clock menu more-arrow no yes mobile

Filed under:

Are analytics for losers too?

The Wildcats are simultaneously one of the most forward-thinking yet analytically regressive teams in college football.

NCAA Football: Northwestern at Wisconsin Jeff Hanisch-USA TODAY Sports

Down 24-9, Northwestern made a bold decision, to attempt a two-point conversion rather than the conventional wisdom of kicking the extra point. After failing on their first attempt, the Wildcats doubled down, rather than opting to put off a needed two-point conversion to tie the game.

However, while those decisions may be forward-thinking, the majority of Northwestern offensive play calling could not be less so. It’s time for us to break down the analytical anomaly that was Northwestern football’s game plan against Wisconsin. Let’s start with an explanation of the most controversial decisions of the afternoon:

Why it makes sense that Northwestern would go for two

Down 15 points, the conventional football wisdom states to try the extra point. Northwestern kicker Charlie Kuhbander has a 98.7% conversion rate on PATs as a Wildcat. However, advanced statistics show that it makes sense to follow Pat Fitzgerald’s strategy of attempting a two-point conversion, especially if you know you do not want to play for overtime.

After scoring a touchdown to go from being down 21 to being down 15, the game is technically a two-possession game, yet it is actually closer to a 2.5 possession game, because the success rate of a two-point conversion is slightly below 50%. As depicted in the graph below (while slightly outdated), the expected point value (average points per try for both) shows that the success rate of two-point conversions in college football typically hovers between 40 and 50%.

This chart displays points per attempt for conversions after touchdowns from 1958 to 2016

There are many outside factors to add to the equation (such as the fact that the Northwestern offense has proven to be significantly below average thus far, the Wisconsin defense is above average, and home field advantage). But if you base the success rate of two-point tries at 40% (slightly below the actual average for 2016 but closer than 50%), with all other factors standardized, the likelihood of making at least one two-point conversion from two attempts is 64% (given that if you make the first try, you don’t attempt a second), while the chance of missing both are around 36%.

This means that, assuming (as ended up being the case) Northwestern scores another touchdown, after that second touchdown, going for two down 15 gives you a 39% chance of being down six (making the first, kicking on the second), a 24% chance of being down seven (missing the first, making the second), and a 37% chance of being down nine (missing both). Keep in mind that this is with a generally uncharitable estimate of how often a team would make the two-point conversion.

If you do not go for two when down 15, using the NCAA average of a 97% success rate on PATs, you have a 94% chance of being down seven after two touchdowns, and approximately a 6% chance of being down eight.

Further assuming that Fitzgerald’s plan was to go for the win in regulation (especially on the road), keeping that same 40% two point success rate, in the first scenario a touchdown after the first two two-point decisions would give you a 49% chance of taking the lead (either making the conversion after being down seven or kicking the extra point down six). In the second scenario, said chance would be just 38% (making the conversion while scoring after being down seven).

Even if, due to being the road team and struggling offensively, you downgrade Northwestern’s chances of making their conversions all the way to 30%, or you upgrading Kuhbander’s kicking ability on the PATs to 100%, or both, the math holds. If you want to win in regulation after scoring a touchdown down 21, your best hope (though every victorious scenario would be extremely unlikely) is to immediately go for two.

Another reason for trying the conversion is that it is probable that Pat Fitzgerald would have attempted at least one two-point conversion en route to a comeback, rather than take his chances with the 123rd-ranked offense in the country against a superior team with a home-field advantage in overtime. By trying it earlier, it gave Fitzgerald the ability to get multiple bites at the two point conversion apple.

Either way you look at it, though it flies in the face of conventional wisdom, Fitzgerald and his staff’s two-point math makes sense. Some of the other parts of his squad’s offensive game plan, however, most certainly do not.

How Northwestern remains one of the most regressive teams in college football

While the two point conversion play-calling against Wisconsin may have been forward-thinking, offensive coordinator Mick McCall’s game plan against the Badgers was one of the most regressive in college football. Northwestern has averaged 3.6 yards per carry so far this season. But yet the Wildcats opened the game by running on first down in eight of their first eleven drives.

Excluding the 4th quarter (which saw a substantial difference in play-calling), Northwestern ran the ball on 11 out of 18 first down plays. The average of the first down gains were 6.54 yards per carry. However, once you eliminate the 31 yard run by Drake Anderson and Hunter Johnson’s 25 yard scramble, the average falls to an anemic 1.77 yards. Also, when the Wildcats ran the ball to begin drives, which they did on six of their first eight attempted treks downfield, they gained just 0.83 yards per carry.

Crucially, of the aforementioned 11 first down runs, eight possessions included a run on second down. The average 2nd down yards gained in a run-run combination was a horrendous 1.125 yards.

Eight of the 11 drives that began with a first down run ended with a punt (with a ninth ending in a field goal). Six of the eight run-run drives followed the convention of run-run-pass. Of those six, four ended with punts, and only one, which resulted in a field goal, included even a single first down. Both times Northwestern ran the ball three times, they punted.

In the game as a whole, Northwestern passed on first down 21 times against Wisconsin. 62% of the first down passing series ended up with first-down conversions. The average first down gain per pass attempt was approximately 4.4 yards, and the average yards per catch on a first down pass was 8.8 yards. The other attempts included two fumbles, four punts, one turnover on downs and one interception.

The first down play-calling, though it was extremely predictable and unsuccessful at the very beginning of drives, was actually varied enough to portend at least some success. The real thing that crushed Northwestern’s offensive scheme was the non-stop back-to-back runs.

Of course, occasionally any offense needs to mix in a second straight run at times to keep the defense guessing. But as shown above, this wasn’t a case of just changing things up every so often. The Wildcats went back to the well repeatedly with regards to starting a series with two straight runs, and the Badgers made them pay.

Analytically, Northwestern’s decisions on two point conversions were actually defensible. But that aggressive strategy doesn’t jibe with their decisions to become extremely predictable offensively in the first half.

The stats show that when the Wildcats ran the ball repeatedly, Wisconsin stuffed them. But the seemingly forward-thinking coaching staff didn’t make an adjustment until far too late.