Jimmystats: The 2016 Schedule by Stats Comment Count

Seth February 23rd, 2016 at 4:24 PM


As some noticed on the twitters I've begun putting together the stat boxes for this year's HTTV opponent previews. I figured I might as well share some of that data here in one place.

FEI and S&P+ things: Champion stats by the two resident best internet football stats guys. Brian Fremeau (@bcfremeau) of Football Outsiders, and Bill Connelly (@SBN_BillC) of that and Football Study Hall on SBNation.

Connelly is responsible for, among other stats, S&P+ ratings (for offense, defense, etc.), which are derived from play-by-play and drive data of every FBS game. S&P+ measures four of five factors that determine game outcomes: efficiency, explosiveness, field position, and finishing drives. The fifth, turnovers, is relatively random so it's left out except as extra weight on sack rates, a thing that will effect at least one weird number we'll see. Garbage time is removed, and it's all weighted by opponents.

He also puts out pre-season projections based on recruiting, returning production, and front-weighted S&P+ of the last five years. Michigan's opponents by Bill C's Projected 2016 S&P+:

School S&P+ Rk/128 Recruiting Ret Prod 5-yr
Michigan 19.3 6th 14th 5th 17th
Ohio State 16.4 14th 5th 18th 3rd
Michigan State 13.5 22nd 18th 30th 12th
Penn State 11.3 28th 17th 39th 29th
Wisconsin 8.3 37th 33rd 60th 15th
Iowa 8.1 38th 49th 32nd 48th
Indiana 3.9 56th 55th 57th 75th
Maryland 2.9 62nd 47th 65th 77th
Illinois 0.4 76th 67th 76th 73rd
Colorado -2.2 82nd 50th 87th 101st
Rutgers -3.1 87th 60th 93rd 84th
Central Florida -7.0 99th 57th 113th 70th
Hawaii -13.4 118th 102nd 116th 120th

Not a lot of play in that schedule; the big rivals look to remain tough tests but that's it for the expected Top 25. The first two games should be good tuneups for O'Korn/whoever.

Brian Fremeau made FEI and F/+, based on opponent-adjusted drive efficiency. Clock kills and garbage time are filtered out, and strength of schedule is factored in. Since it's an overall efficiency thing I prefer to use FEI as a single-stat measure of an offense or defense, while going to the play-by-play nature of S&P+, I tend to use that and the raw, sack-adjusted* yards per play, to represent an offense and defense's run/pass splits.

Those and more after [The jump.]

* [The NCAA treats sacks as rushing, which doesn't make sense. So every year I take the NCAA's base stats and treat sacks as pass plays.This makes a huge difference. I've put them in a Google Doc if you want at 'em.]

2015 Offenses of Michigan's '16 opponents (ranked by FEI)

Offense Run% OFEI Rush YPC Ru S&P+ Pa YPA Pa S&P+
MSU 54% 0.61 (23rd) 4.2 (113th) 94.6 (97th) 6.8 (48th) 115.8 (30th)
Indiana 54% 0.57 (25th) 4.9 (79th) 105.6 (56th) 7.9 (20th) 126.3 (11th)
OSU 59% 0.54 (26th) 6.0 (12th) 124.7 (8th) 6.8 (45th) 118.5 (26th)
Iowa 56% 0.39 (31st) 5.1 (63rd) 108.0 (44th) 6.6 (53rd) 108.2 (42nd)
Michigan 48% 0.36 (33rd) 4.5 (97th) 104.6 (60th) 6.9 (40th) 124.5 (17th)
Rutgers 55% -0.01 (56th) 5.1 (70th) 105.1 (58th) 6.3 (66th) 107.7 (46th)
Illinois 38% -0.11 (70th) 4.7 (85th) 103.1 (66th) 5.1 (112th) 101.4 (67th)
Wisconsin 50% -0.15 (80th) 4.4 (108th) 98.7 (80th) 6.3 (64th) 110.2 (39th)
PSU 47% -0.16 (81st) 5.1 (67th) 108.9 (37th) 5.8 (91st) 96.2 (81st)
Maryland 53% -0.28 (91st) 6.2 (7th) 119.6 (12th) 4.7 (119th) 85.8 (107th)
Colorado 49% -0.59 (111th) 4.6 (90th) 97.0 (89th) 5.6 (98th) 88.3 (100th)
Hawaii 49% -1.04 (123rd) 4.6 (88th) 89.6 (110th) 5.0 (114th) 80.7 (113th)
UCF 46% -1.19 (124th) 3.4 (127th) 67.6 (127th) 4.7 (121st) 74.7 (122nd)

The spread between the spread-to-run teams (OSU and Maryland) and the rest is visible. Corey Clement getting hurt was a problem but not the only one that saw Wisconsin's typically top 10 rushing offense drop below even the problematic passing game, though the advanced stats aren't as harsh as that simple yards per carry I highlighted. If the Terps only had a guy capable of throwing a football last year they might have been trouble.

2015 Defenses of Michigan's '16 opponents (ranked by FEI)

Defense DFEI Ru YPC Ru S&P+ Pa YPA Pa S&P+
Ohio State 0.70 (10th) 4.1 (16th) 118.1 (18th) 5.3 (7th) 137.2 (4th)
Michigan State 0.66 (15th) 4.5 (36th) 116.3 (21st) 6.7 (75th) 126.6 (10th)
Penn State 0.59 (19th) 5.0 (70th) 108.7 (39th) 5.3 (8th) 126.7 (9th)
Michigan 0.58 (20th) 4.5 (37th) 113.3 (28th) 4.9 (1st) 125.9 (11th)
Iowa 0.51 (25th) 4.4 (24th) 116.1 (22nd) 5.8 (24th) 105.9 (45th)
Illinois 0.11 (57th) 4.8 (53rd) 102.3 (62nd) 5.3 (9th) 118.4 (17th)
Colorado -0.05 (73rd) 5.7 (110th) 91.7 (95th) 6.1 (40th) 98.7 (72nd)
Maryland -0.08 (77th) 4.4 (26th) 111.8 (34th) 7.3 (95th) 97.2 (76th)
Indiana -0.61 (106th) 5.9 (120th) 90.6 (97th) 7.0 (85th) 91.9 (92nd)
Hawaii -0.68 (110th) 4.9 (66th) 96.8 (81st) 6.9 (84th) 87.5 (106th)
Wisconsin 0.68 (11th) 4.0 (6th) 119.5 (14th) 5.3 (6th) 126.7 (8th)
UCF -0.88 (121st) 5.2 (83rd) 88.2 (106th) 8.5 (123rd) 88.0 (105th)
Rutgers -0.94 (122nd) 5.4 (96th) 90.4 (100th) 8.7 (125th) 75.1 (127th)

Yeah I saw that Michigan State pass S&P+ and was like "What the deuce?" Remember how above I said that Connolly weights sack rate to account for turnovers, since that's the only reliable correlation? MSU's pass rush was fierce to behold, but their sack rate wasn't THAT high. I mean, I watched football last year…

Using base stats (total sacks divided by pass attempts+sacks) their sack rate (7.66%, 27th overall) was around the same as Michigan's (24th overall), and Fremeau's adjusted sack rate had them 17th. For comparison Penn State was 2nd, and they didn't fold against Rutgers, or let Tommy Armstrong gunsling his way to an upset. Granted the Spartans did run into Oregon minus a quarterback with a functional hand, caught the Urban Meyer on the one day of his career he'd wake up as Al Borges, and Bama barely needed to pass when their line was opening running lanes big enough to drive Spartan logic through them. Or perhaps Bill C. found a way to calculated DISRESPEKT to make up for a tempo-free YPA comparable to Hawaii or Indiana's defenses?

AYG, OL Yds, Havoc, TO Rates, etc.

I also included a few readily available stats. The % run traditionally came from the basic sack-adjusted stats, but Football Outsiders included those, presumably absent the garbage time rushes that throw off the numbers, with their FEI stats so I used that. Ditto turnover rate.

Here's the schedule (and Michigan's at the end) and I'll explain those stats below:

Offense Base O O AYG OL Adj. LY T.O. Rate O Adj. Pace
Hawaii Pistol spread 32% (120th) 100.2 (72nd) 4.01% (125th) -7.7 (very slow)
UCF Modern spread 30% (124th) 78.5 (124th) 4.23% (126th) -8.4 (very slow)
Colorado West Coast 43% (79th) 98.1 (83rd) 2.08% (55th) 5.5 (high tempo)
Penn State Zone stretch 39% (106th) 104.6 (45th) 2.17% (62nd) -9.0 (very slow)
Wisconsin Inside zone 45% (65th) 100.5 (70th) 1.93% (41st) -0.1 (normal)
Rutgers Spread option 42% (87th) 104.4 (48th) 2.23% (66th) -4.1 (slowish)
Illinois Spread to pass 39% (102nd) 96.3 (90th) 1.87% (37th) 1.7 (up-tempo)
Michigan St IZ/WCO 55% (17th) 104.3 (49th) 1.41% (14th) -0.5 (normal)
Maryland Modern spread 37% (113th) 110.9 (22nd) 4.34% (127th) -2.5 (slowish)
Iowa Zone strech 49% (43rd) 103.8 (51st) 1.71% (22nd) -3.7 (slowish)
Indiana Tempo IZ 53% (24th) 108.2 (35th) 1.33% (9th) 10.0 (hyperspeed)
Ohio State Power spread 52% (25th) 120.2 (7th) 2.02% (50th) -1.5 (normal)
Michigan Power I/WCO 51% (32nd) 103.7 (52nd) 1.77% (26th) -1.6 (normal)

AYG: Available yards gained. It's a simple measure of the % of yards to the goal line an offense attained during non-garbage time. It's a check on scoring stats by showing offenses that tend to at least move the ball fairly consistently (OSU, Indiana, MSU) versus those that have a lot of three-and-outs that put their own defenses in a bad position.

OL Adjusted Line Yards: Football Outsiders borrowed this from their NFL stats. It's on a median 100 scale (100 is average), and is opponent-adjusted. From the glossary:

Adjusted Line Yards (ALY): Statistic that attempts to, even to a small extent, separate the ability of a running back from the ability of the offensive line. Adjusted Line Yards begin as a measure of average rushing yards per play by running backs only, adjusted in the following way:

  • 0-4 yards: 100% strength
  • 5-10 yards: 50% strength
  • 11+ yards: not included
  • runs for a loss: 120% strength

Each play is also adjusted based on game situation as well as quality of opponents faced. Adjusted Line Yards can be listed as total or broken down by direction to attempt to isolate ability of specific linemen

Note that Maryland's OL wasn't its problem (and they had three NFL prospects on it) and Ohio State's was its strength, while Michigan's intact OL was about comparable to MSU's at many times patchwork one.

Pace: Another gift from Bill C, he calculates this based on a function of plays per game as a function of run/pass ratio, then compares actual plays versus the number he would expect.

I didn't use rank because the distribution of these isn't a line:


You can see at the top are teams that use it as a strategy (Baylor, Indiana, etc.), and at the other end are teams that have to be trying to eat up clock during a game (Penn State was the slowest). So I broke them up by descriptors whenever there was a convenient spot, like so:

  • Hyperspeed: 10.0 (Indiana) to 15.5 (Baylor)
  • High-tempo: 5.5 (Colorado) to 9.8 (ASU)
  • Up-tempo: 1.6 to 4.8 (MTSU)
  • Normal: –2.0 (UTEP) to 1.1 (VT)
  • Slowish: –5.2 to –2.3 (SDSU)
  • Very Slow: –9.0 (PSU) to –5.3 (Oregon St)

Indiana and Colorado will push the pace on Michigan, and Illinois can bring it out of the bag, but the rest of those on the schedule are generally plodding. Notably Ohio State will get to the line with no huddle and then eat up a lot of the play clock shifting around.

Here's the defensive side of those, with one extra thing of questionable usefulness that happens to tickle me fancy:

Defense D Havoc D AYC Adj DL Yds Adj Sk RT T.O. Rt
Hawaii 13% (bendy) 52% (103rd) 89.0 (110th) 95.6 (71st) 1.05% (127th)
UCF 13% (bendy) 59% (121st) 88.3 (113th) 62.6 (117th) 1.53% (112th)
Colorado 16% (normal) 52% (102nd) 90.3 (104th) 103.2 (58th) 2.32% (60th)
Penn State 20% (vicious) 36% (14th) 104.3 (47th) 197.1 (2nd) 2.49% (40th)
Wisconsin 19% (aggressive) 30% (2nd) 104.5 (45th) 130.3 (23rd) 2.65% (29th)
Rutgers 16% (normal) 62% (125th) 101.1 (57th) 69.9 (110th) 1.98% (87th)
Illinois 18% (aggressive) 41% (37th) 104.6 (44th) 95.2 (73rd) 2.44% (45th)
Michigan St 19% (aggressive) 42% (44th) 117.6 (10th) 135.5 (17th) 3.13% (10th)
Maryland 17% (aggressive) 49% (83rd) 109.6 (36th) 137.6 (14th) 1.99% (86th)
Iowa 16% (normal) 38% (23rd) 98.8 (71st) 114.5 (41st) 2.81% (18th)
Indiana 16% (aggressive) 51% (95th) 101.1 (56th) 94.9 (74th) 2.12% (74th)
Ohio State 17% (aggressive) 32% (5th) 109.5 (37th) 157.4 (4th) 2.34% (59th)
Michigan 19% (vicious) 37% (16th) 112.0 (25th) 127.3 (26th) 1.47% (117th)

Sack rate is self-explanatory and the rest are the other side of the coin of those I just discussed but let's get into Havoc.

Havoc Rate: A stat by Bill Connolly that accounts for % of plays that result in a TFL, an incomplete pass defensed (including INTs), and forced fumbles. Good defenses will create more Havoc, but there's a ton of wiggle, as evidenced when I plot the Havoc rates by S&P+:


Plotting at least shows some intent, for example Ohio State had a relatively "aggressive" havoc rate but the graph makes them look more like they were playing conservative and just happened to generate a lot of those events in the normal course of being good at defense. On the other hand you have Rutgers and Indiana appearing to play pretty loose but unsound.

Anyway I did the same thing with labeling instead of ranking because of this:

12.2 or lower: Pushovers
12.8 to 14: Bendy
14.2 to 16: Normal
16.2 to 19.1: Aggressive
19.3 and up: Vicious

But I'll take another look at it later to see if I can do something with that to show teams that are purposefully generating CHAOS versus those that come by it naturally in the course of having Joey Bosa.



February 23rd, 2016 at 4:42 PM ^

One thing you could do in terms of seeing how "organic" the Havoc is would be to look at who makes the plays - are LBs or DBs getting TFLs? That would mean more blitzing (or trying to throw a bubble at Peppers). I'm not sure how to do that with passes defended, as that could be a lineman batting a ball, or an LB/DB in coverage.


February 23rd, 2016 at 7:54 PM ^

Too hard to separate that from quarterback runs and options. Teams that scramble more are most likely had quarterbacks who can run anyway. And once a Qb crosses the LOS run things are happening, such as blocking and breaking tackles, which would inform rushing stats. So there's no little to count QB scrambles. Sacks on the other hand are failed passes with negative yardage that affect passing team stats to the point that it distorts the results beyond usefulness.


February 24th, 2016 at 10:52 AM ^

Fair enough but the same can be said for QB scrambles that result in a sack.  If Rusell Wilson scrambles and he gets tackled 1 yard short of the LOS it comes off his passing total.  If he gets 1 yard past the LOS it goes towards his rushing totals even on a similar play.


I think a better metric might be sacks that occur in the pocket come off passing totals, everything else is rushing.  But I agree that would be difficult to assess without a new stat being tracked.


February 25th, 2016 at 7:10 AM ^

the area for Jabrill Peppers to show the greatest improvement is in his ability to create turnovers for our defense. There was one game last year (Maryland?) where Peppers broke up a pass but he did not catch it and the play would have resulted in a pick 6. Those are the plays he needs to finish for us to take another step forward on defense. Not to mention they will help him get to NYC for the Heisman presentation :)