clock menu more-arrow no yes

Filed under:

Processing the Numbers, Bye Week | Where Things Stand

New, 66 comments

Taking stock of the Tide's standing in the advanced metrics, plus looking at some game charting results

Jason Getz-USA TODAY Sports

All statistics are courtesy of Football Outsiders, home of the F/+ Combined Ratings for college football.
The Fremeau Efficiency Index (FEI) was created by Brian Fremeau; check out his website BCFToys for other goodies.
The S&P+ rating was created by Bill Connelly; check out his college football analytics blog, Football Study Hall.
Hat tips to Addicted to Quack's kalon and FO's 7th Day Adventure column for the inspiration.

How'd last week go?

Wow. Florida got completely obliterated. The game was still up for grabs in the 3rd according to From the Rumble Seat’s millsgt49 and his win probability model, but by the end of that backbreaking 16 play, 66 yard drive midway through the third, the game was well in hand. Not quite 60 minutes, but one or two other things go Florida’s way and it might’ve been.

The Goods
Overall Quality
ALABAMA
F/+ 27.1% (1)
FEI 0.241 (7)
S&P+ 262.3 (1)
OS&P+ 127.9 (5)
DS&P+ 134.4 (4)

(bold) numbers indicate national ranking.

Wondering what all these terms are?

  • FEI: The Fremeau Efficiency Index, an overall team quality metric that is drive-based and opponent-adjusted. For a more detailed discussion of FEI, check out the PTN primer.
  • OFEI: The offensive component of FEI.
  • DFEI: The defensive component of FEI.
  • FPA: FEI Field Position Advantage, a measure of how much field position value a team earned against its opponents.
  • S&P+: Another overall team quality metric, S&P+ is primarily play-based and consists of three components: Success Rate, Equivalent Net Points per Play, and a drive efficiency component. The "+" refers to opponent adjustments. For a more detailed discussion of S&P+, check out the PTN primer.
  • OS&P+: The offensive component of S&P+.
  • DS&P+: The defensive component of S&P+.
  • Rush OS&P+: S&P+ calculated on rushing plays for the offense — a good measure of a team's effectiveness at running the ball.
  • Rush DS&P+: S&P+ calculated on rushing plays for the defense — a good measure of a team's effectiveness at stopping the run.
  • Pass OS&P+: S&P+ calculated on passing plays for the offense — a good measure of a team's effectiveness at throwing the ball.
  • Pass DS&P+: S&P+ calculated on passing plays for the defense — a good measure of a team's effectiveness at defending the pass.
  • PD: Passing Downs, defined as later downs with medium yardage or more to go (3rd, 4th downs in excess of 5 yards to go), as well as 2nd down with more than 8 yards to go.
  • SD: Standard Downs, defined as all downs that are not Passing Downs.
  • SD OS&P+: S&P+ calculated on standard downs for the offense — a good measure of a team's offensive effectiveness on earlier downs and short yardage.
  • SD DS&P+: S&P+ calculated on standard downs for the defense — a good measure of a team's defensive effectiveness on earlier downs and short yardage.
  • PD OS&P+: S&P+ calculated on passing downs for the offense — a good measure of a team's offensive effectiveness on later downs and long yardage.
  • PD DS&P+: S&P+ calculated on passing downs for the defense — a good measure of a team's defensive effectiveness on later downs and long yardage.
  • F/+: The F/+ combined ratings combine FEI and S&P+ into one metric that serves as the official college football rankings of Football Outsiders. For a more detailed discussion of F/+, check out the PTN primer.
  • Off. F/+: The offensive component of F/+.
  • Def. F/+: The defensive component of F/+.
  • ST F/+: The special teams component of F/+.
  • Swanson Giddiness Index: Easily the most accurate predictor of success in college football, the Swanson Giddiness Index is a qualitative, completely unsupportable metric that is presented via the tone of that week's image/animated gif of Ron Swanson — beloved Parks and Recreation character and official spirit animal of Processing the Numbers.

Wondering what all of this means? Check out the PTN primer!

The Disclaimer

For the first six weeks of the season, these metrics are based partially on a few projection factors, namely recent program performance, the effects of roster attrition, recruiting rankings, and sweet, sweet voodoo*.

As the season progresses, data from games played will be factored in, with a progressively lighter emphasis on the projection factors. Starting in Week 7, these metrics will be based purely on games played this season. At that time, we’ll also get splits for offense, defense, and special teams, as well as insights on how teams handle passing and short-yardage situations and how they manage field position.
* This last one's not true.

So, what do we know?

Skimpy table this week, but that’s what happens when you don’t have an opponent.

Due mostly to FSU and Oregon struggling with inferior opponents this past week, Alabama is now #1 in both F/+ and S&P+. In another highlight of the inherent goofiness of FEI, FSU actually got slightly better, thanks to Clemson’s epic Clemsoning probably due to holding Clemson to only 17 points despite their numerous scoring opportunities. I suspect some of the preseason projection factors are still monkeying with things here, because FSU is sitting at the top of the heap despite putrid strength of schedule and Game Efficiency. It will be interesting to see where these rankings stand in a couple of weeks when the projection stuff goes away.

Time to play with some charting data.

You can do a lot of things with play-by-play data (like the metrics discussed above), but there’s a lot of questions that data can’t answer. How much passing yardage is due to yards after catch and how much is from the quarterback’s throw? Are teams more successful out of certain formations than others? What about when they run to certain parts of the field, or when they run play action?

Enter the FootballStudyHall Charting Project. In addition to the typical play-by-play stats (down, distance, yardline, play type, yardage gained / play success, target/rusher), charters also grab data on formation (number of backs and receivers, numbers of rushers and blockers, whether it was shotgun, pistol, wildcat, etc.), playcalling details (no huddle, play action, blitzing), passing details (air yards, yards after catch, pass direction, reason for incompletion), rushing details (QB movement, run direction), and yards after contact/broken tackles data. The number of things you can look at with this kind of stuff is limited only by the amount of data you can acquire (charting a game in this fashion takes most folks 3 to 4 hours, and it’s all volunteer).

I’m participating this year, and have charted all 4 of Alabama’s games. I pulled a sampling of things from this data for comparison below. I should note all of this falls under the caveat of small sample size – this kind of stuff works a lot better when you look at a large group of teams over an entire season. Also, I would love to put all of this in the context of the entire country or among all quarterbacks, things of that nature, but capturing the entirety of FBS games on volunteer labor is a pipe dream.

Instead, I pulled data from last year’s project for the 2013 Crimson Tide for comparison. Keep in mind though that the 2013 numbers come from games like Virginia Tech, Texas A&M, Auburn, and Oklahoma, versus the slate so far this year. Somewhat different levels in team quality there.

Passing Splits
Frequency Success Rate
Down McCarron (2013) Sims (2014) Coker (2014) McCarron (2013) Sims (2014) Coker (2014)
1st 43.88% 26.49% 10.60% 52.46% 60.00% 50.00%
2nd 43.69% 31.68% 8.91% 42.22% 59.38% 44.44%
3rd 66.10% 42.86% 14.29% 38.46% 71.43% 57.14%
Air Yards McCarron (2013) Sims (2014) Coker (2014) McCarron (2013) Sims (2014) Coker (2014)
Under 5 13.93% 43.96% 15.96% 50.00% 67.50% 46.67%
5 - 10 23.12% 21.98% 23.40% 51.81% 65.00% 59.09%
10 - 15 27.86% 16.48% 26.60% 53.00% 66.67% 60.00%
Over 15 35.10% 17.58% 34.04% 53.17% 50.00% 53.13%
Update: Should have clarified that "Success Rate" here is the same as in the excellent RB Success Rate articles Kevin has been doing for us this season. As a refresher, "success" is when you achieve 70% of required yardage on 1st down, 50% on 2nd down, and 100% on 3rd and 4th downs. Your success rate is just how often you're successful.

Observations

First I should probably define what I mean by Air Yards. Simply put, Air Yards are the down-the-field or vertical distance the ball travels from the line of scrimmage to the receiver — yardage across the field is not considered. Obviously if you throw a ball 50 yards down the field from the left hash to the right hash, that ball traveled a lot farther than 50 yards. But for this, we’re just concerned with the down-the-field distance. In the case of underthrown balls, the Air Yards refer to where the ball lands. In the case of overthrown balls, it’s where the receiver is.

I left out 4th down data here — not a whole lot to go off of, so not worth listing. So far in 2014 Alabama tends to throw the ball a bit less on 1st and 2nd downs, but are a little more successful when doing so. Blake Sims in particular has been absurdly good on 3rd downs — anytime you’re successful at something on a football field 70% of the time you’re doing it right. Will be interesting to see if this trend holds through the meat of the schedule.

The yardage frequency is pretty telling, as Sims has a very different distribution from McCarron and Coker, who have pretty similar numbers. McCarron and Coker are both down-the-field throwers, whereas Sims has been making a lot of his hay on screens and other short-yardage routes. We saw a lot more vertical passing from Sims against Florida, it will also be interesting to see if that trend continues down the line.

Yes, McCarron was more successful the farther he threw it. No, that is absolutely not normal. As I recall, when Bill Connelly introduced this kind of breakdown in Study Hall, the same thing was going on with McCarron in 2012, and it was just as odd then as it is now. Almost like the guy was a good quarterback or something.

The other notable thing from this quadrant is as long as Sims is passing within 15 yards of the line of scrimmage he’s been pretty unstoppable. Some of that is still due to the receivers, but you can see from Coker’s numbers that some of it has to do with the quarterback as well.

Rushing Splits
Frequency Success Rate
Direction 2013 2014 2013 2014
Left End 8.78% 17.53% 61.54% 55.56%
Lt. Tackle 13.51% 12.34% 45.00% 52.63%
Middle 46.62% 37.01% 50.72% 50.88%
Rt. Tackle 21.62% 17.53% 62.50% 62.96%
Right End 9.46% 15.58% 42.86% 54.17%
Down 2013 2014 2013 2014
1st 56.12% 62.91% 53.85% 54.74%
2nd 56.31% 59.41% 51.72% 61.67%
3rd 33.90% 55.10% 55.00% 55.56%

Observations

Same thing as before — I left out 4th down data because it didn’t tell us much. What jumps out to me from the Run Direction data is that Kiffin spreads things out a bit more than Nussmeier did. There’s still an emphasis on running up the middle, but more use is made of the edge and there’s not a huge slant to the right side like there was last year. Part of that is undoubtedly due to the departure of Anthony Steen. Aside from a puzzling dip on success rate around the left end (something to do with Kouandjio, perhaps?), the 2014 Tide have been more successful running the ball than the 2013 squad along the rest of the line.

Huge jump on 3rd down runs, probably due to the creampuffs (lots more garbage time/running out the clock) as well as the higher success rates on 1st and 2nd downs leading to shorter yardage on 3rd down. As with the run direction data, the 2014 running game so far has been more successful than the 2013 version regardless of down.

Formation/Play Action Splits
Frequency Success Rate
Formation 2013 2014 2013 2014
Shotgun 43.42% 24.68% 43.94% 64.47%
Pistol 13.82% 16.56% 59.52% 64.71%
Under Center 42.76% 58.77% 52.31% 52.49%
Play Action 9.87% 10.71% 50.00% 57.58%

Observations

I didn’t include the two Wildcat plays, since none of that was run in 2013 to my knowledge. Main thing to note here is we run a lot less shotgun in favor of a slight uptick in pistol and a big jump in under center. I think the latter has to do with all of the bootleg action we’ve seen from Sims — I don’t recall much of that from McCarron. Success Rate doesn’t tell us much here, as the offense is just better so far than it was last year. Same thing with play action — slight uptick in usage, and a better success rate.

What do yall think about this stuff? Is this something you’d like to see more often than just bye weeks? If there’s something else you’d like insight on, let me know in the poll or in the comments!

Swanson Giddiness Index

Rs-drunk_medium

Ron’s still celebrating last Saturday’s beatdown. Good thing it’s a bye week!

ROLL TIDE