Jared uses the work of of David Berri and Rob Simmons to show something the NFL draft for quarterbacks is pretty random. Yes, when we examine higher draft picks they have more yards, completions and touchdowns (all things that good quarterbacks do) But when when we control per possession, we see that top drafted QBs aren’t that really any better than their peers drafted later.
The real key Jared uses is the chart above. If we look at hall of fame quarterbacks (consensus greats) we see a drastic slope downwards based on draft pick. But, the key is we see “jumps” at the end of the chart. Jared says this is a J shaped graph. In essence, the presence of outliers on the far right shows that we’re likely missing gems early.
Jared follows the logic Dave and Rob put forth. When we draft a quarterback high, we heap lots of playing time on them. We believe them good, so we play them like they are. When a player is drafted late, we don’t think of them this way, so we don’t give them minutes. Sound familiar?
I think this kind of trend is important to notice in sports. The NBA is another prime market to consider for this. Yes, a majority of our stars, franchise players, and hall of famers are top picks. But the presence of greats picked outside of the lottery says a lot (Dennis Rodman, Ben Wallace, Manu Ginobili to name a few) And we still have active players that have played very well and were second round picks (David Lee, Carlos Boozer, Paul Millsap, Marcin Gortat) We like to think that our evaluations are really good. But if we find jumps at the far end of our spectrum, we definitely need to revisit the problem. And it’s a problem that definitely exists in the NFL and the NBA. (that great people like Arturo have tried to help with, if only GMs would listen.)