On Bayesian Probability and Why The Eagles Won’t Draft Landon Collins

Think about the last time you had to make a difficult decision. How did you go about it? Did you consider all of your options? Did you weigh the consequences of each option? Maybe you made a list detailing the pros and cons of your options. Perhaps your past experiences or the experiences of someone you know influenced your decision. If any of this describes how you came about your decision, then without knowing it, you’ve become acquainted with Bayesian probability.

According to Wikipedia, Bayesian probability is an extension of propositional logic that enables reasoning with hypotheses when given propositions whose truth or falsity is uncertain. In layman’s terms, Bayesian probability is a form of decision-making that is based on statistics, facts, or past occurrences when the actual outcome of the impending decision is impossible to know.

In the NFL, Bayesian probability is taken to another level. While we’d all like to consider ourselves experts at predicting which draft prospects each year will be NFL stars, it’s impossible to actually know until said prospects actually perform and produce on the field against NFL competition. While some teams use the concept of Bayesian probability more than others, team scouts begin forming player reports and evaluations more than a year in advance of when a given player will be drafted. In turn, the decision-makers in each organization will take the data compiled by scouts and match it with their own evaluations and opinions before finalizing a player’s position on a team’s draft board. While impossible to calculate an individual player’s future production in the NFL, teams will match a prospect’s height, weight, speed, coverage skills, ball skills, tackling ability and college production to current or former NFL players with the same measureables. In short, decision-makers around the league will use what they know about a player whose outcome is already evident to project an unknown future outcome for a similarly-measured draft prospect.

In his first appearance with members of the media since being named the Eagles’ Vice President of Player Personnel, Ed Marynowitz, without mentioning it directly, alluded to several concepts utilized when dealing with Bayesian probability. Below is one of Marynowitz’s quotes on using analytics to project draft prospects.

“Big picture wise, you want to play with the odds, not against the odds. And the odds are telling you that a majority of these guys that are under a certain prototype do not play at a starting level in the NFL. If you have seven draft picks, do you really want to waste one, especially in the top three rounds, on a guy that history is telling you… typically these guys with these types of measureables don’t produce at this level.”

So let’s say Prospect 1 is projected across-the-board to be taken in the first round of the draft. He also plays wide receiver, which is a position of need for Team A. Let’s also say that Team A assigns Prospect 1 a numerical value for each category of Prospect 1’s body of work as it relates to Team A’s offensive scheme. When finalizing the scouting report, Team A decides that Prospect 1’s size and vertical leap are elite for the wide receiver position, but his lack of separation skills make him a questionable fit for Team A’s vertical offense. They assign Prospect 1 an overall rating of 12.4 and compare him to current NFL wide receivers who were given the same rating by Team A. Later in their evaluations of wide receivers, Team A gets to the scouting report of Prospect 2, who is projected as a fourth round pick. Though his size and vertical leap are below-average for an NFL wide receiver, Prospect 2 has plus-ability to separate against any defensive coverage, which makes him a great fit for Team A. Like Prospect 1, Prospect 2 is given an overall rating of 12.4 and is compared to current NFL players and other prospective receivers with a similar score. When Team A sits down to map out a concrete draft plan, they realize there is better value in waiting until the fourth round to draft Prospect 2 instead of drafting Prospect 1 with their first pick of the draft. Team A used concepts of Bayesian probability to maximize the value of their draft picks.

The Eagles find themselves in a similar scenario at the safety position as the 2015 NFL Draft approaches. Though he certainly fits the height and weight requirements of a prototypical NFL safety and is exceptional in run support, Landon Collins’s lack of skills in coverage make him a questionable fit for the Eagles’ defense, especially since the team would likely have to select him in the first round. While declining to get into the specifics of individual players on Thursday, Marynowitz seemed to give a brief overview of whether or not Collins can thrive in the Eagles’ defense as a cover safety. “He can do it to a degree,” Marynowitz said. “All of these guys have strengths and weaknesses. Landon has the ability to do that. Maybe not to the degree that other guys do, but he certainly has the ability to do it.” While not entirely eliminating him from the realm of possibility of being selected with the Eagles’ first round pick, Marynowitz’s take on Collins was not exactly a ringing endorsement, either.

A real life example of Marynowitz’s “He can do it to a degree” assessment of Collins, while trivial, is something we’ve all done. Whether you use ear buds to listen to music, converse with the person on the other end of the line during a phone call, or while watching a hilarious video a friend sent you, we’ve all made the mistake of sticking the ear bud meant for the left ear into our right ear or vice versa. Yeah, you still hear the music, and it’s not a life-or-death situation, but the inconvenient fit is so uncomfortable (and at times painful) that we almost always immediately correct our mistake and shift the ear bud to the correct ear.

For the sake of this lighthearted example, Landon Collins is a high-priced left ear bud that would be forced to fit into the right ear of the Eagles. Sure, he would have his moments in Billy Davis’s defense, but why a use first round pick when better fits for the Eagles defense will be available later in the draft? Unless Marynowitz and Chip Kelly stray away from their apparent philosophy of selecting players who fit their schemes perfectly, Landon Collins is unlikely to hear his name called when the Eagles are on the clock on April 30th.

Follow Sidelines and Headlines creator Ray Butler on Twitter: @RayButlerII

Alabama v Auburn
Landon Collins’s coverage skills make him a risky pick for the Eagles in the first round.
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s