David Beaty says analytics led to surprising 4th down decision
Imagine you’re the coach of the KU football team.
First off, congratulations on the gig! Now it's time to earn your paycheck.
You’re playing against rival K-State and, despite being a 24-point underdog, you're down only 11 points five minutes into the third quarter.
Your quarterback is Carter Stanley, who threw for 181 yards in the first half. That's an extremely welcome sign considering your team was shut out in the nine quarters preceding the game and only combined for 127 yards in the two games before it.
Your special teams unit has struggled, but a defense that was once a huge minus for the team has actually played quite well.
Your team has the ball at the K-State 41-yard line. It’s 4th-and-3. Make the call: Should you go for it or punt?
If you’re like me — or the KUsports.com staff — you go for it.
In that exact situation in its 30-20 loss to K-State on Saturday, KU decided to do the opposite.
Cole Moos came on to punt. He kicked a line drive right to D.J. Reed, a dynamic K-State return man who already had a 99-yard kick return for a touchdown on the day.
Reed returned the ball to the K-State 46-yard line, meaning KU actually lost five yards of field position on the decision. K-State marched down the field and kicked a field goal. It would end up being a difference-maker at the end of the game — but first let’s talk about the decision.
Beaty was asked, twice, about that sequence in his postgame news conference. Both times, he maintained KU was playing the numbers.
Here's the transcript:
Reporter: “Coach, you guys had been, especially early in the year, a lot more aggressive going for it on fourth down. I was just curious what went into that decision in the third quarter, 4th-and-3 from the 41?”
Beaty: “I think I said it a little bit earlier. There’s a lot of study that has been done throughout sport, particularly in football, obviously, on fourth down and the value of touchdowns. And there’s a lot of analytics that go into it and we stay pretty close to the analytics. We stay pretty close to it. There are certain things where you have to go with your feel and situationals, but that one there was an analytic call. It was the right call based on the analytics that we had.”
Reporter: “The analytics were to punt, 4th-and-3 on the 41?”
Beaty: “Like I said, what our analytics told us at that point was that was the right thing to do. And with our defense playing the way that we were playing we felt like we had a good chance to get the ball back.”
Assuming Beaty wasn't thinking about another sequence — or tripped up by the wording of the question — that's a surprising answer.
In 2014, the New York Times came out with a fourth down bot.
It analyzes fourth down decisions, weighing the expected number of points for the team in the case of a conversion and for the other team in the event that it fails, based off of 10 years of data. It then spits out the decision of whether or not to go for it on Twitter — but for our purposes, there’s also a chart that explains the decision-making.
Now, that data isn’t 100 percent applicable to college football, as it’s based on NFL teams, but it’s a good place to start. In fact, KU’s fourth-down decision making has lined up with it at several points this year.
After charting 80 of KU’s fourth down opportunities going into the K-State game, including plays that were wiped out or changed due to penalty, the bot would’ve advised KU to go for it 19 times. In actuality, KU went for it or attempted to do so on 12 of them, with the biggest difference being 4th-and-1s.
The bot recommends always going for it on 4th-and-1. KU punted on four such opportunities, each time holding the ball on either their own 34 or 35 yard-line, which makes sense. There's nothing to complain about there.
Still, it bodes well for KU that these numbers match up, and it shouldn’t be all that surprising given that offensive coordinator Doug Meacham said earlier in the year it was “kind of a buzzkill” to have to settle for field goals and punts.
“(KU's fourth down chart) is probably a little bit more aggressive than your norm — than what the norm is, I guess,” Meacham said back in September. “It’s just like if I coached baseball I’d probably steal a lot more than a lot of these guys do. I mean, make them make a play. You go for it on fourth down. Make them make a play. You punt or try a 48-yard field goal, that’s just kind of a buzzkill to me. Let’s just go.”
But even if KU’s analytics, for whatever reason, disagreed with the NYT bot on Saturday, given KU’s track record this year, things don't quite add up.
Prior to the K-State game, KU had five 4th-and-1s between the opponent’s 39 and 45 and went for it all five times. That shouldn’t come as any surprise, regardless of the situation.
Then there was the Ohio game, which featured perhaps the two most comparable instances.
Down 11-0 in the first quarter, KU faced 4th-and-4 from the Ohio 42. KU went for it, but Peyton Bender’s pass fell incomplete, resulting in a turnover on downs.
In the second quarter, down 25-7, KU faced 4th-and-3 from the Ohio 38. Again the Jayhawks went for it and were successful, eventually scoring on the drive off a nine-yard touchdown pass from Bender to Chase Harrell.
Entering the week, KU had converted on 11 of 18 fourth downs, ranking in the top-half of the nation. They had broke the huddle planning to go for it 19 times, with two situations being altered by penalty and an additional official attempt being added on a muffed punt.
“Everything we do is calculated. The coaches have a formula for that and it’s been great for us this year," said Stanley. "We’re definitely one of the best teams in the country on fourth down. And obviously as a quarterback, sure, I want to stay on the field.”
It's worth noting that Beaty said the team can override decisions based on feel, but in both cases against Ohio — and the one on Saturday — the NYT bot says to go.
Add to that the third-down call, and it sure seemed like KU was gearing up to go for it.
Facing 3rd-and-3 — again, from the 41 — Stanley lofted a pass down the left sideline deep for Steven Sims. Sims was able to find some separation after giving Reed a little nudge but couldn’t come down with the ball.
That type of call, an aggressive, deep-play in a short-yardage situation, would seem to align with what Meacham said earlier in the year.
“There’s a lot of math involved, a lot of data involved in when and when not to. It’s kind of like when you go to Vegas, you know the blackjack card, when to take a hit. We kind of have that for fourth downs,” he said. “They let me know. And it helps me because on third down I can maybe do something that is a little abstract, because I know I’ve got another one.”
That would certainly qualify as “a little abstract,” only there wasn’t “another one,” to follow.
It cost KU any chance of coming back at the end.
Leading by a touchdown, the Wildcats started their final offensive drive of the game at the KU 45. They were only able to gain 21 yards, but in doing so, they kicked a 41-yard field goal to go up 10 with 1:21 to play.
“That put it out of reach,” Beaty said. "We were just two scores out, and that made it difficult.”
It didn’t have to be that way.
Had KU gotten the first down and simply kicked a field goal, even if everything had worked out the same way, they would've at least had a shot to tie.
And who knows what might have happened?
If the ensuing kickoff after a score had pinned K-State back at the 25 instead of, say, the 46, could KU have kept the Wildcats from scoring?
If KU had gotten a touchdown on the drive, would that have put pressure on K-State that forced a critical mistake?
Might KU have thrown a pick-six on the next play, falling behind three scores and making this whole conversation pointless?
One thing is for sure: We'll never know.