GM Boot camp lesson 2
Yesterday the statistics (and I) told you the value for running backs begins with the second round. To most stats geeks, if you've seen one running back you've seen 'em all.
Today I bring you a lesson in player value. We often talk about points charts when evaluating draft pick trades. Let's look at it from the perspective of expected career approximate value over average (eCAVOA).
eCAVOA is defined as the expected value of that player's career over the average.
With that in mind let's look at the Robert Griffin III trade as explained by the Harvard College Sports Analysis Collective:
From an expected value perspective, the Redskins definitively lost this trade (to put it mildly). The second overall pick carries an expected Career Approximate Value Over Average (eCAVOA) of 435.4. The 6th and 38th overall picks have a combined eCAVOA of 525.1. If the Redskins had given up just these picks, they would have lost 89.7 eCAVOA, which is the equivalent of the 114th overall pick (the middle of the 4th round).
If this price had been the extent of the trade, it would have been defensible. A 525.1 eCAVOA translates to a CAV of 78.7, essentially equaling Matt Hasselbeck’s CAV. So RGIII would have only had to equal Hasselbeck for this trade to be equal.
However, the Redskins gave much more: their next two first round draft picks. The average expected value of a first round pick is 276.8 eCAVOA, which brings the total eCAVOA the Redskins gave up to 1078.7. The Rams only gave up 435.4 eCAVOA, giving them a net gain of 643.3 eCAVOA, equivalent to the first and 57th overall draft picks.
That brings us to GM lesson number 2: don't trade high draft picks for less value. Seems like a no-brainer right? Wrong. History is littered with teams trading picks based on the points chart and getting screwed. It's a variant of the commonly used term, "a reach." Want a good example? Look no further than Tony Ugoh.
Ugoh has a weighted career average value of 16. What did the Colts give up to draft him in terms of eCAVOA? The cumulative eCAVOA of those picks was the equivalent of the 9th overall pick. What could you have gotten with a lottery pick like that? Ryan Clady.
GM boot camp lesson #2: learn to value your trades appropriately in order to never overpay for draft day trades.
"What did the Colts give up to draft him in terms of eCAVOA? The cumulative eCAVOA of those picks was the equivalent of the 9th overall pick. What could you have gotten with a lottery pick like that? Ryan Clady."
Can you show me your math there? Are you crediting them with having given up the 'average' 1st round pick instead of the 27th pick that they actually gave up?
@Nate Dunlevy Sure. To get Tony Ugoh they traded the first round pick from 2008 plus a fourth round pick. The career approximate value over average for those two picks is 211.7 (first round) plus 77.5 (fourth round) for a total of 289.5. The value of first round picks has varied a bit but over the last ten years it would have earned you as much as the 9th overall pick depending on a variety of factors including GMs involved, their pick valuation methodology and players on the board. It's always possible the Colts would have gotten much less for their first overall pick in 2008 as well--something impossible to measure because at the time no one knew if they'd pick 15th or 32nd.
We could go a step further and discount the 2008 first round pick as teams like to do and we get closer to the 42nd overall pick (where they got Ugoh). Teams typically discount next year's pick at 173% according to HSAC. I would argue that's a mistake driven largely by the outdated points chart teams have been using. In that chart the first overall pick is worth nearly 6 times what the 33 overall pick is worth. In eCAVOA terms it's about 3 times the value or approximately a discount reduction of 50%. Further, assuming that a player will contribute what amounts to their CAV in their rookie year (the only year involved in the discount process) is simply unlikely. In that light using the old points chart it is remarkably easy to justify the Ugoh pick.
Need? That's another story. In that light the Colts badly needed someone to fill in the for the retiring Tarik Glenn. Those considerations are at least part of what lead to the Colts taking a huge risk and paying too much (in terms of pick valuation) for what ended up being a very average at best player.
Except they couldn't have had Ryan Clady, because there's no way on God's green earth an NFL team trades down out of the top-10 for a fourth-rounder, no matter what Harvard's numbers say. You can argue that top-10 picks are inflated by a broken system but that's the system that exists.
What the Colts could have had was Dashon Goldson and Kentwan Balmer. You can argue if spot-welding those two guys together provides as much value as what the Colts hoped they were getting in Ugoh.
@squirrel Shanahan made some similar deals while in Denver: http://en.wikipedia.org/wiki/List_of_Denver_Broncos_first-round_draft_picks
@Nate Dunlevy Oh one last thing Nate. If you decide to play by their rules for discount you'd have to value the 2007 4th round pick much higher. So if you take the points for the 28th overall pick in 2008 (211.7) and the adjusted value of the 2007 4th round pick (134.075 after adjustment) you end up with the 6th overall pick in 2008. That's a stretch to me by the way but it's how the numbers work out if you use the average discount percentage teams have paid over the years for picks in consecutive drafts.
