Cleveland Browns: Kevin Meers at the forefront of analytics

The Cleveland Browns retain premier talent evaluators, including John DePodesta and his protégé Kevin Meers, who is a rising star in the new analytics era in the NFL.

The Cleveland Browns dismissed analytics specialist Sashi Brown in favor of John Dorsey in December of2017, but they still retain premier talent in the front office from the Sashi era including John DePodesta and his protégé Kevin Meers, who is a rising star in the new analytics era in the NFL.

“Analytics” refers to the process of using tools of mathematics and statistics to guide decisions on personnel. “Moneyball” might be an apt synonym. It doesn’t replace the gut feel of a professional scout, but it does place a framework around it. Cleveland went all in for analytics during Sashi Brown’s tenure, but even after his departure John Dorsey and the front office have analytics strategists helping them to make decisions.

Chief strategy officer Paul DePodesta is the most well-known of these strategists, having succeeded brilliantly with Moneyball in Major League Baseball. DePodesta has Kevin Meers working under him as director of research and strategy. Meers is a Harvard graduate, as are DePodesta and Brown.

While a student at Harvard, Meers had already made a name for himself by creating a new draft value chart, which attempts to provide a guide for trading draft picks. This was based on a Pro Football Reference stat called Career Approximate Value (CAV), then statistically correlated the player’s value to their draft position based on historical data.

Meer’s estimates can be compared with the unofficial league standard, sometimes referred to as the Brandt point system created by the Dallas Cowboys in the late 1980s and early 1990s. VP of player personnel Gil Brandt is generally credited with doing most of the actual analysis, aided by GM Tex Schramm, and head coach Jimmy Johnson.

Meers’ estimates may be found here, as well as a comparison to the older Brandt estimates.

Meers Draft Valuation Method

Brandt Point System Explained

There are major differences between Meers and Brandt. Meers doesn’t value the early draft picks nearly as strongly as the Brandt estimate, but Meers values late round picks more than Brandt. Some examples:

• Brandt estimates that the first overall pick could be fairly swapped for 3 mid-first-round picks (15th, 16th and 17th overall).  Meers would trade first overall for only two later picks (18th and 19th)
• Brandt estimates the No. 16 pick on the second round is worth 42 percent of the No. 16 overall pick in Round 1. Meers estimates the difference to be 58 percent. That is, Meers believes that the second round is not that big of a drop from the first round.
• Brandt assigns a very low value to late rounds, while Meers is much higher.  So, for example, Meers would be willing to trade a fourth-round pick for two seventh-round picks. Brandt would want two fifth-round picks to make the same deal.

Who is right?  Meers believed, as a college kid, that he had proved that NFL teams overestimated their talent evaluation ability. However, he neglected the other obvious possibility, that the Dallas Cowboys might be much better at scouting and evaluating player performance than Pro Football Reference.

More from Browns Draft

• Browns mock draft: Mazi Smith headlines 2023 class
• One and Done: First year Browns players that didn’t earn a second season
• 3 Michigan Wolverines Cleveland Browns could draft
• 3 positions the Cleveland Browns must draft in 2023
• 2023 draft class suits Browns needs at wide receiver

Probably the results of Meers’ analysis should not be taken too seriously, as he was only a student at the time, not a paid professional as he is today. The problem is that since the entire analysis used a stat from Pro Football Reference, the correlation can no more accurate than Pro Football Reference’s ability to evaluate talent. If you think about it, if some random stat were used, then there might be no correlation at all to draft position. The fact that it does correlate at all proves that PFR’s CAV has some validity, but not necessarily the best estimate.

Again, it may not be the methodology but the quality of the CAV that shows a poorer correlation to draft position. If there were a stat is more accurate than CAV, Meers’ model might look more like Brandt’s.

Very likely a pro team’s personnel department is much more accurate than estimates made by a football publication, no matter how meticulously assembled, and there is no evidence produced that says that Pro Football Reference is more accurate than the Dallas Cowboys for evaluating player performance and value.

We can only hope that the Browns’ analytics crew can populate the inputs to Meers’ model more accurately than the CAV stat. Otherwise, the Browns are going to do cuckoo trades like accepting two late round first round picks over Baker Mayfield. Perish the thought.

Kidding aside, it is very likely that Browns do have a much higher valuation of first-round draft picks than suggested in the original Meers model, because if they were really following that,  some team would have for sure swept in and traded for the Browns’ 4th overall pick last year, which was kept and used on Denzel Ward.

Anyway, Meers probably didn’t intend to establish himself as a draft authority. At that time he was a college kid with a great interest in sports, but his blog piece took off and became widely cited.

He now has access to much more sophisticated models of the NFL draft, and no doubt has greatly improved the model that he first created as an undergraduate student at Harvard. Fans need not worry that he will usurp power from John Dorsey, but he will be there to offer advice on trade proposals.