ASU has an eye on analytics this year
The date: Sept. 8, 2018.
The event: Arizona State vs. Michigan State football game at Sun Devil Stadium.
The situation: Trailing by three points with 6:33 left in the game, the Sun Devils are within reach of an upset in Herm Edwards’ second game as coach. They have the ball, but it’s fourth and 1 at the ASU 35. Everyone expects the punt team to enter the game, but Edwards decides to go for it.
The decision: What could Edwards possibly be contemplating? You punt here, hope your defense holds up to give you time for at least another possession, don’t you? But on fourth down, quarterback Manny Wilkins hands to running back Eno Benjamin.
Brace yourself, Sun Devil fans. Something very similar to that scenario could happen this season because Edwards and his staff are embracing quantitative analysis designed to give ASU its best chance to win.
And analytics might advise going
for the first down in the above scenario. If so, Edwards might be willing to follow the numbers.
“We’ve been practicing it and looking it and seeing how it works,” Edwards said. “I think everybody is going to it. It’s just, how much do you believe?”
There are several companies that supply the type of analysis ASU will be using, but it’s unknown at this point just which one the Sun Devils have chosen.
They had several conversations with EdjSports, a data intelligence firm, which, according to its website, “enables smarter decision-making and performance” and is “committed to helping our clients win more – no matter the game, race or match.”
ASU did not choose EdjSports, but several other universities and NFL teams, including the Eagles, have. And decisions made by Eagles coach Doug Pederson last season in what EdjSports calls “critical call decisions” opened eyes at all levels of the game. Or should have.
Critical call decisions are defined as fourth downs, points after touchdowns and kickoffs. What EdjSports and other companies have found is that fourth downs are packed with opportunities. Coaches often pass on those opportunities because it’s how they’ve learned the game, and they often fear the downside of failure more than they embrace the possibility of success.
Pederson and the Eagles have embraced the use of analytics more than any other football team, and to EdjSports co-founder and chief innovation officer, Frank Frigo, it’s no coincidence they won the Super Bowl.
“I think you really saw something with the Eagles and NFL last year that was distinctly different than anything we ever witnessed,” Frigo said. “That’s an indication there could be a bit of a shift going on.”
On average, an NFL team will give up 2/3 of a victory each season because of poor decisions on fourth downs, according to Frigo.
Even a bold leader such as former Cardinals head coach Bruce Arians was guilty of it. According to EdjSports, the Cardinals lost a 1/2 game last season on critical call decisions.
The Eagles, in contrast, were nearly perfect on those calls. It’s not that they went for it on fourth down an inordinate amount of times – 26, ranked third in the NFL – it’s that they did it 11 times in the first half.
Most coaches won’t do it until the game nears the end.
In the Super Bowl, Pederson went for it twice on fourth down. The Eagles converted both, including once for a touchdown late in the first half.
“Is it always going to work?” Edwards said. “No, but it means you put the team in the best possible situation to win a football game.”
EdjSports and other analytics firms can’t be there on game days for clients, nor are college and pro teams allowed to use a computer to crunch numbers during games. The companies run years of empirical data through a computer model that can simulate thousands of games with adjustments made for teams’ strengths and weaknesses, as well as decisions made by coaches.
Teams can use the data to create charts and graphs that can be used on game day. So when Edwards is faced with a critical decision, he conceivably could have an assistant in the coaches’ booth look at what the chart suggests he do.
Edwards expects most other schools will be doing the same thing in the coming years, and networks broadcasting games could start to use graphics that give fans the odds of various decisions working. Sort of like what happens in televised poker games.
Edwards won’t always go by what the charts and graphs upstairs suggest, however. You can never take “the gut” out of making decisions, he said, and specific matchups, for instance, might dictate going against the numbers.
“You can look at a game and say, ‘We can’t block that guy. I know what this thing says, but we’re punting,’ ” he said.
Asked about the risk of being second-guessed after making an unconventional decision, Edwards smiled.
“If it works, it’s great,” he said, “If it doesn’t, it’s, ‘Why you did you do that?’ That’s part of it.”
Epilogue: The scenario at the top is exactly what the Cardinals did against the Texans last November. Adrian Peterson lost a yard, the Texans scored on the next play and won by 10 points.
Minutes after the game, Arians blamed himself for the loss.
But his opinion had changed a day later.
“Take all that s--t back I said yesterday. That was a damned good call.”
An analysis by EdjSports backed up Arians’ day-after reversal.
A punt would have resulted in a 6.1percent loss in win probability.