Advanced Stats Glossary
Week 2 Five Factors Box Scores
Team Advanced Stats Profiles
Greetings, friends and Frogs, and welcome to the first real, honest-to-God, we’re going to look at the numbers preview of the season. The sun has risen this morning, the advanced stats profiles are live, and we have the number 2 ranked team (in S&P+) coming in to town on Saturday. We’re in for a solid week.
Before jumping into the preview, I want to segue with a comment about using advanced stats, especially towards the beginning of the season. In baseball, when a pitcher has a good ERA to start the season., analysts and fans are quick to run and check his “peripherals”, the underlying stats that contextualize a performance. In this column, I will be doing the exact same thing - examining the peripherals of college football, searching on the margins for underlying trends, highlights, and explanations. We have plenty of quality “big picture” work elsewhere on the site, and I’ll cover some of it, by my aim for this space is to put the Frogs and their opponents underneath a microscope to see what we can discover about the hidden aspects of football. That will include some summary of and pontification on the statistics, along with some new and improving graphs about play tendencies, but will also come with a heavy dose of wariness towards any kind of metric as decisive. The more I use statistics, especially here as teams have played meaningless games against inferior opponents, the more I feel the need to contextualize them. I hope you’ll enjoy a granular look at the Frogs, and as always, feedback and questions in the comments are appreciated.
Last Week: TCU Horned Frogs (99.9% Post-Game Win Probability) 42, SMU Mustangs (0.1%) 12
Let’s start with charts!
What went well:
TCU dominated SMU, after a rough start, mostly because the Frogs kept putting themselves in opportunities to sustain drives and convert, even if they failed to do so in the first quarter. TCU averaged a starting field position of 37.2 yard-line and gave themselves 7 scoring opportunities (trips inside the opponent’s 40 yard line), even if TCU only averaged 4 points per trip, by continuing to move the ball well (55.2% success rate), TCU allowed themselves time to normalize and wear down SMU, and ran away with the game as a result.
The defense had themselves a great game as well - holding SMU to 3.36 yards per play (31.5% success rate), and only allowing two scoring opportunities - both in the first half. A takeaway on defense (and another one that Innis Gaines should’ve had in his hands) plus a special teams score took the game from a grind-it-out win to a clear victory.
Also encouraging was the downfield passing defense; that will come in handy this weekend, as I’ll discuss below, even if SMU didn’t rely on the play-action pass as often as Ohio State likes to. The defensive pressure took some time to develop over the course of the game, but three sacks and three passes defended indicate a team creating havoc in the short game.
What went poorly:
TCU’s tackling was sloppy on the first drive, hence that skewed histogram - allowing 51 yards on one play really throws off the distribution. When SMU ran the ball successfully, though, it was to the left side or the left middle. On the broadcast, the announcers mentioned a “known weakness” of TCU was the middle seams, but it doesn’t look like SMU was particularly effective at exploiting that weakness, if it even exists.
On offense, TCU really moved the ball well, which is to be expected in the presence of a clear talent advantage. What concerns me about their performance (aside from the slow start) was the lack of big plays - TCU had some big plays, for sure, in the passing game, and was able to rush for 13, 18, 14 yards here and there, but all of their passing big plays came from receiver talent. One could argue that wide receiver skill in isolation a feature, not a bug of the TCU offense; I’d counter with the fact that Robinson had 10 poor throws and was unable to connect downfield even once. Yes, let’s chalk some of that up to the rain, but in the context of the Southern game, we’re seeing a pattern where Shawn’s eagerness to make a throw downfield results in laser line drives; to really make the deep game a threat, Shawn will have to put some more touch on deep balls and let his receivers get under them.
Shawn stared down that interception. I saw it. You saw it. He can’t get away with that - as others have noted, his eyes have been a little sloppy all season. He either figures that out or has a very long Saturday. That’s all I’ll say about that.
All in all, similar to my predictions, I might add, a superiorly talented TCU initially struggled with the rain and a hyped up rival opponent, but when the Frogs got their feet underneath them, TCU performed at a high level.
TCU Horned Frogs (22nd in S&P+, 39th Offense, 22nd Defense) vs. Ohio State Buckeyes (2nd, 2nd, 15th)
The 2018 Ohio State Buckeyes, subject of offseason controversy, are nonetheless following in their predecessor’s footsteps as one the elite teams in college football. The Buckeyes sit at 84th nationally in returning production, but a seasoned offensive and defensive line and a quarterback who’s been in the system make that more appealing than it sounds. Ranked 2nd in S&P+, the Buckeyes have moved the ball methodically (62.4% success rate), looking almost bored in tune-up blowouts against Rutgers and Oregon State, and scoring every time they get into opposing territory (6.78 points per scoring opportunity). The Buckeye passing attack completes 82.1% of their passes, and their hyper-efficiency (10th in the nation) on the margins is daunting. The defense has held inferior offenses to success less than a quarter of the time (24%, fifth in the nation) and effectively shut down the rushing game - Ohio State’s defensive front ranks first for preventing rushing opportunities and fifth in stuff rate, stopping a rush at or behind the line 36.7% of the time. Their 3rd ranked Sack Rate is an additional testament to the quality of defensive line play the Buckeyes have enjoyed to start the season.
The New Look, Same-Old Buckeyes:
Last March, SB Nation’s own Ian Boyd explored the possibilities for the Buckeyes under new starting quarterback Dwayne Haskins:
Haskins at his best is not even as strong a runner as Barrett but at worst he’s still capable of pulling the ball on a zone read and doing some damage in the open field. But he brings a totally new dimension to the Buckeye offense with his arm, which is vastly superior to Barrett’s in terms of velocity and accuracy either on the move or throwing outside the hash marks.
Initially, Boyd’s prognosis holds up. In two games as the Buckeye starter, Haskins has run the ball only 4 times, a far cry from JT Barrett’s run-first approach the last few years. The Ohio State offense still operates the same, but their bread-and-butter has shifted with the emergence of Haskins and his passing abilities away from a QB run and towards a “pull you in and beat you over the top” kind of scheme. A staple of the Urban Meyer offense has been the jet sweep motion. The Buckeyes will run a motion on most plays, bringing a sweep across to add another dimension of confusion on a zone read, or creating an opportunity for a runner to get in space, to create a better matchup, or simply to see what a defense will do and then attack that adjustment. OSU runs a sweep slightly different from TCU. The Frogs’ jet sweep motion creates space and speed; often the Frogs will run a jet sweep motion and pass/pitch the ball wide, in a psudeo-run play to isolate a receiver on the edge. OSU focuses a little more on power and confusion in their approach.
In running the ball and adding horizontal motion, the Buckeyes, can catch linebackers being dishonest, and the added wrinkle of matchup switching via jet sweep yields itself towards a scary PA attack. Of the five non-garbage time drives in their game against Rutgers, Ohio State had one designed QB run, but threw three touchdowns, two of which were 35+ yard play-action strikes. The Haskins offense features the play-action deep ball following successful runs; twice against Rutgers the Buckeyes ran successfully and then scored deep on play action.
They did allow a sack, and struggled with some pressure (again, these are only non-garbage time plays), but on the whole imposed their will on Rutgers; one drop, making Haskins (and one drive of Tate Martell) 14/16 = 87.5 true completion percentage with three touchdowns. Not a bad game.
First Downs and Penalties Matter:
Rutgers gained 39 yards on 21 first downs all game. That’s 1.86 yards a first down, which resulted in them facing an average of 2nd and 6.95. Getting behind the sticks is how the Ohio State defense lures you in - first down success matters. The Frogs rank 18th in the nation in first downs coming on first or second down, and will have to keep that trend going if they want a win in Jerryworld. Additionally, Ohio State had 7 touchdown drives on Saturday - Rutgers had penalties on five of said drives. Twice (twice!) Rutgers had Ohio State stopped and committed a penalty on third down that kept a drive going. TCU’s penalty issues are well-documented. A key to success this weekend will be not shooting themselves in the foot.
They Punted on the Second Drive:
On that drive, Ohio State failed to complete a short pass on second down and found themselves facing third and five. A Rutgers sack got the defense off the field. The Ohio State offense is not perfect; to have any measure of success, though, teams need to prevent third down conversions and get off the field.
The Matchup: How will the Frogs Fare?
1. Success Rate: 61.8%, 5th / 28.9%, 14th
2. IsoPPP: 1.09, 90th / 1.17, 75th
3. Avg. Field Position: 38.7, 4th / 24.0, 10th
4. Pts. Per Scoring Opportunity: 4.47, 77th / 3.40, 32nd
5. Turnover Margin: +1, 43rd
1. Success Rate: 62.4%, 2nd / 24.0%, 5th
2. IsoPPP: 1.15, 69th / 1.35, 102nd
3. Avg. Field Position: 30.6, 59th / 26.0, 38th
4. Pts. Per Scoring Opportunity: 6.78, 2nd / 4.25, 65
5. Turnover Margin: +3, 16th
So far, TCU ranks 5th in offensive success rate, 90th in IsoPPP. The Frogs are successful at getting necessary yardage, lacking in that deep/over the top threat (as discussed above) or a break out runner. TCU is 3rd in Opportunity rate - that is, the percentage of runs where the offensive line provides at least five yards for a rusher to work with. In a season with offensive line as the largest question mark, it’s encouraging to see the hogs taking care of business up front - against arguably the best defensive line in college football, though, the running backs are going to have to find ways to break lose to keep the offense moving. That IsoPPP number will be especially important against Ohio State, who let up a few long plays against Oregon State.
The Frogs are 4th in field position - huge for the Ohio State game. Starting drives at, on average, the 38.7 yard line is a result of TCU’s defensive performance and the talented special teams crew. Against Ohio State, the floundering Rutgers averaged starting drives at the 21.7 yard line, and Ohio State has allowed opponents only a 26 yard line average start so far this season. Flipping the field will be crucial for TCU’s offensive success and rhythm against an elite Buckeye defense.
The Frog defense has played well in the face of lesser competition this season, but still has some areas to improve on; a 28.9% success rate allowed is great, but the Frogs rank 75th in defensive IsoPPP - both Southern and SMU converted big plays against the Frogs, a disconcerting trend.
What to expect:
We don’t know much about the real Ohio State; both of their games have been lopsided endeavors, and in the Rutgers game, by the middle of this first quarter Ohio State had packed it in, running only their base plays, imposing their will. Ohio State gave up 31 points to an awful Oregon State team. Yes, throw out week one results, but also, Oregon State! There are clear areas of weakness - small areas, but extant nonetheless.
The Frogs are young and talented and have a high ceiling, but Ohio State’s ruthless efficiency on offense and stingy defense (5th in success rate) is going to be hard to handle. If things go well for the Frogs, Shawn Robinson will have connected on a long pass early in the game, getting him comfortable and in a rhythm. TCU expects to go pound for pound against Ohio State, both in offensive firepower and defensive prowess, and if the Frogs can lock down third downs, expect a closer game than most think.
What Else I’m Watching:
Thursday/Friday: Hurricanes have scheduling all bungled up this weekend, but that 4:30 Boston College (13th in S&P+) at Wake Forest (53rd) game is one I’ll have recorded. It might be fun to watch Memphis (20th) blow off some steam against Georgia State (119) on Friday, as the Tigers lost a game in which they had a 97.8 post game win probability against Navy last week.
Saturday (early): I might be going Big 12 only this weekend, especially in the early slot. Oklahoma (4th) and Iowa State (76) will probably get ugly, but could start as a scrappy game. I’ll have Rutgers (101) and Kansas (105) pulled up - you can’t miss the worst Power Five matchup in history. (Kansas is favored! Against a P5 team! They’ll probably lose!). And if the early match-ups get away from you, Troy (86) at Nebraska (29) might be an actual game worth watching, but I’m biased about Troy and their beautiful throwback unis.
Saturday (mid-late/pre-post Frogs): Houston (37th) at Texas Tech (50), North Texas (56) at Arkansas (54), sure have those in the background, but you want to watch Boise State (5th!) at Oklahoma State (11th!). Preferably watch with an Oklahoma State alum. Cancel your plans, watch this game. This is the other marquee game. Post Frogs, I’d check in on Arizona State (34th) - SDSU (74th), because even though SDSU is bad, who knows when or if Herm Edwards will ever trip up in a bad game?
What are you watching? What did I miss? What are your thoughts about the Ohio State Offense and Defense?