clock menu more-arrow no yes

Filed under:

Graphing the Tide vs. Western Kentucky

New, 22 comments

Saban wasn’t happy, but the data feel surprisingly okay about Saturday’s win over the WKU Hilltoppers.

NCAA Football: Western Kentucky at Alabama Marvin Gentry-USA TODAY Sports

Author’s note: the graphs will look a bit different than last week’s, as the data available have changed. As always, these will shift (and hopefully improve) as the season continues.

Saturday’s victory against Western Kentucky had a familiar story line: the Tide defense suffocates an opponent while the Alabama offense struggles through a first-half identify crisis: “Who am I? What does it all mean? Is a jet sweep actually a pass?”

Saban was notably grumpy after the game on Saturday, citing execution and “arrogance” as the main reasons, but the stats actually show a fairly average game. Maybe a fair expectation for playing a non-Power team is to be above average... but all in all, the first team offense made it work.

Metric definitions

  • A "successful" play, as defined by Football Outsiders, is basically when a play gains enough yardage to keep the offense on track, i.e., 50% of needed yardage on 1st down, 70% on 2nd, or 100% on 3rd/4th.
  • A "big play" (aka an "explosive play") has different definitions depending on the analyst, but I use a generous one: a run of ≥12 yards, or a pass of ≥16, is a big play.

Success rates, big play rates

Big play rate (XR) and Success rate (SR)

* NCAA average SR = 40%

Not seeing a chart here?

That 41% total Success Rate for the Tide is very average: it’s what you’d expect a mortal NCAA offense to do in a game. The pacing is interesting, though: three solid quarters brought down by one sore one (the 2nd). That quarter presents an unfortunate scapegoat in Blake Barnett: after a 3-and-out series with Hurts at the helm, Barnett was put in and asked to pass (or perhaps chose to pass on RPOs) on basically every snap he played in the 2nd. The Tide was also called for two untimely penalties on offense during the 2nd quarter, which certainly contributed to stalled drives.

As for big plays, Alabama came out of the gates with a fairly explosive 1st quarter, tallying some big passes to Ridley (x2), Stewart and O.J. Howard. Things fell off in the 2nd and 3rd quarter, though, and the Tide ended up slightly less explosive than the week prior against USC (11% to 9%) for the game. B.J. Emmons helped out with two 13-14yd runs in the 4th, which was a bright spot.

The defensive performance was encouraging across the board: the vaunted WKU offense we’d heard so much about didn’t manage a single quarter of average or above-average performance. BOOM! The Hilltoppers managed a modest surge during the second quarter—first on their running game, then on late-half passing—but didn’t manage to get any points from it (the blocked FG helped). Eddie Jackson’s pick-6 in that quarter likely had some effect on both offenses, keeping Bama’s off the field and WKU’s off their game. The Hilltoppers never achieved much efficiency through the second half, though they did rally off a few explosive plays during garbage time (scoring a touchdown on backup safety Trevon Diggs in the process). In the end, WKU’s 27% Success Rate looks a lot like USC’s from Week 1 (26%).

Running and Passing (#RTDB)

Run rate (runs ÷ total plays), cumulative

Success rate for runs and passes, cumulative

* NCAA average SR = 40%

Not seeing a chart here?

The running and passing charts have changed since last week: the percentages are still cumulative (as in, including everything up to that point), but they only update once per quarter, at the end of each quarter. We’re left with less of that fun “squiggle,” but we’ve still got a good view.

By the end of the game both running and passing had settled into a comfortable slightly-above-average efficiency, but they took very different journeys to get there. Lane Kiffin had Jalen Hurts slinging the ball early, and with some success (note the 1st quarter SR and XRs in the earlier chart). The play-calling followed to push more passing—especially from Barnett—but efficiency took a dip across the board. The passing SRs fell further during the second half, with several unsuccessful passes from Jalen Hurts (including dropped long passes from Dieter in the 3rd and O.J. Howard in the 4th), but fortunately the running game picked up the slack, with Hurts, Harris, and Scarbrough in the 3rd quarter, then B.J. Emmons in the 4th.

When all is said and done though, passing slightly outperformed running the ball. Which is something that we should talk about: there’s a day in every Gump’s life that he or she needs to learn about the birds, the bees... and passing the football. In some games (i.e., when the defense is stacking the box), it just works. Take what the defense gives you, and pass the ball. Last year’s Cotton Bowl versus Michigan State was a prime example of this: MSU largely shut down the run, so Jake Coker just passed. You probably remember how that game went. The result wasn’t so dramatic against WKU—and Alabama was likely throwing the ball more to press QB development—but think about it next time you find yourselves yearning for a 100% run rate. #RTDBLK.

BONUS: Run direction

Run Direction Plays Success Rate Big Play Rate
Left end 8 38%
Left tackle 4 25%
Left guard 1 100%
Middle 8 13%
QB draw 2 50%
RB draw 1 100%
Right guard 3 33% 33%
Right tackle 4 50% 25%
Right end 4 50%
Reverse 1 100%

You’ve been well-behaved this week, and everyone’s been talking about the offensive line struggles, so here’s a bonus look at Alabama’s run efficiency by direction. As you’ll see from the play counts, this is very little data (and there’s always gray area in this kind of directional data) so be mindful of your #takes.

That said, there are some interesting tidbits. For one, Bama likes running behind its start left tackle, Cam Robinson, which isn’t a surprise aside from the middling success rates you see there. Maybe defending linebackers expect the behavior and are that much quicker to the ball... or, Cam is having to work harder to make up for having newbie Lester Cotton at left guard. (Some of these stats would have been collected after some second-stringers were in, but lineman don’t show up in the data, so we’ll have to make some assumptions.)

The low success rates running up the middle are not encouraging. There are almost certainly some other factors here—short yardage attempts, abandoned pass plays, bad RPO reads—but it’s not hard to imagine that our shuffled, young interior line is struggling early. The draw plays are a bit different, and had good results, but the rest of this needs to be cleaned up for the Tide to be able to reliably run the ball inside.

Young linemen aren’t all bad, though: that right side, especially on the outside, did well in this game: 50% success rates behind the right end and right tackle is a solid, above-average success rate. Plus, the small sample size aside, both explosive runs (featuring B.J. Emmons) came on the ride side. Right-side run attempts happened throughout the game, so there are a few suspects as to our mysterious road-grading lineman: is it true freshman starter Jonah Williams? Or backup Korren Kirvin? Only a detailed rewatch will tell, but in any case, let’s hope for more 50%+ success rates out of the line in the coming games.

Parting thoughts

  • In watching the game, one can see what Saban was grumpy about: Alabama won big, but not in the efficient, run-first manner it generally strives for. That said, it was a pretty good effort against a supposedly-underrated team, with quarterback auditions still ongoing (one could argue that Jalen Hurts may have had a better 2nd quarter if given the time).
  • Stewart is still a favorite target for both quarterbacks (with 15 targets), but Calvin Ridley came around this week (16), along with a few others from the Tide’s deep receiving core (the next few were Dieter with 6, Foster with 5, and O.J. with 3 targets)
  • Our plays-by-outcome flow chart is currently broken and in the shop: we’re hoping to have it fixed (and even improved) for the Ole Miss game, but you’ll see it soon either way. Fingers crossed that our next stats review tells a story as good as (or even better than) this one. Roll Tide.