So I’m kinda infatuated by the NFL draft, but mostly because of how it works as an economic system of resource maximization. Limited picks having to maximize returns and savings to construct a team that can win. A big part of it is the crap shoot nature of the draft, how do you maximize returns with such uncertainty? If scouting was perfect and accurate we wouldn’t see teams perpetually drafting in the top 10 so often. However it does happen, frequently. Why?
In an attempt to illustrate how much of a crap shoot drafting in the NFL is I want to examine what scouts said about individual players and put that against their NFL production. I’m going to focus on running backs. Mostly because I think Running backs have good statistics to use and are less married to their system. For example if a scouting report calls a QB accurate, what stat do we use to judge that?
-But that’s very much a design of system, if teams run quick pass systems and west coast offenses it will produce higher accuracy by stretching the field horizontally and shortening the field for the QB. However playing in a team that wants more of a downfield passing game will produce with the same QB lower accuracy. Also a QB can complete a lot of bad passes, they can be completed because the defense isn’t bothering to defend that short a route, or because his WRs make great plays on the ball.
Yards after catch?
-Makes sense, if a QB can put the ball right on his receiver the player has a greater chance to produce positive yards after the catch (since he can catch in stride and not have to adjust). Once again this is very scheme dependent.
Yards Per Attempt?
-To a large extent this evens both of the previous out to reward high completion percentages that produce advancable balls. This does however favor guys in systems that pass less and play action pass down the field more.
So to dissect this the best would probably be to go to the film room. I don’t have that kind of time or access to game tape. and then it comes down to perception and my ability to judge these things. If I disagree with a Kiper or Mayock well odds are they’re better at watching tape than me so it proves nothing.
Running backs are dependent on their opportunities and offensive line, this is true but we can see certain things based on readily avaliable stats, total yards, yards per touch, touchdowns etc.
What I’m interested in is also looking at their yardage distribution. If someone is power-back without great speed the expectation is to see a distribution of runs with out many for a loss and not many for over 10 yards, with the largest bins being between 2-6 yards or there about. This should be fairly independent of offensive line, we might not see as many +4 gains from them but we should still see a low distribution of runs for a loss. Simple reason is that powerful backs tend to run down hill hit the hole quickly and rarely are stopped for a loss because of the difficulty of stopping them 1-on-1 and their tendency to not dance in the backfield looking for extra yards.
If a guy is supposed be quick but lack ‘between the tackles’ power we would expect to see more of a boom and bust distribution. Since the player won’t be running straight ahead as often and may have to get to the edge to gain yards, or might try cutting in the backfield to create wider running lanes they can be expected to be stopped for a loss or small gain more often. Ideally this will be paired with a highish rate of big impact plays for 10 or 20+ yards.
Being used as a situational back can affect this. Running a lot of draws will reduce the effects of lacking power, and running at the goal line a lot will potentially increase the amount of no-gain or loss runs from a power guy since the Defense is geared to stop the run. This is something that can be parsed out by the numbers and commented on to explain a discrepancy etc.
Some back might get the ‘football jesus’ scouting report, basically: is fast and powerful, can cut on a dime and pound the ball between the tackles decisively. Here we’re basically expecting a guy with a very good distribution, not very many negative plays, a lot of 3+ and 10 or 20+.
As you can imagine this will take a while to do. My goal is to use the 2006 through 2008 draft classes since I think there’s enough time passed to judge the players on their own merits now, and it’s not so far back I won’t be able to find scouting reports. I’m also hoping for a similar age group so I can (as much as possible) have a similar data set. If for some reason in 2009 there were a lot of good run defenses then good lets let that be a part of the data.
The big snag I’ve hit is finding scouting reports. I want to get multiple (3-5 would be great) for the players to aggregate things out, and I also hope to not just pick on some guy with a webpage, I’d like to use well known scouts.
I’m also hoping to avoid having to pay for ESPN’s Insider pass because really? this day and age you’re going to charge me to read your articles when so many of your writers are crap? haha, no.