Chad Peltier
Guest
Advanced stats review: Big defensive plays fueled Ohio State’s comeback win
Chad Peltier via our friends at Land-Grant Holy Land
Visit their fantastic blog and read the full article (and so much more) here
It wasn’t pretty, and it wasn’t comfortable, but the Buckeyes emerged with one of the best wins of the season by any team
Along with LSU’s win over Auburn and Georgia’s win over South Carolina, Ohio State now has one of the best wins in the country through three weeks of college football action.
The Buckeyes’ 12-point margin of victory was similar to the Vegas spread and advanced stats projections, but the actual game didn’t play out as many expected. Maybe I’m projecting on the rest of the fanbase here, but I expected a 2-3 touchdown win, but with the Buckeyes maintaining a comfortable lead for most of the game; maybe TCU would score a garbage time touchdown to cut it down to two-ish scores.
Instead, Ohio State was down by 8 points until there were only seven minutes left in the third quarter. According to ESPN’s FPI, Ohio State’s win probability at that point was just 34 percent. After the big throw to Parris Campbell, the Dre’Mont Jones pick-six, and the blocked punt/touchdown, the win probability jumped to 93 percent in just ~5 minutes of game time.
That’s an insane five minute run for the Buckeyes in the third quarter.
So how do we contextualize this win? Let’s get to the advanced stats:
Stats definitions
Here’s the full advanced stats glossary.
The two final kneel down plays were filtered out.
How the game was decided
Big defensive plays (the Nick Bosa sack/fumble recovery touchdown and Jones’ pick-six) and special teams (punt block), which fueled the Buckeyes’ 21 points off of turnovers, were the big reasons why Ohio State won.
But we shouldn’t overlook the role of Ohio State’s poor red zone success rate (touchdowns on just two of five red zone trips), points per scoring opportunity (5.2, with three field goal attempts), and relatively low rushing success rate (49 percent), were all major factors that influenced the game’s closer-than-expected outcome and Ohio State’s eventual win.
Here were my big questions heading in to the game:
Can the offensive line handle TCU’s pressure?
It depends on what you look at here. TCU’s defense was 22nd in sack rate and 17th in overall havoc rate through the first two games, and ranked 15th in adjusted sack rate last season. So the offensive line holding that kind of aggressive group to just a single sack was impressive.
But Ohio State’s stuff rate — which is the percentage of runs stopped at or behind the line of scrimmage — was 21 percent, or 8 percentage points higher than what Ohio State’s defensive line held TCU’s run game to. It gets much worse if you include 1-yard runs; then the Buckeyes’ stuff rate jumps to an insane 41 percent. Their rushing success rate overall was solid — 49 percent against a great defense — but they ran for only one yard or less just as often as they had a 5-yard run.
Dobbins had longer runs than Weber, and averaged a full 3.1 yards per carry more, and he had a 56 percent rushing success rate to Weber’s 44 percent, but it was also seemed like the Buckeyes’ lack of a quarterback running threat and TCU’s aggressive defense contributed to the lower overall rushing success rate.
I also took a look at how Ohio State has performed in big games over the last year. I’ll update this table after each big game this season. Here’s rushing:
The rushing offense fell off of a cliff at the end of last season. Against Oklahoma and Penn State, the Buckeyes averaged a 52 percent rushing success rate and a similar opportunity rate. But, in the last three games of the season, the Buckeyes averaged just a 39 percent rushing success rate. Part of that was playing Michigan and Wisconsin, two excellent defenses, but part of it was also with defenses realizing that they could load the box, dare Barrett to pass, and force him to keep on reads (keeping the ball out of Dobbins’ hands or making him get tackled for short gains).
So putting TCU’s rushing performance in that context — 10 percentage points higher rushing success rate, but a higher stuff rate — says a lot. First, it means that TCU’s defense is incredibly aggressive, and they’re excellent at creating havoc (which we’ll get to later). It also means that Ohio State ran well against a solid TCU run defense, even if there were too many run stuffs.
Can Haskins spur a quick lead?
No. But it wasn’t really his fault, and he was excellent overall. Ohio State’s first half offensive drives were: field goal, punt, punt, punt, field goal, missed field goal. The Buckeyes play really well with the lead (it’s kind of a hallmark of a high-efficiency offense), but the offense didn’t have a quick lead to rest on this week.
Thank goodness for defensive scores, because the Buckeyes had two first half drives that stalled in the red zone (and three if you count the missed field goal drive at the end of the half).
Ohio State didn’t half a single successful play in the red zone in the first half on six attempts. Of those six plays, four were incomplete passes, and the two runs lost a yard and gained a yard. On the day, Ohio State scored touchdowns on just two of their five red zone trips. So flag “red zone success rate” as a key stat to watch against better defenses.
Here are the passing numbers in big games:
I was more interested in comparing the passing performances between Barrett and Haskins than the rushing success in big games. Obviously this is just 1.5 games worth of big game data for Haskins, but I think there are still some takeaways here.
Without charting out all of Robinson’s throws, the TCU passing game was every bit as effective as Ohio State’s, with a 48 percent overall success rate. That’s a good sign for the Horned Frogs and a credit to Robinson, but it also shows that things are not yet settled at a few defensive spots for the Buckeyes.
Can Pryor and Wade stabilize the pass defense?
The safety spot opposite Jordan Fuller remains an issue. My thought going into the game was that Shaun Wade might be the answer there, but his extensive playing time as the starting nickel back makes me think that the Buckeyes still need someone to emerge. Because while maybe Wade can play that position in three-linebacker sets, the defense needs another member of the secondary to emerge when Wade is at nickel.
And there’s also reason for concern with the three-man corner rotation. While Okudah and Sheffield looked relatively solid against TCU, Arnette was beaten on some notable plays.
Can TCU exploit linebacker inexperience for a solid run game?
Yes! Well, mostly. TCU had a 41 percent rushing success rate, which was a little lower than their 46.7 percent average last season, but significantly higher than Ohio State’s 34 percent allowed last year.
TCU found a lot of success running early on (and with screen passes), but the real issue was the 93-yard run with 9 minutes left in the second quarter. Not only was that Ohio State’s longest play allowed from scrimmage, but now Ohio State is one of only three teams to allow three plays of 70+ yards this year (they’re also one of two teams to allow more than one 80+ yard plays).
The linebackers seemed to be out of position frequently, still failing to maintain their gaps and not keeping contain. The marginal efficiency numbers haven’t been updated for this week yet, but it was notable to me that the linebacker havoc rates were by far the lowest of any unit on the defense — they just don’t seem to be around the ball as much as they should. Malik Harrison was the only linebacker in the top 7 tacklers and the unit as a whole had just one of the team’s 7 tackles for loss.
Can the Ohio State defensive line overwhelm a young TCU offensive line and create negative plays?
With seven total tackles for loss, a 16.7 percent havoc rate, and absolutely game-changing plays by Nick Bosa and Dre’Mont Jones — the answer is absolutely yes.
I was encouraged by what we saw from Jonathan Cooper, but should note that Chase Young didn’t record any tackles even if he did seem to be bringing a lot of pressure.
What about the hidden factors, like field position, turnovers, and scoring opportunities?
Obviously the defensive scores and punt block were absolutely critical to Ohio State getting the win, but field position overall was key.
Thanks in no small way to punter Drue Chrisman, eight of TCU’s 14 drives started inside their own 20, with three starting inside their own 10 yard line. That’s a tough position to be in, even if TCU still managed to use explosive plays to score on three 75+ yard drives.
Ohio State averaged starting on the 26.5 yard line, but it was just the 22.1 yard line if you exclude the drive after the punt block.
Scoring opportunity efficiency isn’t really hidden yardage like the others, and we’ve already mostly covered that point above — the Buckeyes could have done much better with their efficiency in the red zone and in scoring opportunities more broadly.
Moving forward
Just a couple of things I’ll be watching moving forward, moving from greatest to least concerns:
Continue reading...
Chad Peltier via our friends at Land-Grant Holy Land
Visit their fantastic blog and read the full article (and so much more) here
It wasn’t pretty, and it wasn’t comfortable, but the Buckeyes emerged with one of the best wins of the season by any team
Along with LSU’s win over Auburn and Georgia’s win over South Carolina, Ohio State now has one of the best wins in the country through three weeks of college football action.
The Buckeyes’ 12-point margin of victory was similar to the Vegas spread and advanced stats projections, but the actual game didn’t play out as many expected. Maybe I’m projecting on the rest of the fanbase here, but I expected a 2-3 touchdown win, but with the Buckeyes maintaining a comfortable lead for most of the game; maybe TCU would score a garbage time touchdown to cut it down to two-ish scores.
Instead, Ohio State was down by 8 points until there were only seven minutes left in the third quarter. According to ESPN’s FPI, Ohio State’s win probability at that point was just 34 percent. After the big throw to Parris Campbell, the Dre’Mont Jones pick-six, and the blocked punt/touchdown, the win probability jumped to 93 percent in just ~5 minutes of game time.
That’s an insane five minute run for the Buckeyes in the third quarter.
So how do we contextualize this win? Let’s get to the advanced stats:
Stats definitions
Here’s the full advanced stats glossary.
- In the tables below, points per trip scoring opportunity looks at the average points scored per scoring opportunity. Scoring opportunities are drives with a first down past the opponents’ 40-yard line.
- Drive efficiency looks at the percentage of drives that were scoring opportunities.
- Rushing opportunity rate is the percentage of runs that gained five or more yards.
- Rushing stuff rate is the percentage of runs that were for no gain or a loss.
- Explosive plays are those that gain 15 or more yards.
- Success rate: A common Football Outsiders tool used to measure efficiency by determining whether every play of a given game was successful or not. 50 percent of necessary yardage on first down, 70 percent on second down, and 100 percent on third and fourth down.
The two final kneel down plays were filtered out.
How the game was decided
Big defensive plays (the Nick Bosa sack/fumble recovery touchdown and Jones’ pick-six) and special teams (punt block), which fueled the Buckeyes’ 21 points off of turnovers, were the big reasons why Ohio State won.
But we shouldn’t overlook the role of Ohio State’s poor red zone success rate (touchdowns on just two of five red zone trips), points per scoring opportunity (5.2, with three field goal attempts), and relatively low rushing success rate (49 percent), were all major factors that influenced the game’s closer-than-expected outcome and Ohio State’s eventual win.
Here were my big questions heading in to the game:
Can the offensive line handle TCU’s pressure?
It depends on what you look at here. TCU’s defense was 22nd in sack rate and 17th in overall havoc rate through the first two games, and ranked 15th in adjusted sack rate last season. So the offensive line holding that kind of aggressive group to just a single sack was impressive.
But Ohio State’s stuff rate — which is the percentage of runs stopped at or behind the line of scrimmage — was 21 percent, or 8 percentage points higher than what Ohio State’s defensive line held TCU’s run game to. It gets much worse if you include 1-yard runs; then the Buckeyes’ stuff rate jumps to an insane 41 percent. Their rushing success rate overall was solid — 49 percent against a great defense — but they ran for only one yard or less just as often as they had a 5-yard run.
Dobbins had longer runs than Weber, and averaged a full 3.1 yards per carry more, and he had a 56 percent rushing success rate to Weber’s 44 percent, but it was also seemed like the Buckeyes’ lack of a quarterback running threat and TCU’s aggressive defense contributed to the lower overall rushing success rate.
I also took a look at how Ohio State has performed in big games over the last year. I’ll update this table after each big game this season. Here’s rushing:
The rushing offense fell off of a cliff at the end of last season. Against Oklahoma and Penn State, the Buckeyes averaged a 52 percent rushing success rate and a similar opportunity rate. But, in the last three games of the season, the Buckeyes averaged just a 39 percent rushing success rate. Part of that was playing Michigan and Wisconsin, two excellent defenses, but part of it was also with defenses realizing that they could load the box, dare Barrett to pass, and force him to keep on reads (keeping the ball out of Dobbins’ hands or making him get tackled for short gains).
So putting TCU’s rushing performance in that context — 10 percentage points higher rushing success rate, but a higher stuff rate — says a lot. First, it means that TCU’s defense is incredibly aggressive, and they’re excellent at creating havoc (which we’ll get to later). It also means that Ohio State ran well against a solid TCU run defense, even if there were too many run stuffs.
Can Haskins spur a quick lead?
No. But it wasn’t really his fault, and he was excellent overall. Ohio State’s first half offensive drives were: field goal, punt, punt, punt, field goal, missed field goal. The Buckeyes play really well with the lead (it’s kind of a hallmark of a high-efficiency offense), but the offense didn’t have a quick lead to rest on this week.
Thank goodness for defensive scores, because the Buckeyes had two first half drives that stalled in the red zone (and three if you count the missed field goal drive at the end of the half).
Ohio State didn’t half a single successful play in the red zone in the first half on six attempts. Of those six plays, four were incomplete passes, and the two runs lost a yard and gained a yard. On the day, Ohio State scored touchdowns on just two of their five red zone trips. So flag “red zone success rate” as a key stat to watch against better defenses.
Here are the passing numbers in big games:
I was more interested in comparing the passing performances between Barrett and Haskins than the rushing success in big games. Obviously this is just 1.5 games worth of big game data for Haskins, but I think there are still some takeaways here.
- First, TCU was excellent in pass defense last year, ranking 15th overall in passing S&P+ and 15th in adjusted sack rate. I’d guess they’re going to rank similarly this season, too.
- Second, the Buckeyes averaged a 34 percent passing success rate in six big games last season. Yikes. Outside of Barrett’s otherworldly comeback against Penn State, Haskins’ 47 percent passing success rate game is the best performance by an Ohio State quarterback since at least the beginning of last season.
- Third, while completion percentages varied (averaging 61 percent last season) and were similar to Haskins’ complete rate vs. TCU, he averaged an insane three full yards per attempt more than Barrett last season (with sack yardage deducted). This may be the biggest takeaway so far — while Barrett and Haskins completed a similar percentage of their passes, Haskins’ performance vs. TCU was far more successful because he averaged more per attempt. Haskins was more successful targeting receivers on intermediate and downfield throws.
Without charting out all of Robinson’s throws, the TCU passing game was every bit as effective as Ohio State’s, with a 48 percent overall success rate. That’s a good sign for the Horned Frogs and a credit to Robinson, but it also shows that things are not yet settled at a few defensive spots for the Buckeyes.
Can Pryor and Wade stabilize the pass defense?
The safety spot opposite Jordan Fuller remains an issue. My thought going into the game was that Shaun Wade might be the answer there, but his extensive playing time as the starting nickel back makes me think that the Buckeyes still need someone to emerge. Because while maybe Wade can play that position in three-linebacker sets, the defense needs another member of the secondary to emerge when Wade is at nickel.
And there’s also reason for concern with the three-man corner rotation. While Okudah and Sheffield looked relatively solid against TCU, Arnette was beaten on some notable plays.
Can TCU exploit linebacker inexperience for a solid run game?
Yes! Well, mostly. TCU had a 41 percent rushing success rate, which was a little lower than their 46.7 percent average last season, but significantly higher than Ohio State’s 34 percent allowed last year.
TCU found a lot of success running early on (and with screen passes), but the real issue was the 93-yard run with 9 minutes left in the second quarter. Not only was that Ohio State’s longest play allowed from scrimmage, but now Ohio State is one of only three teams to allow three plays of 70+ yards this year (they’re also one of two teams to allow more than one 80+ yard plays).
The linebackers seemed to be out of position frequently, still failing to maintain their gaps and not keeping contain. The marginal efficiency numbers haven’t been updated for this week yet, but it was notable to me that the linebacker havoc rates were by far the lowest of any unit on the defense — they just don’t seem to be around the ball as much as they should. Malik Harrison was the only linebacker in the top 7 tacklers and the unit as a whole had just one of the team’s 7 tackles for loss.
Can the Ohio State defensive line overwhelm a young TCU offensive line and create negative plays?
With seven total tackles for loss, a 16.7 percent havoc rate, and absolutely game-changing plays by Nick Bosa and Dre’Mont Jones — the answer is absolutely yes.
I was encouraged by what we saw from Jonathan Cooper, but should note that Chase Young didn’t record any tackles even if he did seem to be bringing a lot of pressure.
What about the hidden factors, like field position, turnovers, and scoring opportunities?
Obviously the defensive scores and punt block were absolutely critical to Ohio State getting the win, but field position overall was key.
Thanks in no small way to punter Drue Chrisman, eight of TCU’s 14 drives started inside their own 20, with three starting inside their own 10 yard line. That’s a tough position to be in, even if TCU still managed to use explosive plays to score on three 75+ yard drives.
Ohio State averaged starting on the 26.5 yard line, but it was just the 22.1 yard line if you exclude the drive after the punt block.
Scoring opportunity efficiency isn’t really hidden yardage like the others, and we’ve already mostly covered that point above — the Buckeyes could have done much better with their efficiency in the red zone and in scoring opportunities more broadly.
Moving forward
Just a couple of things I’ll be watching moving forward, moving from greatest to least concerns:
- The linebackers fulfilling their assignments. (Can we play Browning outside? Will more Borland help? Will Greg Schiano/Alex Grinch stealthily takeover as the linebackers coach?).
- The other safety spot.
- Taking advantage of scoring opportunities on offense / red zone play calling.
- Rushing efficiency, especially on non-read inside zone plays. (Will other opponents be similarly aggressive as TCU? Can the line be more dominant earlier in games?).
- Haskins’ continued development.
Continue reading...