Five minutes into the third quarter, hope was quickly draining away for Georgia Bulldogs fans. Georgia Tech led 17-0, its offense was all but owning the time of possession, and the ugly word "rout" was on the tip of countless tongues.
Then, at the 9:57 mark, Oscar Delp caught a pass in the end zone. It was 17-6. That’s when head coach Kirby Smart made a decision still being debated: he went for two.
The move puzzled fans, and all the more because it failed. Why not take the easy kick and make it a ten-point game. Afterwards, Smart said the decision was “by the book,” and offered no further details.
What book would that be?
“Analytics,” of course. The influence of advanced and specialized data has transformed baseball and is now threatening hidebound assumptions about football strategy.
UGASports reached out to one of the sports data consulting firms, Championship Analytics of Cumming, Georgia, for clarification. Michael McRoberts, president of CA, offered his insight.
UGASports: Hi, Michael. Georgia fans are fascinated and even a little freaked out by that two-point conversion call. Who called the Code Red? Was it your company?
McRoberts: Thanks for reaching out, Rob. I can't comment on relationships with individual teams, but we've been working with teams on game strategy for a decade now and do consult with a very high percentage of SEC teams and colleges nationally.
UGASports: But you can talk data and strategy, right?
McRoberts: I’m always happy to discuss analytical strategies in general terms—particularly because such choices aren’t often explained clearly. And that’s a disservice to coaches when they go outside the norm.
UGASports: Analytics challenge the conventional wisdom.
McRoberts: Correct. Take scenarios where you’re trailing in the second half, as in this case. Conventional wisdom tends to focus on what a team trailing needs to do to tie a game. The analytics approach wants teams to focus on what’s needed to take the lead and ultimately win the game—not just tie it.
UGASports: Interesting. Yet in the third quarter in Athens, as things worked out, a kick there would have ultimately won the game in overtime for UGA, 28-27. So why the call?
McRoberts: So, down 17, conventional wisdom says, aim for two touchdowns and a field goal. To pull that off in a quarter and a half would be phenomenal. However, it just ties the game, and you need something good even beyond that to take the lead. Two-point strategy is one way to get that extra boost for the win.
UGASports: So it’s the aggressive move, with a bigger difference to make up.
McRoberts: You’re down 17, fairly late, and you get six to pull within 11. Now consider a two-point play would position a team down nine, and a TD plus field goal wins instead of ties the game.
UGASports: But as we know, the attempt failed.
McRoberts: Yes. So we’re down 11 and we score a second touch. Now the difference is five points, and in the analytics approach, the team would try to go for two again.
UGASports: Smart goes the more conventional route then, however, after its next touchdown. (Note: Tech added a field goal.)
McRoberts: Right. But for the sake of discussion, let’s say you went for two after a second touchdown; at 17-14 you’re now down three—the same spot as if you kicked twice after those same touchdowns. So if we look at the down-17 two-point strategy as a whole, an advantage is gained if a team goes 1-1 on two-point plays. There’s a safety net to get back to neutral if we split on two attempts, and only if you go 0-2 on two-point plays would this strategy hurt you.
UGASports: So the difference is in a tie-mindset vs. a win-mindset, along with the time left and the flow of the game. Georgia wasn’t expecting to get the ball back many times.
McRoberts: Yes. There’s a multitude of ways a game can play out. We can’t just assume a team will pitch a shutout moving forward or that all kicks will be made. We have to account for any and all possibilities. Certainly the UGA-Tech finish was a wild one that went outside that mold.
UGASports: And the argument that Georgia later wins by one with a kick at 17-6?
McRoberts: It’s never as simple as that. If UGA had kicked, Tech almost certainly would’ve gone for two after their last touchdown to cover the point difference. Also, if they’re behind rather than tied on that final possession, they probably behave differently. There’s never certainty in how a game plays out. Our job is to consider all the potential outcomes, weigh probabilities, then offer a verdict on what’s most likely to win a game in a given situation.
UGASports: Are there other examples of this particular call?
McRoberts: Going for two when down 11 late is actually a common strategy. Just this past weekend, five college teams did just that in the second half. Oregon State and Colorado State converted to pull within 9, while Georgia, San Diego State, and Pitt missed. The Jaguars also did this in the NFL Sunday—they missed the first two-point play to stay down 11, but got the second one after their next touchdown to pull within three late.
UGASports: As I heard another commentator say, we’ve probably seen it a whole lot more than we realize. Thanks, Michael! Best of wishes crunching the numbers.