switched.

[nik-]

hmmm, maybe Monahan is the Flames' guy.