Before the 2019-2020 college football season started, I watched a video made by Jon Bois of SB Nation about finding the saddest punt in the world. It’s an incredible video for more reasons than I can tell you in this post, but the part that stuck to my nerdy brain was when he talked about a report written by an economist about fourth down decision making.
The report was written by David Romer in 2002, and it’s findings are directly in contrast to the decisions that football coaches make. It shouldn’t surprise anyone that NFL teams are far more conservative than is good for them.
Basically, if it’s fourth down and 4 or less, you should be going for it. That’s pretty wild to even my brain and I accept mathematically justified analysis pretty easily. I’m not going to go that far looking back at this past season in the MAC, but I did evaluate the coaches’ decisions on 4th down this season. I wanted to see who made the best decisions, per the report written by Romer, and what each coach’s philosophy is on 4th downs.
Gathering a mountain of data and sifting through it
You can tell that we are in the middle of a pandemic. With no live sports or developing interesting storylines in 2020, I went game by game and logged every fourth down play by a MAC team. There were more than 1200. It took a second. I logged the date, opponent, home or away, the betting line, yard line, distance, score differential, time, and outcome of the play.
The next thing to do was to define an opportunity to be aggressive. The David Romer report and personal opinion mixed to get that definition. I think that fits best since his data is from the NFL which will skew some of what happens. The biggest change is in field goal kicking. In his analysis, the value of kicking a field goal starts at the 33-yard line.
That is not MAC football. I don’t want to pick on young men that try their damnedest to make every field goal they get to kick, but only half the teams made a 50+ yarder last season. Only two teams made multiple 50+ yard field goals and only nine teams tried one that long.
Defining an aggressive opportunity on Fourth Down
Now that the field goal tangent is over, let’s do what I meant to do 129 words ago. Once a team is past the 40-yard line and the distance is less than 5 yards to go, you have an opportunity to be aggressive. Behind your own 40 yard line, I’ll give a pass on a punt. Really the line is set that far back is if a coach is being that aggressive, I want to see it in the data.
I cut out the 4th quarter to limit garbage time or final drive decisions. Those are desperation decisions, not aggressive decisions. I’m hoping that was more overly restrictive than under but it’s hard to tell. There are probably some games that were put away in the third quarter and the score influenced the decision, but hopefully only a handful or two of plays.
Buffalo and Western Michigan going all the time
Based on the established definition of an aggressive situation, Western Michigan lead the way going for it on 77% of their opportunities. There are plenty of reasons for that. Future parts are going to go deeper into the numbers but on the surface, the Broncos and Buffalo were aggressive teams.
Buffalo went for it on 76% of their opportunities. The Bulls also converted the highest rate of their attempts. Did that this season with a backup quarterback. Not bad.
Frank Solich is as old school as old school gets. He at least had the luxury of kicking field goals in 9 of 19 opportunities. Still, how many of those drives could have ended in the endzone instead of field goals. When he decided to take the chance, the Bobcats were successful. It’s weird how that works out.
Chuck Martin. He really impressed me this year when he found a way to win the MAC. I was sure that it wouldn’t happen until he stopped getting nervous in close games. I thought maybe he was over the hump. Nope. It doesn’t matter since it paid off, but he was still very conservative this season. I looked a little deeper into Miami’s numbers, I’ll go into them more in Part 2, and “conservative” does not do it justice.
Ohio should be more aggressive, and so should everyone else
According to Mr. Romer, the teams would be best served on average to have gone for it in 100% of these situations. The teams should consider their matchup at the same time. Romer’s analysis is of the average NFL team and how they should behave. By definition, half the teams are worse than that.
That being said, the Bobcats had an outstanding offense, very capable of getting 5 yards on any given play. They averaged 5.5 yards per rush and 8.6 yards per pass attempt this season. They would’ve been far better off being far more aggressive.
Bowling Green is quite the anomaly. They were conservative this season, and possibly rightly. Despite choosing to go for it 9 times in aggressive situations this year, the Falcons converted only once. That is incredibly low for only considering 4th and shorts.
I’m willing to bet that Western was aggressive this year because their field goal team had their issues. A true freshman did their kicking and distance didn’t seem to be a factor in if he made it or not. I would bet that if Coach Lester had an automatic three points from closer than 40 yards, he would be less aggressive. It may have been to his benefit to have a struggling kicker.
Miami and Chuck Martin had a great kicker in contrast, he was drafted in the seventh round of the 2020 NFL draft. That could explain the low aggressiveness they showed, but they had a 71% conversion rate in those situations. A number that high tells me they had an advantage often and should have used it more.
Staying ahead of the chains
Every coach has “Stay ahead of the chains” in their bag of cliches. It is important. The fewer yards there are to gain the more likely the team is to gain it.
The chart above shows the distance to go in every 4th down this season. The chart looks like what is expected. A spike at 10 for drives that finished where they started. The most common fourth down distance is 1 yard to go. Fifteen or more yards to go is the second most common distance. Maybe teams should have more plays for that situation than broadcasters have led me to believe.
This is what a good offense looks like. Toledo consistently moves the ball forward. Their chart matches the progression that the All-MAC chart shows. The Rockets had as many fourth down and 4 or shorter decisions to make as 4th and 5 to 11. Right in the sweet spot.
Now, this is Akron. No real trend, and 4th and 10 is tied for the most common singular situation they found themselves in this season. It’s less than ideal. To Akron, their short yards to go situations are more valuable than Toledo’s. There are far fewer of them. When Akron has a chance to extend a drive, they should take it seriously.
Ohio’s chart is the most interesting to me. All options forward are about even. They are the only chart that looks that way. Can a bobcat walk backward? Based on the chart, no. They catch a holding penalty or sack occasionally, but otherwise it’s always forward.
In my day job, I’m an engineer. I have to trust analytics. My job wouldn’t be possible if I didn’t trust mathematical models to inform reality. I believe the report written by David Romer, and also believe ever MAC team should be more aggressive.
The only reason to not be as aggressive as the report suggests is to gain an advantage over a small sample size. On average, the advantage is to be aggressive. There is going to be an average conversion rate for any situation. For example, let’s pick 3 yards and shorter, first three quarters, any score, and ball on your own half of the fifty.
There were 105 of those situations. It was punted 76 times. It was converted 18 times on 24 attempts. The conversion rate on all 4th and 3 attempts was 63.5%. The field position informs the decision to go or not, but not the outcome in the given scenario. The penalty for failing is heavy for giving the ball to your opponent with a short field. The average still says go for it.
Ball State only had 7 of these situations all year. The Cardinals were 2 for 2 converting them. In value-added by the coaches’ decisions for Ball State, they avoided a big penalty by not turning the ball over to a short field. They kept two drives alive and at least controlled the ball. I can’t do the math to know if they beat the book with the mix of modest success and conservative decisions with no failures, but they are at least close. Small samples mess up playing the averages.
The only reasons to not follow the new book are the small sample sizes and being in situational disadvantages. Whenever football comes back, we’ll see who has done their homework.