Photo by Randy Hampton
Nebraska Football

Hot Reads: Osborne Shoulda Gone for Two . . . Earlier

August 7, 2017
7,448

Here’s a good story from Ralph Russo on the growing role of analytics in college football. It’s the sort of trend story you see once or twice every year until the trend becomes the norm and then it’s time to move onto the next trend story.

But that hasn’t happened yet with statistical analysis at the the college level, so here we are again charting its progress. Russo’s story mentions two companies that are helping lead the way. Zcruit, a company I’ve highlighted before in Hot Reads, is trying to use data to help coaches predict which recruits are most likely to commit to which schools. That’s an interesting idea and an almost totally untapped market.

The second company mentioned in the story is probably more along the lines of what people think about when they think of coaches using analytics.

When it comes to in-game strategy, a six-year-old company named Championship Analytics, Inc., is quickly making a mark. CAI has gone from three schools subscribing to its service in 2014 to 53 this year, including 38 FBS teams.

Using a patented system of statistical analysis, CAI provides its clients each week with a game book, a three-ring binder stuffed with pages of color-coded charts and a by-the-numbers breakdown of the matchup. Taking strengths and weaknesses of each team into account, the game book lays out possible scenarios and gives strategic recommendations based on which option provides the best odds of winning.


I wasn’t familiar with CAI so I went over to its website. You can get a look at what those “game books” look like. If you click on the “WHY CAI” section, you’ll find a collection of stories from coaches and programs touting the advantages offered by CAI.

You’ll also find a short case study on the 1984 Orange Bowl. Everyone remembers Tom Osborne’s decision to go for 2 at the end of the game. Nebraska rallied from down 14 points in the fourth quarter to get to that deciding play and CAI’s analysis agrees with the decision to go for 2, but CAI says it should’ve happened on the first touchdown that quarter, a 1-yard run from Jeff Smith with just under 7 minutes remaining that made it 31-23 pending the point after attempt. CAI’s recommendation: Go for two now and, if successful (and assuming the game plays out the same), kick an extra point on the second touchdown for the win (emphasis mine):

This is the optimal strategy. This strategy gives Nebraska the same chance of winning as any other strategy while leaving the tie in play. If Nebraska had tied the game at 31 to 31 with a two point conversion after being down by 14 points, and had made the prior attempt at winning the game, it is almost certain that Nebraska would have won the national championship.

Taking into account all of the possible outcomes, the CAI simulator concluded that the odds of winning the national championship in a manner consistent with the team’s principles would have increased 5% by deploying this new strategy. With the CAI Game Book, these decisions would have been pre-programmed for the team with a chart showing the recommended strategy. Utilizing such an approach could have changed the outcome of one of the greatest football games ever played.


Nebraska did the opposite, kicking the extra point to make it a 7-point deficit, then going for 2 the second time and the rest was history.

Now, before you call this slander it’s worth noting that Osborne is listed as a CAI “advisor and shareholder.” You can hear him talk about his approach to analytics in a CAI webinar here. And you can hear him talk about the 1984 Orange Bowl decision knowing CAI’s recommendation here.

It’s an interesting idea to consider for a few reasons but let’s just tackle the biggest one: If Osborne employs CAI’s optimal strategy it changes everything.

Osborne’s decision to go for 2 is the defining moment of his career. Maybe it shouldn’t be, but history isn’t particularly fair when it comes to how people are remembered and with that one choice Osborne became the coach who put a national championship on the line out of principle, which isn’t a bad way to be remembered by any means. But let’s give that up for the other approach.

Say Nebraska goes for 2 on the first touchdown. If we’re still writing trend stories about how analytics are still only gaining ground in the uphill battle against conventional wisdom in 2017, it’s likely that almost nobody understands this decision in 1984. People could’ve done the math of course — the Huskers are going for 2 here in hopes an extra point will win it later — but it would’ve seemed bizarre at best.

Assume the Huskers get that 2-point conversion and then go on to win the Orange Bowl with an extra point in the final minute. That outcome is still based on the parameter that Osborne was unwilling to kick two extra points, tie and take his chances with the voters, but because the reveal of that decision happens with 6:55 left in the game rather than in the drama of the final minute — when the stakes and decision were clear to everyone — is it what people remember? Or does the narrative become about how Osborne was two steps ahead, how he made a decision almost no coach would’ve made with nearly 7 minutes left? Does the story become essentially about Osborne’s brilliant decision (rather than the principled one)? I think it might have and heroes with an intricate knowledge of the odds are less fondly remembered than heroes with an unshakeable set of principles.

But the other outcome might be even more interesting. Nebraska goes for 2 on the first touchdown and doesn’t get it. That leaves the Huskers down 8 and thus there’s no decision to be made on the second touchdown. If the Huskers convert the 2-point try there the game probably ends in a tie and Osborne’s first national title is discounted forevermore. He can’t walk into the postgame press conference, for example, and say, “Well, I knew that our odds of winning the national title were 5-percent better if we went for 2 right away, so that was the basis of that decision.” Nobody wants to hear that. It would have been defensible statistically, but disappointing emotionally.

And that’s still the tension with analytics in sports, which are as popular as they are because sports are really a hero-building mechanism.

To put it another way, CAI’s optimal strategy for the 1984 Orange Bowl would’ve been slightly better for Osborne’s title odds that year but almost certainly worse for his legacy.

The Grab Bag


Today's Song of Today


 
×
Verify your student status
See Subscription Benefits
Trial only available to users who have never subscribed or participated in a previous trial.