Moving the (Stati)Sticks: Week Three

[Upchurch]

If you’re like most people, watching movies-- especially movies in the same genre-- has a sort of comfort to it, an ease of accessibility that comes from well-tread narratives that we’re familiar with. That isn’t a mistake. There’s a good piece from The Atlantic that details the great lengths studios go to in order to produce thoroughly average and increasingly similar movies. They’re happy releasing something that won’t be remembered long after it’s released but is good enough to bring an audience in and make the money needed for the studio to see some return on its investment, and they’re finding that the key to bringing audiences in is giving them something they’ve basically already watched.

As Derek Thompson wrote in the aforementioned piece:

They are the product of Hollywood's exquisitely designed factory of average-ness, which has evolved as the industry has transitioned from a monopoly to a competitive industry that can no longer afford to consistently value art over commerce.

Hollywood keeps making the same movie over and over because we don’t really want things that are all that different. We get used to something and there’s a standard that’s set, and that becomes what we expect to see. If you’ve watched Michigan this season, you’re getting familiar with this feeling. The stats for the UNLV game are strikingly similar to those of the Oregon State game. If the UNLV game didn’t feel as interesting as the Oregon State game it’s because in most ways we’ve seen it before. Michigan deviates from Hollywood in the quality of its output, but right now they’ve both flipped the switch on the assembly line and are letting the same thing roll off over and over.

-----------------------------------------------------------------------------------------------------------

Michigan again faced an outmatched opponent in week three and did about what you’d expect to them. The second half didn’t have the same feeling of total dominance the Oregon State game had mostly on account of not wholly and completely dominating, but the stats are still kind to Michigan’s performance. Getting the whole wet blanket thing out of the way early, keep in mind that though Michigan’s defense looked great they’ve done their best work against offenses ranked 114th (Oregon State) and 121st (UNLV) in S&P+. But hey, Utah’s offense is ranked 66th and the defense did alright against them. That sounded more comforting in my head.

The advanced box score for Michigan v. UNLV is lopsided from the most basic stats on down. Michigan scored 21 more points than UNLV despite having one less drive (13 to 12), though Michigan did run one more play (64 to 63). Michigan outgained UNLV in yards per play (5.67 to 3.89) and had six scoring opportunities to their two, though Michigan’s points per opportunity was a meh 4.67. Then again, UNLV’s was just 3.50 and they only had two scoring opportunities, so that’s basically a recipe for disaster. As if things weren’t bad enough, Michigan’s average starting field position was the 37.1 while UNLV’s was the 22.8. Success rate was similarly ugly for UNLV, with Michigan holding things together at 53% compared to UNLV’s 32%. Michigan even had a positive turnover margin(!). (It was just +1, but this is Michigan and any movement toward the mean is appreciated.)

[After THE JUMP: Mathlete’s four factors and a whole lot of S&P+ stuff]

Getting all five factors-y, Michigan was more explosive (yards per play) and efficient (success rate), finished drives better (points per scoring opp), had better field position, and won the turnover battle. That would be a clean sweep of the five factors, Michigan’s first of 2015. Bill Connelly’s data (from 2013, but he says it pretty much stays the same each season) shows that you win 86% of games when you win explosiveness, 83% of games when you win efficiency, 76% when you win drive-finishing, 72% when winning field position, and 73% when winning the turnover battle. Win all five? You probably won the game.

Sticking with factors (albeit four instead of five), The Mathlete was again kind enough to send me his four factors data. Getting that weekly allows us to track Michigan’s cumulative performance over the course of the season. His four factors, in his words:

Conversion rate = [1st Downs gained]/[1st Down plays (including first play of drive)]. A three and out is 0/1. A one play touchdown is 1/1. Two first downs and then a stop is 2/3, etc.

Bonus Yards = [Yards gained beyond the first down line]/[Total plays from scrimmage]

This is an adjustment to how I have previously calculated, to account for the plays a team runs.

Field Position = The expected point difference per game for where a team’s offense starts and where a team’s defense starts. Each drive is given an expected value based on the start of scrimmage, all of the drives for the offense and defense are totaled and compared. This accounts for all elements of field position: turnovers, special teams, drive penetration etc.

Red Zone: Points per red zone trip (TD’s counted as 7 regardless of PAT)

Offense:

Michigan’s favorable field position has started to manifest itself as a higher number of expected points, going up 0.3 points this week. The offense’s efficiency also bumped up conversion rate, a rate which isn’t terrible but is just barely in the upper half of the Big Ten. The most noticeable jump the offense made in week three was in Bonus Yards Per Play, which increased by an incredible 0.93, enough for a 52-spot jump in the national rankings. Michigan’s long runs undoubtedly played a big part in the sudden rise in Bonus YPP. Overall, the offense is something of a mixed bag. They have one area that looks really weak (Field Pos.) while everything else ranks in the upper half of the conference.

Defense:

As the offense’s starting field position has gotten better their expected points have gone up, and intuitively the defense’s has dropped as teams have been starting from further back. Every category has been going down defensively with the exception of Red Zone scoring, the story of which you probably got from watching: the defense holds teams almost the entire game but gets sloppy for a drive and lets the opponent score. If the trends above continue after the BYU game we can start talking about whether Michigan’s defense is, for lack of a less PFTCommenter term, elite.

A new feature that debuted at Bill Connelly’s Football Study Hall this week is a collection of team statistical profiles. I already mentioned offensive S&P+; Michigan’s is ranked 76th. Their defensive S&P+, however, is 12th, and that’s bolstering their overall S&P+. (They’re ranked 25th.)

The offense is ranked 118th in explosiveness, 27th in efficiency, 14th in field position, and 89th in finishing drives. Defensively, Michigan is ranked sixth in explosiveness, 37th in efficiency, 28th in field position, and 17th in finishing drives. The offense is way below average in explosiveness, while the defense is one of the best units in the country at preventing big plays. I can only imagine what practice is like. As for turnovers, Michigan’s expected turnover margin is 34th while their actual margin is 91st, so it’s fair to expect that turnover generation will improve as the season goes on.

It’s pretty clear that what’s weighing down Michigan’s offensive S&P+ ranking is a lack of explosiveness, but a bit more evidence: IsoPPP is a measure of explosiveness, and Michigan’s rushing IsoPPP is ranked 107th while their passing IsoPPP is 106th. Compare that with Michigan’s rushing Success Rate (25th) and passing Success Rate (38th) and it’s clear that Michigan has been good at staying in favorable down and distance situations but hasn’t broken off many big plays.

There’s a weird exception to the aforementioned lack of big plays, however, and it’s on passing downs. Michigan’s passing downs Success Rate is ranked just 88th, but their passing downs IsoPP is 22nd. That’s a stark contrast to their standard downs stats, where they’re ranked 19th in success rate and 118th in IsoPPP.

Defensively, the numbers reflect what we’ve seen in opponent’s inability to run on Michigan; they’re ranked 16th in rushing S&P+, 28th in rushing Success Rate, and 12th in rushing IsoPPP. One of the more cheshire-cat-grin-inducing stats I’ve seen is Michigan’s 29.7% stuff rate (which ranks 10th in the nation), a stat that measures how often a runner is hit at or behind the line of scrimmage.

If you’ve ever microwaved a frozen meal at work only to take it back to your desk and find out that part of the middle’s still kind of cold then you have first-hand experience with the culinary version of Michigan’s offense. The food’s not inedible and is actually fine in most ways, but there’s definite room for improvement.

The defense so far is, uh…pizza. Or maybe you don’t like pizza. In that case it’s whatever food you like. More accurately, it’s a choice between your favorite food or freezer-burned vegetables; one thing is pretty clearly awesome, though it’s not exactly like it’s facing stiff competition.

-----------------------------------------------------------------------------------------------------------

It may not have rolled off the line at the factory of average-ness, but we’ve seen this before. The UNLV game’s plot was familiar and felt comfortable, like re-reading your favorite book or listening to your favorite album. At least, I think it did; it’s been a while since I watched football and wasn’t filled with intermingling dread and boredom.  The defense may have played some bad teams, but the stats they’ve posted have been impressive nonetheless. The offense isn’t explosive, but they’ve been efficient and have benefitted from good field position. These plotlines are becoming more familiar by the week; I wouldn’t mind watching the same thing again this weekend.