# On Relative Worth of Draft Picks

Chemmy
March 16 2011 02:42PM

Avalanche blogger Jibblescribbits went to the trouble to figure out a way to value NHL draft picks to determine their relative worth here (go read it now, I'll wait).

I went ahead and started with this approximation:

Or Player Value = (Career TOI/g) / (Position average TOI/g)

I decided that a player who plays more would, on most occasions, be a better player than one who played little minutes.

That seems reasonable enough to me. I wanted to play with the numbers a little bit to see what they looked like. I had seen an NFL draft pick value chart on ESPN a few years ago and thought it was an interesting little thing. For example in the NFL the top pick is worth "3000" fun points while the second pick is worth only 2600. I found Jibblescribbit's piece on draft pick valuation through Tangotiger's post at "Inside the Book" where he determined a formula to value NHL draft picks based on Jibble's research:

Setting the #1 pick’s value to a “1”, I get this equation as the best-fit:

17 / (pick + 16)

So, a #1 pick has a value of “1” (or 100%).  A #7 pick has a value of 74%.  A #18 pick has a value of 50%.  A #52 pick has a value of 25%.  Therefore, the suggestion is that trading up from 18 to 7 is equivalent to trading up from 7 to 1 or from 52 to 18.

I multiplied that formula by 1000 to set the #1 overall pick equal to 1000 draft zone super points and determined the relative worth of every one of the 210 picks in the NHL Entry Draft. The way to use this table is that the worth of the trade should balance out. If you want the 3rd overall pick worth 895 zany coins you'd need to give up picks and players worth 895 smackers.

I then thought about that NFL chart. How does the value of the 2nd overall pick in the NFL compare to the value of the 2nd overall pick in hockey? What about the 100th pick? I divided the ESPN numbers by 3 to set both 1st picks equal to 1000 units and graphed them:

While the NHL second overall pick is worth 89% of the first overall pick the NFL second overall pick is worth only 86%. It seems that the NFL draft is considered much more top heavy by the people who made ESPN's chart. Whether that's because of differences in the sport or simply that the NFL chart includes weighting based on the perceived worth of a 1st overall pick relative to 30 6th rounders is unclear without seeing their method.

An interesting exercise for sure. Be sure to visit Jibblescribbits' page to tell him what you think.

Like many people Chemmy was born. He lives with his wife and a dog named Wendel Bark.
#1 SkinnyFish
March 16 2011, 02:57PM

Shouldn't the value of NHL draft picks decrease at a higher rate than NFL draft picks simply due to the higher number of player positions in football versus hockey? In football there are at a minimum 13 positions (when not counting sides of the ball, LT, RT, etc...) In hockey there are 3.

In the NFL draft you could conceivably get the 3rd best player at any position with the 30th overall pick. You could never get the 3rd best F or D in hockey with that pick. (Goalies are enigmas)

Going further, the talent pool for the NFL draft is considerably larger than the talent pool for the NHL draft so one would expect overall talent to also decline at a lesser rate than in the NHL draft, thus giving late round picks more value.

From Jibble's post, I have a hard time equating a jump from 18th to 7th with a jump from 7th to 1st. Feel that undervalues the 1st overall relative to the other picks.

Just my two cents.

#2 MF37
March 16 2011, 03:05PM

The book Scorecasting has a chapter dedicated to the NFL valuation you mentioned. It was originally developed by the Dallas Cowboys.

The discuss it in brief here (see the first Q): http://www.freakonomics.com/2011/02/03/the-authors-of-scorecasting-answer-your-questions/