A team that runs a 4-3 defense would probably carry six linebackers. One starter and one backup for each position. The same team usually carries three QBs, so one starter and two backups, but at any given time, no more than two guys are generally in the running for the starting gig. (The third guy is usually a developmental guy that won’t see any playing time unless the first two guys get hurt.) Teams might play 4-5 linebackers throughout the game whereas they’re most likely to play only one QB.

To sum up, your point might account for a small part of the increase in the chances that a LB could develop into a starter but not for all of it. One could even argue that it makes LBs even more draftable due to their versatility. If a better player comes along, you could move your starter-caliber LB to another spot and still get something out of him. Not so with a QB that gets usurped.

WR’s, on the other hand, have always been a toss up. I suspect most of that is because of the looseness of the college game. It’s easy for a WR to catch 90 balls when they don’t have to deal with press coverage, complicated zone schemes and, more importantly, run blocking. (By which I mean, it’s easier to catch more balls on a team where passing is the first option).

You’re taking this saying too far. While it’s true that correlation doesn’t automatically prove causation, it doesn’t disprove causation, either. That is, the correlation could very well be a result of causation.

If you read the article, we clearly discuss “all else being equal” scenarios — a team has scouted and interviewed all available options, and has narrowed the group to just a few possible picks. Only then are we saying that this table can be of use.

Correlation does not ever indicate causality. Ever.

In this chart you make the argument that because some positions ended up with a much higher percentage of starters, etc. for a NFL team, a team should use these to assess the quality of a potential draft choice.

But there is no evidence that a third factor is heavily influencing both the “star” and “draft pick” numbers:

Perhaps when someone is a high draft pick at a line or special teams position, it means they are truly outstanding *because only an outstanding player at these positions would ever make it into the draft!*

That is, a very good player at certain positions is likely to be drafted, but a very good player at other positions is unlikely to be drafted early. Thus when you see a center among the top draft picks, you know this guy is amazing, exactly because centers–even very good centers–don’t usually make it into the early draft picks.

Another possibility is that not all teams are alike, that is to say, certain types of teams (or teams at a certain point in the “build a good team, be on top for a decade, then drop again” cycle) are more likely to have collected high draft picks.

A team that trades away a decent player for a high draft pick is indicating more about their own aspirations in the long term/short term category than about the players they draft. Thus the list of “teams in great need of some star power in certain positions” may be very similar to the list of “teams with multiple high draft picks due to the team’s recent (poor) performance and their willingness to “trade for the future.”

It would be reasonable to expect that teams who tend to have several high draft picks tend to have the greatest need for a strong player around which to build an offense and thus a future team.

Thus it might be likely that the list of teams with the highest share of top draft picks is strikingly similar to the list of teams looking for “impact players,” players who, by virtue of their positions, are typically more likely to assume team leadership roles.

And it is also completely feasible to assume that teams making particular early high draft picks are at least as interested in selling season tickets than in compiling a 20-5 rather than a 16-14 record over a couple of seasons.

Like it or not, fans are much more likely to buy jerseys and paint themselves in strange colors over a new and promising QB, LB, RB or pass receiver rather than, say, a new offensive tackle. Probably not fair, and not even particularly wise, but it’s the reality.

Yet another possible explanation for the correlation of position and player success in the draft might be the possibility that a team is selecting a draft pick they believe will be of high value in negotiating a trade for a proven team leader from another–perhaps specific–team.

Or it might be that a team considers it likely to pick up solid players in non-visible positions outside the draft than to find star players in high-profile positions. Wouldn’t it make sense for a team to use their few draft picks to find those players not likely to be available through other recruiting means? At least one might expect that a team would hold off on drafting even a superior player in an easy-to-find if the odds are high that the team would be able to recruit players in some positions outside of the draft while slim that certain other positions would be filled with promising players except through the draft.

So while it probably would not be particularly smart to bet against this table, use of its numbers as a guide also might tend to lead teams toward making bad decisions when it comes down to the smoke and mirrors and dramatic high-security, nationally televised draft pick events.

It might not be the case that picking players in certain positions early in the draft would lead a team to increase their odds that their pick is “safe.” And at some level, as with all constructed events such as the NFL draft, unnoticed factors might well be much better predictors of the wisdom of a particular team’s choice.

morawk