The sports analytics community has Mike Tomlin’s back – Pittsburgh Post-Gazette

Mike Tomlin might not have the backing of the armchair coaches who criticize him for going for two-point conversions more than any other head coach in the NFL. But you won’t hear any criticism from those in the world of sports analytics who praise him and say a simple math equation supports his penchant for two-point conversions, including the controversial second attempt Sunday afternoon against the Cowboys that has drawn the ire of many professional football traditionalists.

Before going 0 for 4 on two-point attempts against the Cowboys, the Steelers had made 14 of their previous 17 two-point tries. But beyond the Steelers’ success rate, elementary school math supports two-point attempts. 

In 2016, NFL teams are converting the 33-yard extra point kick at a 94.8 percent rate, which yields .948 points per attempt. They are converting two-point conversions at a 54 percent rate which yields 1.08 points per attempt.

But football games aren’t won with math equations, and there are some who believe Tomlin erred when going for two after the Steelers’ second touchdown. Did he kill his team’s momentum? Did he unnecessarily chase the first missed point the rest of the game?

There is no math equation for coaching from the gut. The Steelers led,12-3, and could have gone up by 10 by kicking the traditional one-point attempt. When a pass on the second attempt fell incomplete and the lead was nine, and the miss played a role in the crazy ending of the game.

If the Steelers opted for one point and a 13-10 lead, and the game had played out with the traditional one-point attempts for both teams the rest of the game, they would have led by three, not one, when defending against the Cowboys on the final drive. And in theory, it would have meant playing a more conservative defense with nine seconds remaining when Ezekiel Elliott raced 32 yards for the winning score.

“My argument is it does make sense,” said Kurt Bullard, a member of the Sports Analysis Collective at Harvard University. “Being up 9 or 10 late in the first quarter, I don’t think that changes the game plan going forward. Before Sunday, it was crazy. They had made something like 14 of 17. Obviously, they’re very good at it. After you fail the first time you still look at the matchups, and the Cowboys defense is average or slightly below average. You feel good about the matchups.

<!–

addResponsivePlayer(‘1r2c450wc8d581ocopccqn080t’, ”, ”, ‘perf1r2c450wc8d581ocopccqn080t’, ‘eplayer41’, {age:1431963427000});–>

“What I admire about that decision is the Steelers entered that game as a slight underdog. It was smart and prescient for Tomlin not to change the game plan. I think what he was acknowledging was they needed all the points they could get in the game.”

Arvind Pendurthi of Carnegie Mellon University Tartan Sports Analytics has a similar outlook on Tomlin and his two-point decisions. He cites the Steelers’ high success rate and the expected points model as the main reasons to support Tomlin.

But what do a couple of college seniors know about the NFL?

Plenty of NFL teams, including the Steelers, have hired young people with sports analytics backgrounds into their front offices in recent years. Karim Kassam is in his second year as the Steelers’ analytics and football research coordinator. Previously, he was senior director of quantitative analytics at Legendary Entertainment and a professor in the department of social and decision sciences at CMU.

The Cleveland Browns, the Steelers opponent Sunday, hired several people to their front office to help their rebuild. Paul DePodesta is a Harvard graduate whose previous experience in football came as a player at Harvard two decades ago. He spent the first 20 years of his career working in Major League Baseball. Several other Browns executives have backgrounds in analytics.

That doesn’t mean Tomlin and the Steelers can’t be questioned when it comes to two-point conversions. Bullard and Pendurthi pointed out how the Steelers almost never run on two-point tries, which gives the defense an advantage if they know they only have to defend the pass.

“If there is something to question it’s that all four attempts were pass attempts,” Pendurthi said. “Why not try to factor Le’Veon Bell into it?”

Pendurthi also is of the belief the 54 percent conversion rate will dip because teams will start game planning more for two-point conversions. Last season was the first time the NFL moved the one-point kick back to a 33-yard attempt.

Pendurthi interned at the NFL Network last summer and hopes to work for a professional football or basketball team once he graduates. Bullard also has aspirations to someday use his analytics background in sports.

The final question for two aspiring analysts was why does Tomlin continue to be mostly alone when it comes to going for two. Four-time Super Bowl winning coach Bill Belichick, for instance, hasn’t gone for two once since the 2014 season, and he only tried it once that season. 

Bullard theorized that’s because the Patriots have been so dominant over the years that they don’t need to go for two points very often. For the other coaches who don’t follow the math, Bullard said it’s simply a matter of job security.

“I think it’s the status quo thing,” he said. “If you’re a coach who is worried about his job you’re going to be conservative as possible. The Steelers are a team that likes to keep coaches for a long time. Other coaches almost anywhere else don’t have that kind of job security. It’s all about minimizing risk and defending the decision. You saw the Steelers missed some, and it’s a story. If you kick the PAT there it’s what you expect. Tomlin has more leeway.”

Ray Fittipaldo: rfittipaldo@post-gazette.com and Twitter @rayfitt1.