Calculated Risks From the Coach Who Never Punts

My brother has been hounding me for while now to look into the high school football coach who never punts. I finally did a deep dive by reading a few articles and watching some videos on Kevin Kelley, the head football coach at Pulaski Academy.

Kelley says he read a report from a Harvard professor who analyzed 2,000 football games over a three year period. His conclusion was that field position wasn’t as important as most coaches believe. The huge takeaway for Kelley here was that you should never punt the ball. This is the kind of thing that I have only seen people do when they’re playing Madden so I had to see the numbers on this to be convinced.

Kelley lays out an example, with evidence. Let’s say his team is backed up on their own 5 yard line and it’s 4th down. If they go for it and don’t make it the other team gets the ball and they’re going to score a touchdown 92% of the time. If they instead punt the ball the other team gets the ball at around the 40 yard line. From there the number only drops to 77% of the time that they score. So it’s not a huge difference in terms of scoring odds.

But if you factor in the 4th down conversion rate it makes even more sense to go for it every time. Pulaski says his team converts 50% of their 4th downs to keep drives alive. Now going for it every time doesn’t sound so insane. They don’t even have a punter on the team.

They also onside kick every time. Kelley says you can tell who wins the game 75%-80% of the time just by looking at the turnover battle. He’s just trying to force turnovers through onside kicks. On an average kick-off the other team would get the ball at the 33 yard line. With the average onside kick (when they don’t recover it) the other team would get the ball, on average, at the 47 yard line. So they’re only giving up 14 yards when they don’t recover the onside kick. Their onside kick recovery rates average around 20%, so they get the ball back one out of five times. This can really change the complexion of the game.

There is also a stat that the team with the most big plays — 20 yards or more — wins 80% of the time. So he’s structured his offense to make more big plays by utilizing laterals and throwing the ball deep.

Here are some takeaways from this story:

• Analytics aren’t perfect, but they’re necessary. There’s the old saying that not everything that counts can be counted and not everything that can be counted counts. But how can an individual or organization ever truly understand their performance if they’re not defining and measuring the impact of their decisions?
• Measurement can improve performance. Performance measurement does you no good if you’re not going to implement your findings in some way. Kelley said once he discovered this data he felt he would have been providing a disservice a coach had he not utilized these techniques in his program.
• Process over outcomes. This system isn’t foolproof. He’s looking for minor advantages and trying to improve his team’s chances for success. He estimates his game plan could improve their odds of success by 15% or so. That’s no guarantee. It’s a process over outcomes system that doesn’t work every time, but it works more often than not. That’s all you can ask for.
• Make sure you’re being compensated for the risk you take. The coach knows going for it on fourth down, doing onside kicks and trying for big plays isn’t going to work out every time. There are times when these risks will back-fire. He had to factor in the failure rate when making the decision to take these calculated risks.
• Going against conventional wisdom is not easy. Everyone else thinks this is crazy and gimmicky but Kelley thinks it’s just smart. He’s not trying to be a contrarian to try to stand out. He says, “There are those kinds of people that want to be different because they just want to be different. And there’s those kinds of people that want to be different because they’re looking for something to help them be more successful.”

So why don’t more coaches try Kelley’s methods? I’m shocked no small-time college programs have tried this yet. You would think it would help bring some attention and recruits to those programs that can’t keep up with the larger programs. There is of course the problem of career risk and the herd mentality. I’m reminded of the old Keynes quote where he said, “Worldly wisdom teaches us that it is better for reputation to fail conventionally than to succeed unconventionally.”

Here’s a video from the now-defunct site Grantland:

And here’s another one from 538:
The High School Football Coach Who Never Punts

Also, if you’re interested in sports analytics, check out the old Michael Lewis book Moneyball if you’ve never read it, which covers a similar topic in Major League Baseball. One of the better value investing books there is.