Research Communications

Team Player 

Artificial neural network makes smart NFL draft picks

September 29, 2011

In the 2007 National Football League (NFL) draft, the Oakland Raiders selected one of the biggest draft busts of all time.

Jamarcus Russell was picked first overall and signed to a $61 million contract. Three losing seasons later, the quarterback had been cut from the team. The wins the Raiders bet on evaporated just like the $32 million Russell had been guaranteed.

William Young says he could have saved them that money.

William Young
William Young. Photo credit: Robb DeCamp.

Young, adjunct assistant professor of industrial and systems engineering at Ohio University, constructed an artificial neural network that showed Russell should have been selected ninth, not first, and clearly hired for a lot less. Entitled Heuristic Evaluation of Artificially Replaced Teammates, or HEART, his system “replaces” existing players on teams with prospective draftees and evaluates how many games a team could win if they drafted one player over another.

The algorithmic system, which developed as Young’s doctoral dissertation at Ohio University, gets 30 to 40 percent of its data from the NFL combine events, such as the 40-yard dash or long and high jumps. Veteran players don’t have current combine stats, so Young forecasts their performance using on-the-field statistics.

Young’s model also can assess player personality attributes more impartially than people can, he says. It can help teams find players that suit their organization’s culture by converting attributes such as character traits, performance with teammates, or work ethic into numbers. Almost any trait can be measured numerically and evaluated.

“I see my system as a supplemental tool that NFL teams and their scouts can use to maximize team efficiency,” Young says. “They all have limited budgets and only so many players they can hire. My tool helps them choose the right ones at the right price. Paying more for superstars doesn’t mean the team will win.”

Follow Ohio University Research News on Twitter and Facebook.

By Daniel Kelly

This article will appear in the Autumn/Winter 2011 issue of Perspectives magazine, which covers the research, scholarship, and creative activity of Ohio University faculty, staff, and students.