What"s An NHL Draft Pick Worth? Analytics revealed

[PotentPowerPlay21]

For those of you who are analytics geeks like me, the following pdf creates a statistical analysis of the value of all draft picks. It's a handy reference to evaluate trades involving draft picks. It gives each draft pick a numerical value for comparisons.

http://myslu.stlawu.edu/~msch/sports/Schuckers_NHL_Draft.pdf

 

 

[GASabresIUFAN]

That study is 11 years old.

[Curt]

19 minutes ago, PotentPowerPlay21 said:

For those of you who are analytics geeks like me, the following pdf creates a statistical analysis of the value of all draft picks. It's a handy reference to evaluate trades involving draft picks. It gives each draft pick a numerical value for comparisons.

http://myslu.stlawu.edu/~msch/sports/Schuckers_NHL_Draft.pdf

 

 

While interesting as a theoretical examination, this isn’t very useful or reflective in terms of what it actually takes to trade up/down in the NHL draft.

For example, according to this, you could trade two early 3rd round picks for 20th overall, or trade up from 5th overall to 1st overall just by tossing in an early 3rd round pick.

Seems way off.

[matter2003]

It's always amazed me that teams will draft player in the middle of the first round with the expectation they will be a solid third line player.

Imagine the NFL...drafting a player at 15 overall hoping he is a good backup player that gets some snaps.

Not sure why there is such a disparity between sports.

Edited July 3, 2022 by matter2003

[Taro T]

14 minutes ago, matter2003 said:

It's always amazed me that teams will draft player in the middle of the first round with the expectation they will be a solid third line player.

Imagine the NFL...drafting a player at 15 overall hoping he is a good backup player that gets some snaps.

Not sure why there is such a disparity between sports.

18 years old vs 22ish years old.

How many of the HS All Americans end up being NFL impact starters?

[jad1]

15 minutes ago, matter2003 said:

It's always amazed me that teams will draft player in the middle of the first round with the expectation they will be a solid third line player.

Imagine the NFL...drafting a player at 15 overall hoping he is a good backup player that gets some snaps.

Not sure why there is such a disparity between sports.

Move the minimum draft age to 20 in the NHL and GMs will begin to look a lot smarter.

[Taro T]

Plus 23 man roster vs ~60 including Practice Squad.

[Curt]

25 minutes ago, matter2003 said:

It's always amazed me that teams will draft player in the middle of the first round with the expectation they will be a solid third line player.

Imagine the NFL...drafting a player at 15 overall hoping he is a good backup player that gets some snaps.

Not sure why there is such a disparity between sports.

A 3rd line NHL player isn’t equivalent to a NFL backup.  NFL team has 22 starters, NHL team 19 starters.  A 3rd liner is a starter.  An NHL equivalent of the NFL “backup” is in the AHL or getting healthy scratched in the NHL.

Also, I’d expect that NHL careers are generally longer than NFL careers, so an NHL starter is going to hold that position longer, which means less available openings for rookies each season.

Also also, NFL players are drafted at age 21-22, compared to 18 for NHL players.  In the NFL draft, you are much more certain of what you are getting, starters early, backups later.  In the NHL there are more busts early and surprises later.

Edited July 3, 2022 by Curt

[Marvin]

1 hour ago, GASabresIUFAN said:

That study is 11 years old.

Is there an update?  I could not find one which is both significantly more recent and significantly different.

[steveoath]

1 hour ago, Marvin, Sabres Fan said:

Is there an update?  I could not find one which is both significantly more recent and significantly different.

This throws up a couple of models before suggesting a trade value chart.

https://soundofhockey.com/2022/06/06/examining-the-value-of-nhl-draft-picks/

Another interesting article that references the Shuckers piece as well as suggesting improvements. 

https://www.actionnetwork.com/nhl/2022-nhl-draft-preview-analysis-top-two-picks-value-chace-mccallum

And this is too mathy for me 

https://hockey-graphs.com/2020/08/07/applied-prospect-pipeline/#more-24440

Edited July 3, 2022 by steveoath

[sabresparaavida]

I feel as this is a pretty poor analysis for the worth of draft picks. The only criteria for worth of a pick is man games played. The picks values are derived only by the expected amount of games played in the league. One example they gave was 8 and 32 for 1 because you have an expected return of roughly the same amount of games. This has no way to track the quality of the games though. In 2015, McDavid was 1 OA and Zachary Werenski and Christian Fischer went 8 and 32. This system would likely give more value to Werenski+Fischer than to McDavid (because they together will likely play more games), even though that would be a clearly lopsided trade if it were to happen 

[steveoath]

 

13 minutes ago, sabresparaavida said:

I feel as this is a pretty poor analysis for the worth of draft picks. The only criteria for worth of a pick is man games played. The picks values are derived only by the expected amount of games played in the league. One example they gave was 8 and 32 for 1 because you have an expected return of roughly the same amount of games. This has no way to track the quality of the games though. In 2015, McDavid was 1 OA and Zachary Werenski and Christian Fischer went 8 and 32. This system would likely give more value to Werenski+Fischer than to McDavid (because they together will likely play more games), even though that would be a clearly lopsided trade if it were to happen 

Was this in ref to the OP? 

This is why I like the sound of hockey model. It doesn't look at the quality of the player but actually analyses how teams have valued the picks based on what they do. He says "... I tried to get to how teams have–by their actions–actually valued draft picks. To do this, I looked at the pick-for-pick trades made by NHL teams in recent drafts with the goal of understanding how teams have actually stacked up the value of one pick versus another"

Edited July 3, 2022 by steveoath

[sabresparaavida]

2 minutes ago, steveoath said:

 

Was this in ref to the OP? 

This is why I like the sound of hockey model. It doesn't look at the quality of the player but actually analyses how teams have valued the picks based on what they do. He says "... I tried to get to how teams have–by their actions–actually valued draft picks. To do this, I looked at the pick-for-pick trades made by NHL teams in recent drafts with the goal of understanding how teams have actually stacked up the value of one pick versus another"

Yeah, in reference to the OP. This is a better analysis IMO, but it is just so situational with how the draft differs and how rare early trades are.

[dudacek]

2 hours ago, matter2003 said:

It's always amazed me that teams will draft player in the middle of the first round with the expectation they will be a solid third line player.

Imagine the NFL...drafting a player at 15 overall hoping he is a good backup player that gets some snaps.

Not sure why there is such a disparity between sports.

I think typically only 50-60 players from each draft go on to be NHL regulars.

It would stand to reason that guys between 15 and 45 would tend to be mid-roster players, by simple math and averaging, no?

Edited July 3, 2022 by dudacek

[GASabresIUFAN]

https://www.broadstreethockey.com/2013/4/25/4262594/nhl-draft-pick-value-trading-up

I still think this is the most accurate chart that I've seen.

As to drafted players playing in the NHL, Hockeydb.com has a running total of guys who reached the NHL for at least 1 game and what their average career looks like.

This is from the 2013 draft for example.

Quote

 

Draft Stats

Total drafted players to play in NHL: 106

Percent of players to play in NHL: 50.2

Average NHL Career Games: 212

Average NHL Career Goals: 35

Average NHL Career Points: 93

Average NHL Career PIM: 93

 

So far from 2013 - 46 players have played 200 or more games and are still in the NHL as of last season.  This list includes JAGs like Hayden, Hagg, Butcher.  Another 28 who appeared in at least one game last season.  

Edited July 4, 2022 by GASabresIUFAN

[Brawndo]

Adams was asked this question during the presser last week. He mentioned that Sam Ventura and the analytics department have a draft pick value chart constructed, Adams also joked that other clubs certainly have different values assigned to their picks. KA said that in most trades involving draft picks, the other GM wants to call you back after consulting His Numbers Guy

[mjd1001]

15 hours ago, GASabresIUFAN said:

https://www.broadstreethockey.com/2013/4/25/4262594/nhl-draft-pick-value-trading-up

I still think this is the most accurate chart that I've seen.

 

I still think that one undervalues the VERY top picks and overvalues everything else.

According to that chart, the Sabres should be able to package #16 and #28 together and for something like the 7th or 8th pick.  I think a team picking #7 or #8 would want more than 16 and 28 for that.  Of course there may be rare years where a team doesn't think there is much of a difference between the 7th best guy and the 20th, but more often than not I think the higher pick would require more.

[DarthEbriate]

If the chart takes career stats into account and not just games played, I wonder how these various analyses would be impacted if Gretzky had been drafted (and in which draft slot).

[Weave]

2 hours ago, mjd1001 said:

I still think that one undervalues the VERY top picks and overvalues everything else.

According to that chart, the Sabres should be able to package #16 and #28 together and for something like the 7th or 8th pick.  I think a team picking #7 or #8 would want more than 16 and 28 for that.  Of course there may be rare years where a team doesn't think there is much of a difference between the 7th best guy and the 20th, but more often than not I think the higher pick would require more.

I suspect in actual practice the rankings and relative values of the players actually on the draft boards are taken into account and its not just output from an aggregate model.

[dudacek]

24 minutes ago, Weave said:

I suspect in actual practice the rankings and relative values of the players actually on the draft boards are taken into account and its not just output from an aggregate model.

For sure. I would think that if the Sabres ranked Peterka at 24th, they would pay their price for pick 24, even if they were only trading for pick 34.

I also think that pick 24 would be more valuable in a draft where you gave 30 players 1st rounds grades than 1 where you gave only 20 players that rank.