Good morning and happy Week 14 of the college football season. In a regular year, we’d probably be watching conference championships this weekend, and then settle in for a hopefully snow-blanketed Army-Navy game in Week 15 to close out the regular season. Times were good (in fact, if you feel like kicking back and pretending it’s 2017 this weekend, this is your best bet).
This year, we still have a lot of football to play, which, while welcomed, feels weird.
Speaking of pretending it’s 2017, I signed up to be a Zoom College Gameday fan (like 15 weeks ago) expecting to be picked for the Ohio State-Penn State game. Instead, they’ve finally called my number this week for the much anticipated Penn State-Rutgers game. If you have sign ideas, send them over to firstname.lastname@example.org. I’m really trying to lean into the new heights of depression this season has brought, so do your worst.
This week, I thought we’d keep it light and look at an important stat and an important characteristic of great teams. I’m talking about success rate and consistency, respectively.
To refresh you, success rate is a way of determining, via a simple “yes” or “no”, whether a given play was successful. It’s yards gained but with context. So even if you gain 15 yards on a play, if it was 3rd and 20, then no dice. That’s going in the “no” column as an unsuccessful play.
Here’s how we break it down:
1st Downs - At least 50% of yards-to-go gained
2nd Downs - At least 70% of yards-to-go gained
3rd/4th Downs - 100% of yards-to-go gained (earn first down)
As we’ve mentioned in previous newsletters, success rate is a big determining factor in winning or losing games. Teams that are just 5-10% more successful in a game win 75% of the time, according to Bill Connelly.
So let’s take a look at the Big Ten and see how teams do on each down. Special shoutout to Parker (@statsowar) for helping me visualize this is a more digestible way.
Offensive Success Rates
A key to winning football games in general is being consistent regardless of the situation. You’ll see in the below graphs that the top teams are solid on every down, which ensures manageable 3rd and 4th downs. If you are successful on both 1st and 2nd down, you are looking at a 3rd and 1. If you are successful on just one of 1st or 2nd down (and don’t lose yards on the other), then you’ll have either a 3rd and 5 or 3rd and 3, both fairly manageable.
Ohio State’s offense is the most consistent and successful by far. It’s not really close. Beyond that, there are surely some stinkers, but mostly a lot of inconsistent teams in the middle: decent on some downs and weak on others.
You’ll notice that the 4th down success rates are noticeably higher across the board for every team. The reason for this is two-fold: small sample size and selection bias. There are a lot less offensive plays on 4th downs, which can lead to wider variance either positive or negative.
However, the main reason is that teams are only likely to attempt a 4th down if they think they can make it (or they’re desperate). So a lot of the 4th down attempts are on 4th and 1 or 2 yards. Impressively, every team is in the high 60s or 70% range on their 4th down attempts so far this season.
Defensive Success Rates
On defense, I think the most standout team is Northwestern, who allow 36% or less success rate on the first three downs. A Northwestern defense vs. an Ohio State offense that has faced some adversity for parts of games would be an enticing Big Ten Championship matchup.
Some other notable defenses include Michigan St. (although it clearly hasn’t helped them much), Iowa, and, surprisingly, Indiana, who boast the #15 defense in the country in terms of EPA/Play allowed. I know I assumed their offense is what was winning them games, which doesn’t appear to be the full story. That’s another reason why I think Indiana may have a chance this weekend despite their QB being out (at least better than their 17% WP that +480 implies). Wisconsin are the best on paper, but I’m omitting them due to the small number of games played.
This week on the podcast, we talked about the Heisman race, and how 52% favorite Kyle Trask measures up against former Heisman winners as well as the guy hoping to best him in the SEC Championship in a few weeks, Mac Jones. We dive in to a few advanced stats, but for the most part it looks like the Heisman winner is usually determined by one simple number.
If you missed it, you can find all the ways to listen to it here.
I’ll be throwing more graphs up on Twitter over the next few days, and we’ll look at some other conferences as well, so be sure to follow @Staturdays and check them out!
Alright, enjoy your weekend, stay warm, and be sure to keep an eye out for me on College Gameday Saturday morning!