Ohio State’s passing game improved in 2017, but third downs were still a challenge

Since we’ve started the long offseason, it’s time to look back on the 2017 season and start thinking about what the 2018 team might look like.

I wanted to start with a look at the Ohio State passing game. The Buckeyes have had a remarkable run since the Barrett era started at Ohio State, but the passing game has been somewhat of a question mark since the beginning of the 2015 season. And really, besides the 2014 National Championship run, the passing game has been a question since Troy Smith left Columbus.

But, the Buckeyes showed remarkable improvement in 2017, especially compared to the last two years:

This chart shows passing S&P+ (the rating, not ranking), as well as the Buckeyes’ S&P+ rating on passing downs. These are both opponent-adjusted stats, with passing downs counting as, “Second-and-8 or more, third-and-5 or more, or fourth-and-5 or more. These are downs in which passing is easily the most likely option for gaining the necessary yardage, and defenses hold the upper hand. Offenses typically throw about two-thirds of the time on passing downs.” (full definitions here).

Ohio State’s passing S&P+ rankings during these four years: 2nd, 26th, 64th, 4th.

So as you can see, this year saw a big recovery for the Ohio State passing game, almost equal to the Buckeyes National Championship season.

The Buckeyes’ passing downs S&P+ varied nearly proportionally with overall passing S&P+ during this time. This is significant, because it indicates that the offense was successful this season (and in 2014) even when opposing defenses knew that a pass was likely. Put another way, the high overall passing S&P+ ranking might not only be due to successful passing on early downs.

But it is worth noting that Ohio State ran the ball on 26.9% of passing downs, which may not seem like a high percentage, but was 36th-highest in the country (because they had a quarterback who was nearly always reliable for at least four yards a carry).

If we zoom in on just third downs, Ohio State had a total of 185 third down attempts this season, converting 86 (46.5%). J.T. Barrett passed on 73 of those attempts, or about 39.5%, and completed 41 (56.2%). Only 61 of those passing attempts were from 4+ yards, and he completed 34 of those attempts, or 55.7%, and got the first down on 20 (32.8%).

I also looked at J.T. in comparison to other quarterbacks on third down with between 4-7 yards to go. I thought this was an interesting down-and-distance because either running or throwing could be viable options. I took the top-100 quarterbacks in terms of total passing yards in 3rd-and-medium situations, filtered out all quarterbacks who had fewer than 20 attempts in those situations — leaving 81 quarterbacks. Barrett was tied for 28th in 1st down passing success rate in third-and-medium situations, converting exactly half of those attempts (right behind Sam Darnold, who completed 51.2%). Also interesting — Trace McSorley was second overall, getting the first down on exactly two-thirds of attempts.

So what about third-and-long? J.T. isn’t among the top-100 quarterbacks in terms of total passing yards in third-and-long (7-9) situations, with 86 yards on 15 attempts (averaging 5.73 yards per attempt) and getting three total first downs (20%). For comparison, using the same method as above, Trace McSorley was the best in the country at picking up first downs, again getting them on two-thirds of attempts. The rest of the top-10 quarterbacks at converting third-and-long were names you’d expect: Mason Rudolph (No. 2), McKenzie Milton (No. 3), Baker Mayfield (No. 12), etc. Alex Hornibrook, interestingly, was 9th, converting 48.5% of his third-and-long attempts.

I think that may have been the real issue for the offense. When Ohio State did decide to throw the ball in obvious, late-down passing situations, they were overwhelmingly unsuccessful. Third-and-long is always a challenge, and it doesn’t help that the Ohio State offensive line struggled in obvious passing situations (allowing a sack on 9.5% of passing downs, which ranks 92nd), but given how effective the passing game was overall — again, 4th in the country after factoring in opposing defenses — that was a glaring weakness.

And if you take away the running back as a viable ball carrier on standard downs (i.e., forcing Barrett to keep on reads) then the offense can get behind the chains quickly. USC did this extremely well in the Cotton Bowl: Dobbins averaged 3 yards per carry with a 23% opportunity rate, while J.T.’s opportunity rate (which is the percentage of 5+ yard runs) was 165% of Dobbins’. Once the running back is slowed on standard downs, then the only effective play calls were, broadly, throwing on standard downs and quarterback runs.

So can Haskins, Burrow, or Martell improve on those third-and-long numbers next season? We don’t have a lot of data, but Haskins had 5 attempts on third-and-long last season, converting four, and 6 attempts on third-and-medium, converting three. But the interesting thing to me was that every one of Haskins’ third down completions was for a first down — meaning that he didn’t often throw behind the sticks. That’s an encouraging sign, even if it’s on extremely limited data.