Charting the Tide, Defensive Review | Wisconsin Badgers

Bill Connelly invented all of this; check out his college football analytics blog, Football Study Hall.

But first, a brief word about running the ball at Wisconsin

Wisconsin is supposed to be a known quantity, right? Their football philosophy ever since Barry Alvarez took over in Madison has been giant offensive lineman and bruising running backs that grind the opposing defense into dust, after which a “game manager” throws it deep.

The first part of that statement didn’t apply last Saturday, as the Badgers picked up 40 total rushing yards on just 21 carries, a 1.9 YPC average; even removing sacks, they only picked up 64 yards on 18 carries for a Richardsonesque^[1] 3.6 YPC. Half of those yards came in the fourth quarter, when the Badgers got a little offensive boost against the Tide backups. Simply put, the Wisconsin run game was completely and totally shut down from the get-go.

1 | Too soon?

This was the worst rushing yardage performance for the Badgers since October of 2012, when Michigan State held Wisconsin to 19 yards on 37 carries, which is amazing. I’m not going to try and verify this, but I’m guessing it was snowing. Heavily.

Before we get into all of this, two things to keep in mind about these results. First, everything except for individual performance and special teams performance is restricted to non-garbage time only. Garbage time is defined as when the game is not within 28 points in the first quarter, 24 points in the second, 21 points in the third, or 16 points in the fourth. For this game, at the start of the fourth quarter the Tide were up 28-10, an 18 point margin. The Badgers never got closer than that during the quarter, so no plays from the fourth are reflected below.

Second, unlike last year I am not including sack yardage in the rushing numbers. The NCAA insists on counting these as running plays, when they are clearly designed passes that have gone awry. For the time being I’m not throwing them into the passing numbers either, since there was no pass thrown on the play. I’m not sure what I’m going to do with those plays moving forward, but for today at least they aren’t really represented aside from the individual performance chart.

A Humble Request

Figuring out how to present all of this data is challenging to say the least. I’m happy with what I’ve got below, but I am definitely soliciting comments and suggestions on how to improve it. One specific question — for the down and distance matrix and the opposing QB performance chart, would you prefer the tables as designed or a graphic similar to last season’s Map of Quarterbacking Excellence? There will be a poll to that effect at the bottom of the article — please vote!

Overall Defensive Performance

Personnel Breakdown
Group	Plays	Hz	SR	iPPP
11	20	41.7%	30.0%	1.2
21	15	31.3%	26.7%	0.8
12	4	8.3%	25.0%	1.4
22	3	6.3%	66.7%	1.4
30	3	6.3%	33.3%	0.4

NOTE: "Hz" refers to Frequency.

Observations

Wisconsin’s offense is awesome. A three back set (in this case, 30 personnel) is something they use everywhere, not just on the goal line. All told they went with this look on six occasions, though only three occurred outside of garbage time. The majority of their work was out of your standard one or two back, one or two tight end sets, with the second back typically a tight end aligned as an H-back. By far the most effective of these were the two tight end sets (12 and 22 personnel), which were both good for 1.4 iPPP in limited use. The 30 personnel look is a lot of fun, but like most everything associated with running the ball, it didn’t work out too well for Wisconsin on Saturday.

Formation / Playcall Breakdown
Call	Plays	Hz	SR	iPPP
Shotgun	21	43.8%	33.3%	1.2
Pistol	0	0.0%	0.0%	0.0
Under Center	27	56.3%	29.6%	0.9
No Huddle	0	0.0%	0.0%	0.0
Huddled	48	100.0%	31.3%	1.0
Play Action	9	18.8%	44.4%	1.0

Observations

Kind of a boring chart this week, as Wisconsin more or less split time with Joel Stave either in the shotgun or under center. The shotgun was the slightly better look, primarily because they didn’t run out of it much. Wisconsin huddled on every play^[2], so nothing to glean there. Play action was a bit more efficient than the offense was as a whole with a success rate up over 40%, but at the same level of effectiveness.

2 | Notably, they also didn’t use playcards. Ugh.

Quarter Breakdown
Metric	1st	2nd	3rd	4th
Plays	11	18	19	0
S. Rate	36.4%	44.4%	15.8%	0.0%
iPPP	0.6	1.3	0.9	0.0
Pass Hz	45.5%	64.7%	83.3%	0.0%
P. S. Rate	60.0%	63.6%	20.0%	0.0%
P. iPPP	0.7	1.4	0.9	0.0
Rush Hz	54.5%	35.3%	16.7%	0.0%
R. S. Rate	16.7%	16.7%	0.0%	0.0%
R. iPPP	0.4	0.3	0.0	0.0

Observations

I failed to mention in last week’s primer I was going to offer quarterly splits as a regular part of Charting the Tide, but that’s something you can expect to see going forward as well. Nothing to glean from the fourth quarter column this week; as noted, it was all garbage time. Wisconsin’s most successful quarter was the second, which is when they figured out running the ball wasn’t going to work and Stave really started getting into his impression of Stephen Garcia. Wisconsin went to the pass more and more as the game stretched on and the deficit got larger, which isn’t terribly surprising. The third quarter was particularly unkind, as the Tide clamped down both in the passing game (40+% drop in success rate, with a drop in iPPP compared to the second quarter) and especially in the run game, where the Badgers got nothing at all after halftime.

Down and Distance Matrix
Distance	Metric	Down
		1st	2nd	3rd	4th
Short (0-3 Yds)
	Plays	0	0	5	0
	S. Rate	0.0%	0.0%	80.0%	0.0%
	iPPP	0.0	0.0	0.7	0.0
Medium (4-6 Yds)
	Plays	0	5	4	0
	S. Rate	0.0%	0.0%	25.0%	0.0%
	iPPP	0.0	0.0	0.8	0.0
Long (7-10 Yds)
	Plays	19	9	2	0
	S. Rate	26.3%	44.4%	0.0%	0.0%
	iPPP	1.3	0.9	0.0	0.0
Very Long (10+ Yds)
	Plays	1	2	1	0
	S. Rate	0.0%	50.0%	0.0%	0.0%
	iPPP	0.0	1.7	0.0	0.0

Observations

Wisconsin had way too much success in third and short situations, picking up four of their five attempts in that scenario. By comparison, the 2014 Tide only allowed conversions about 54% of the time on third and short, which is significantly better. This will undoubtedly even out over the course of the season, but it’s worth noting regardless. The other seven times the Badgers found themselves on third down, it didn’t go nearly as well, only converting once: the third-and-goal from the six that gave the Badgers their first touchdown of the day.

Another result that jumps out is the iPPP on first and long, which is a suspiciously-high 1.3. However, this a perfect example of why you look at both success rate and iPPP together, because that success rate is putrid. When the Badgers managed to get a good play in on first they picked up pretty decent yardage, but overall value on the down was pretty low.

Front Seven Performance

Rush Splits by Down
Metric	All	1st	2nd	3rd
Attempts	15	10	3	2
Rush Hz	32.6%	52.6%	20.0%	16.7%
S. Rate	13.3%	10.0%	0.0%	50.0%
iPPP	0.0	0.4	0.0	0.3
Opp. Rate	13.3%	20.0%	0.0%	0.0%
Hlt. Yds. / Opp.	0.5	0.5	0.0	0.0
RBR	0.1	0.1	0.0	0.0
LY/Attempt	2.1	2.3	1.7	1.5

Observations

I recognize the front seven and secondary both play a role in defending runs and passes, but rush defense is primarily on the front seven and vice versa, so that’s how I’m splitting it up. You’ll get a better idea of the reason for the designation after tomorrow’s Offensive piece.

For the game, Wisconsin was successful on just 13.3% of their attempts, which is really, really bad. Despite opting to run on over half of their first down plays, the Badgers were only successful once — their first offensive play of the game, a Corey Clement run off right tackle for five yards. That and a five yard Dare Ogunbowale run in the second quarter on first and 18 were your long rushes for the day.^[3]

3 | Outside of garbage time, of course – there was the 25 yarder for Alex Erickson in the fourth.

As such, Wisconsin basically got nothing from their running backs — a grand total of one Highlight Yard for the game. The offensive line was less and less effective at getting a push as the down increased, which will probably be the norm moving forward.

Rush Splits by Distance
Metric	Short (0-3 Yds.)	Medium (4-6 Yds.)	Long (7-10 Yds.)	Very Long (10+ Yds.)
Attempts	2	2	10	1
Rush Hz	40.0%	22.2%	35.7%	25.0%
S. Rate	50.0%	0.0%	10.0%	0.0%
iPPP	0.3	0.0	0.4	0.0
Opp. Rate	0.0%	0.0%	10.0%	100.0%
Hlt. Yds. / Opp.	0.0	0.0	0.5	0.5
RBR	0.0	0.0	0.1	0.5
LY/Attempt	1.5	2.0	2.0	4.5

Observations

More of the same as the previous chart – a whole lot of nothing, and nothing for me to really add.

Rush Splits by Direction
Metric	Left End	Left Tackle	Middle	Right Tackle	Right End
Attempts	1	0	9	4	1
S. Rate	0.0%	0.0%	11.1%	25.0%	0.0%
iPPP	0.0	0.0	0.3	0.4	0.0
Opp. Rate	0.0%	0.0%	11.1%	25.0%	0.0%
Hlt. Yds. / Opp.	0.0	0.0	0.5	0.5	0.0
RBR	0.0	0.0	0.1	0.1	0.0
LY/Attempt	2.0	0.0	2.3	1.6	2.0

Observations

Notice how Wisconsin only ran outside of the left guard once. That suggests there was someone on the right side of the defense they wanted no part of — oddly enough, A’Shawn Robinson was your man at the right defensive end position. Funny how that works.

The Badgers had slightly better luck running up the middle and to the right, but only slightly. The line got the most push on the middle runs, but really there’s not much more to say here. This was a dominant rush defense performance for the Tide, plain and simple.

Secondary Performance

Pass Splits by Down
Metric	All	1st	2nd	3rd
Attempts	31	9	12	10
Pass Hz	67.4%	47.4%	80.0%	83.3%
S. Rate	41.9%	44.4%	41.7%	40.0%
iPPP	1.1	1.5	1.1	0.8

Observations

Throwing the ball was a better option than running for Wisconsin on Saturday, but not by much. Throws on first down were the Badgers’ most successful plays of the game, with a success rate in excess of 40% and an iPPP of 1.5. As the down increased so did the Badgers’ tendency to pass, but success rates and iPPP dropped accordingly.

Pass Splits by Distance
Metric	Short (0-3 Yds.)	Medium (4-6 Yds.)	Long (7-10 Yds.)	Very Long (10+ Yds.)
Attempts	3	7	18	3
Pass Hz	60.0%	77.8%	64.3%	75.0%
S. Rate	100.0%	14.3%	44.4%	33.3%
iPPP	0.8	0.8	1.2	1.7

Observations

Wisconsin picked up three first downs passing on third and short, and all three were passes to Ogunbowale out of the backfield. To me, those short routes to running backs were the one area the Tide defense fell a bit short on Saturday, although most of that can be addressed in film review and perhaps a personnel changeup^[4] or two. Note that passing plays in Very Long yardage will likely always have pretty high iPPPs, as you tend to pick up decent yardage on those throws.

4 | Much like last year’s opener Reuben Foster looked a bit overmatched in pass coverage to me.

Opponent QB Performance
Air Yards	Metric	Left	Middle	Right	Totals
		5	11	12	28
Behind L.O.S
	Comp. %	0/0 (0.0%)	0/0 (0.0%)	1/1 (100.0%)	1
	S. Rate	0.0%	0.0%	0.0%
	iPPP	0.0	0.0	0.0
0-5 Yards
	Comp. %	2/2 (100.0%)	5/6 (83.3%)	6/8 (75.0%)	16
	S. Rate	100.0%	33.3%	25.0%
	iPPP	0.6	0.9	0.4
6-10 Yards
	Comp. %	0/1 (0.0%)	1/2 (50.0%)	1/1 (100.0%)	4
	S. Rate	0.0%	50.0%	100.0%
	iPPP	0.0	0.8	1.2
11-15 Yards
	Comp. %	0/1 (0.0%)	1/1 (100.0%)	1/1 (100.0%)	3
	S. Rate	0.0%	100.0%	100.0%
	iPPP	0.0	1.4	1.0
15+ Yards
	Comp. %	1/1 (100.0%)	1/2 (50.0%)	1/1 (100.0%)	4
	S. Rate	100.0%	50.0%	100.0%
	iPPP	2.3	2.4	1.7

Observations

Stave completed three of his four attempts thrown in excess of 15 yards, which represented most of Wisconsin’s offensive production on the day, as evidenced by the iPPP numbers. In the first three quarters the Badgers accumulated 18.3 net equivalent points^[5] across all plays; 6.3 of those points came on the aforementioned three attempts. Like attempts on Very Long yardage, this isn’t entirely unexpected — if you complete a pass in excess of 15 yards, that’s a very valuable play. It’s still a bit surprising that so much of their significant production came on just those three plays, particularly considering how close one was to being broken up by Minkah Fitzpatrick.

5 | Hey, that’s suspiciously close to the number of points they scored for the game! Hmm…

The vast majority of Stave’s attempts were within five yards of the line of scrimmage, and they came with the high completion percentages you’d expect on those throws. They were not especially successful, however, with a 37.5% success rate at the distance. It’s hard to gauge much from the 6-15 yard distance, as each block only has an attempt or two to look at.

Individual Performance

Disruptive Plays - Game
Player	PBU/PDs	STFs	INTs	Sacks	FFs	BKs	Total
Allen, Jonathan	0	0	0	2	0	0	2
Fitzpatrick, Minkah	1	0	0	1	0	0	2
Tomlinson, Dalvin	2	0	0	0	0	0	2
Foster, Reuben	0	1	0	0	0	0	1
Jackson, Eddie	0	0	1	0	0	0	1

Observations

I generally try to validate my results for disruptive plays with RollTide.com’s play-by-play results, but I don’t always agree with what they have down. We lined up nicely on sacks and stuffs this week, but they credited multiple pass breakups where I saw none. Given that they routinely miscredit passing targets and other information, and the fact I can rewind and really pay attention to minutiae on the play as opposed to trying to do everything in real time, I’m comfortable with not always agreeing with their results.

To that end, we saw three sacks out of the Tide on Saturday, part of what appeared to be a pretty decent pass rush, at least by 2014’s standards. Jonathan Allen was too much for Wisconsin’s inexperienced line to handle, picking up two of those sacks among his three tackles on the evening. Fitzpatrick had a heck of a debut, picking up the Tide’s third sack on a well-executed corner blitz early in the fourth and adding a great pass breakup for good measure. Big Dalvin Tomlinson knocked down two passes at the line, Reuben Foster added a stuff, and Eddie Jackson accounted for the game’s lone turnover, an interception on Wisconsin’s last offensive play of the game. Cyrus Jones also picked up where he left off last season with a nice pass breakup amid consistently good coverage on the evening.

ROLL TIDE