clock menu more-arrow no yes

Filed under:

Ten charts that summarize Ohio State’s season and project 2020

Let’s dive into the analytics to close out the 2019 season and look ahead to next year

NCAA Football: College Football Playoff Semifinal-Ohio State vs Clemson Mark J. Rebilas-USA TODAY Sports

Even though we’re over two weeks removed, the loss to Clemson is just as painful. Throughout the entire 2019 season, the Buckeyes looked like not only one of the best teams in the country, but one of the better teams in the modern college football era. They beat top-25 SP+ Wisconsin (twice), Penn State, and Michigan, and were this close to facing LSU in the national championship game. If you’re like me, it hurts especially bad that it ended up being Clemson to knock off the Buckeyes.

But after some time to digest the loss and get some perspective on the season as a whole, it’s also worth noting that the 2020 Buckeyes could be at least as good as the 2019 team. Losing Chase Young, J.K. Dobbins, and Jeff Okudah undoubtedly hurts, but Ryan Day’s second team will be arguably more talented top-to-bottom, and Justin Fields will be in his second year as well.

So let’s take a look at some data that summarizes the Buckeyes 2019 season and looks ahead to what 2020 might be.


Season summary charts

This first chart shows all FBS teams’ offensive and defensive average expected points added (EPA) for the entire 2019 season (including bowl games). As you can see, for the entire season, Ohio State and Clemson were neck-and-neck, and the most complete teams overall. Importantly though, LSU’s defensive rating has jumped significantly as the season has gone on; LSU’s defense has improved significantly since September.

Next, we have every Ohio State offensive play of the 2019 season by its EPA rating:

Interestingly, the Buckeyes’ median play was rarely ever at or above zero (meaning that the Buckeyes’ median play was worse than expected given down, distance, and field position); but looking at mean EPA, the Buckeyes were third overall in the country and always averaged 0.02 EPA or above. For context, the Buckeyes’ two 0.02 mean EPA games, against Michigan State and Clemson, were still better than 54% of all FBS games this season.

Looking at the Buckeyes’ offensive performances like this, it’s easier to see that Ohio State simply didn’t have as many overwhelmingly explosive plays as in previous games — their EPA top quartile is significantly lower than in most other games this season.

And here’s the same chart for Ohio State’s defensive plays this season (lower EPA is better here). It’s almost the opposite story — there were just a few more really positive plays allowed than usual.

This next chart compares teams’ average EPA success rate by their mean EPA. The 95% confidence interval is shown around the linear regression line, indicating that Ohio State’s offense was about as explosive as you’d expect given their level of efficiency (via this metric):

And here’s the same chart but for Ohio State’s defense. The Buckeyes actually allowed a lower average EPA than you’d expect based on their EPA success rate, even though their defense allowed a second-best -0.27 mean EPA this season.


Offensive Explosiveness

These next few charts look specifically at Ohio State’s offensive explosiveness.

First we have the rate of successful plays (via EPA success rate) and the rate of explosive plays. EPA explosiveness rate is defined here as a play within the top 10% of EPA for all FBS plays this season. Just looking at rates alone, you can see that Ohio State had one of the best offenses in the country, and produced explosive plays about as often as you’d expect (the Buckeyes are right on the edge of the 95% confidence interval) given their high rate of successful plays.

Next we have the rate of explosive play creation by the team’s mean EPA only on explosive plays. Essentially this chart heavily isolates the magnitude of explosiveness so that you can more easily compare it with the rate a team is explosive.

As you can see, Ohio State has one of the highest EPA explosiveness rates this season, but it was only middle-of-the-pack in average EPA per explosive plays. In comparison, Clemson and Oklahoma created explosive plays a little less often, but their explosive plays were of a higher magnitude when they did. Alabama, absurdly, had both a high rate and magnitude of explosive plays.

There is almost no relationship between a team’s rate of explosive plays and their magnitude, but it is interesting to see that the Buckeyes average explosive play wasn’t quite as explosive as one from Oklahoma, Clemson, LSU, or Alabama, despite the fact that they’re producing them at a higher rate than nearly anyone else.

And a big reason for that explosiveness surge this year? J.K. Dobbins’ return to form:

This chart shows Dobbins’ mean rushing EPA for every meaningful game of his career. His sophomore slump was real, as he didn’t produce explosive plays at nearly the same rate as his freshman season, and the team’s overall stuff rate increased as well. But his 2019 performance was one of the best by any running back in OSU history, with only three games out of 14 with a negative average EPA (and two were just barely negative).


Looking forward: Talent in 2020

But as I mentioned above, the 2020 team could be slightly more talented top-to-bottom than even the 2019 team was. This first chart shows every recruit in Ohio State’s last four classes by their 247 Composite player rating.

The 2019 and 2020 classes (as the 2020 class stands now) were definitely a step down from the incredible 2017 and 2018 classes, but the 2020 class does appear to be somewhat of a rebound from the 2019 class, which is understandable given it was a coaching transition year. Ohio State’s median player was better, and they had more top-end players in the 2020 class as well. The 2019 and 2020 classes differ the most from the two earlier classes in that Day seems more willing to take on Ohio kids and high-potential but lower rated recruits.

Finally, this last chart does a four-year rolling average of average player ratings to give a sense for a team’s overall talent levels in any given season. I zoomed in to the top of the ratings, filtering out any team with lower than a 0.9 average player rating:

A couple of interesting things here:

  • Alabama remains at the top throughout the entire span, but in 2020 they will likely be joined by Georgia.
  • Ohio State is very slightly more talented on average in 2020. There’s a solid gap between Georgia/Alabama and the Buckeyes, but there’s almost an equal gap between them and Clemson, LSU, and Texas (and then everyone else).
  • Speaking of the two Tigers teams, both are on a remarkable four-year rise. Neither team had an average four-year player rating over 0.9 in 2017, but are now 4th and 5th in the country in 2020. That’s interesting because with slightly more talent parity, it’s possible that the playoff race is a little more wide open in 2020. Further, other factors, like experience, development, and coaching, are likely to matter even more. Notably, of those top five teams in 2020, only Ohio State and Clemson are returning their starting quarterbacks.