S&P+ Five Factors Matchup: UM vs. Penn State

Submitted by Ecky Pting on

edlt: Oops! Corrected title...

Here's the next installation of Bill Connelly's Five Factors metrics matchup between UM & Penn State. It's a bit busy, but what you see are columns of raw metrics for both offenses and defenses. The Category of the given metric is given in the two columns at the left. To the right of the team offense and defense metrics are the National Averages for that category.  The last two columns are where the rubber meets the road...

The "M Offense vs. PSU Defense" column averages those two metrics to gauge the performance of the UM offense against the CU defense. Likewise, the "CU Offense vs. M Defense" averages the other two to gauge the performance of the CU Offense. From there, the column with the greater aggregate number has the competitive advantage...EXCEPT, in the three categories with asterisks: "Stuff Rate", "SD Sack Rate" and "PD Sack Rate", which are contra-metrics that gauge the offense's ability to avoid the given categorical description.

Anyway, the numbers showing the advantage are in bold, and as such it appears the matchups tilt in M's favor in all but one of the Five Factors: Explosiveness; and all but four sub-categories, those being Passing IsoPPP (passing explosiveness, measured as pts. scored per successful passing plays), SD IsoPPP (pts. per successful standard down), PD IsoPPP (pts. per successful passing down), and the same as against Colorado...PD Line Yards per Carry (bonus yards gained by running on passing downs...a.k.a. breaking contain!). As with the Buffs, gap integrity is going to be an important discipline for Wolverines success against the Nittany Lions. The IsoPPP advantage of PSU in both the standard and passing downs contribute to their advantage in the overall Explosiveness category. It's readily apparent that the most likely means to PSU success against the Wolverines will be through the big play.

If you're reading this you've probably already witnessed what these numbers are quantifying. Of all the individual metrics in the table, the IsoPPP numbers for M's Defense are particularly glaring. IsoPPP overall ranks #123, with Rushing IsoPPP ranked #108 and Passing IsoPPP ranked #109, and SD IsoPPP at #124. What's that mean? M is susceptible to big plays when opposing offenses are ahead of the sticks. Is this an over-aggressive stance by Don Brown's defense? Is it evidence of a learning curve as the team becomes more familiar with Don Brown's schemes? Is it the impact of injuries on the DL and/or in secondary?

FIVE FACTORS
(less T/O Luck)		 M
Off.		 M
Def.		 PSU
Off.		 PSU
Def.		 Nat'l
Avg.		 M.Off v
PSU.Def		 PSU.Off
v M.Def

[1] EXPLOSIVENESS
IsoPPP 

 1.47 1.58 1.36 1.24 1.28 1.36 1.47
[2] EFFICIENCY
Success Rate 		 44.7% 21.3% 43.1% 37.7% 40.1% 41.2% 32.2%
[3] FIELD POSITION
Avg. FP 		 38.4 26.4 37.6 28.9 29.6 33.65 32.00
[4] FINISHING DRIVES
Pts. Per Trip in 40 		 6.63 2.81 4.82 4.56 4.72 5.60 3.82
-- RUSHING --              
Rushing Success Rate 45.2% 17.1% 42.0% 37.4% 41.1% 41.3% 29.6%
Rushing IsoPPP  1.17 1.29 0.86 1.25 1.09 1.21 1.08
Opportunity Rate  39.7% 31.1% 35.1% 39.3% 39.8% 39.5% 33.1%
Power Success Rate 88.9% 60.0% 53.8% 46.2% 68.6% 67.6% 56.9%
Stuff Rate* 12.1% 30.6% 23.8% 23.5% 18.9% 17.8% 27.2%
-- PASSING --              
Passing Success Rate 44.3% 24.5% 44.2% 38.1% 40.1% 41.2% 34.4%
Passing IsoPPP  1.78 1.73 1.86 1.21 1.50 1.50 1.80
-- STANDARD DOWNS --              
SD Success Rate  48.9% 25.5% 49.3% 45.7% 46.0% 47.3% 37.4%
SD IsoPPP  1.3 1.55 1.13 1.12 1.14 1.21 1.34
SD Line Yds/Carry  3.66 1.35 1.77 2.78 2.96 3.22 1.56
SD Sack Rate* 5.1% 12.0% 3.8% 8.9% 4.5% 7.0% 7.9%
-- PASSING DOWNS --              
PD Success Rate  33.3% 15.4% 29.5% 22.9% 30.1% 28.1% 22.5%
PD IsoPPP  2.13 1.64 2.21 1.68 1.75 1.91 1.93
PD Line Yds/Carry  1.79 2.31 3.14 2.74 3.29 2.27 2.73
PD Sack Rate* 4.6% 14.3% 7.1% 12.8% 7.9% 8.7% 10.7%

 

MGoBoard
Football
S%26P+; Bill Connelly; Football Outsiders; Advanced Stats; Matchup; Colorado

slimj091

September 19th, 2016 at 10:20 PM ^

That would be news to most down here in poisonous nut land. Just got an ear full tonight about how Michigan is an overrated 7-5 team because they were down 14 points in the first quarter against Colorado, and how Penn state is going to beat us by 30+.

Thank god for cannabis. Otherwise I wouldn't be able to function on a daily basis as a Michigan fan in this state.

Joined: 01/04/2014

MGoPoints: 9542

Wolfman

September 20th, 2016 at 1:22 AM ^

remember, it is probably the same people who think Carr was a better coach than Bo. And you know the ONE reason. They never factored in the fact that he had never proven himself capable of restoring a dormant program and not just bringing it back to relevance but making it one of the most elite programs in the nation for two decades, starting with his first year in AA.

And these people, who they are I cannot imagine, are somehow forgetting Harbaugh is batting over .800 three games into his second season. And it's pretty frickin simple. People who believe Hoke is somehow better than Harbaugh don't know fuck about football so I would not consider a damn thing they type. 

It is UM who allowed Brady to finish his head coaching career above .500. And many add "but yes, nice guy." I defer to Leo Durocher. 

Joined: 07/23/2008

MGoPoints: 8259

Mgodiscgolfer

September 20th, 2016 at 6:40 AM ^

why there are groups of people who will believe anything. I imagime if you look hard enough mainly in Pennsylvania there is a group of people who believe Joe Pa had no idea what was happening in his teams shower room. Even though other people connected to the team reportedly did and told Joe Pa that they did, But because he had more hair in his ears than under his arms he miss understood one whistle blower to say. I though I saw dairy in the showers with a joy I had to turn my head.

Joined: 01/13/2011

MGoPoints: 5322

Mgodiscgolfer

September 20th, 2016 at 6:40 AM ^

why there are groups of people who will believe anything. I imagime if you look hard enough mainly in Pennsylvania there is a group of people who believe Joe Pa had no idea what was happening in his teams shower room. Even though other people connected to the team reportedly did and told Joe Pa that they did, But because he had more hair in his ears than under his arms he miss understood one whistle blower to say. I though I saw dairy in the showers with a joy I had to turn my head.

Joined: 01/13/2011

MGoPoints: 5322

denardogasm

September 19th, 2016 at 10:58 PM ^

I hope you brought up the fact that they were getting their asses handed to them by Tulsa until it started raining.  Very few top teams go the whole season without putting up a dud.  Even Bama does it.  Clemson had one.  OSU does it just about every year.  Yet for some reason everyone is taking the Colorado game as the true measure of our team while they don't do the same for those teams.  Let's see what things look like in November.

Joined: 09/11/2011

MGoPoints: 10688

Wolfman

September 20th, 2016 at 12:47 AM ^

but a fragile team. They were certainly in no position to lose a main cog, especially there indispensable one. It is my hope he heals quickly and is able to return to action. They are a decent club this year, but just as our depth would have determined the outcome of that game, their talent, if healthy will allow them to beat a few teams they would not otherwise normally beat. 

Joined: 07/23/2008

MGoPoints: 8259

MGoCarolinaBlue

September 19th, 2016 at 10:27 PM ^

Explosiveness is such a high variance thing... let's get some more games under our belt before we start drawing conclusions about whether Michigan's defense is susceptible to big plays based on a small number of big plays given up.

Remember, at this time last year we were saying the reverse about M's offense (well, not me, but you guys were saying it) -- that Jake Rudock isn't capable of connecting on throws downfield.

Joined: 03/19/2013

MGoPoints: 5774

Ecky Pting

September 20th, 2016 at 7:20 AM ^

I suppose I used the term big plays loosely in equating it to explosive plays, and your point is well taken about numbers like this being high variance, but most sports stats are. For example, the point spread on a game alone has a standard deviation of 15 points. Nonetheless, teams that win the Explosiveness metric win 86% of the time, so the correlation is significant, which is why it's one of the Five Factors, not to mention it's not a factor that is determined by random occurrences, like Turnovers. I always find it amusing when commentators or coaches talk about winning the "turnover battle" as a strategem for victory. But anyway, defense against Explosiveness is clearly something that U-M will need to get under control as it moves on through the season, whether that be through making adjustments to formations/alignments/assignments (coaches), getting more practice reps to get up the learning curve (players) or just getting healthy and back on field (trainers).

Joined: 09/06/2013

MGoPoints: 4655

Wolfman

September 20th, 2016 at 12:31 AM ^

and came up with a final of 44-21, but I did not account for the dropped pass so it could have been a good call on my part. Of course I did not account for the defensive TD or the punt in the back the CU's punter. There are variables that simply can't be accounted for early in the year until a pattern emerges. Now two games is insufficient sample size to determine if we are going to score in this manner. It was obvious J.P. was going to return a punt at some time, and I always factor in field position in our favor, thus giving me a realistic chance to figure out how many opportunities we will have to score based on good field position. 

In other words, after game five or so, we should be able to better determine how many special team scores we might expect if this aspect of the game continues to prove an actual part of our scoring.  Might be as significant as a 4 - 5 point swing in predicitions. Beamer's team used to thrive off this and were good, it seemed for 3 to 6 pts per ball game. That would certainly be nice to add on prior to crushing the numbers based on all the other stats. 

Joined: 07/23/2008

MGoPoints: 8259

The Mad Hatter

September 20th, 2016 at 7:42 AM ^

Which teams are included in the national average?  Because we're pretty damn close to average in almost every category.  Unless this is one of those charts where a 0.1 difference is a big deal.

I don't know, I can't math this early.

Joined: 02/17/2015

MGoPoints: 112109

Mongo

September 20th, 2016 at 8:19 AM ^

Missing an All-American, shutdown corner means a lot to this secondary. When JL is healthy and back in the line-up, we take this thing up a huge notch.

Joined: 06/25/2016

MGoPoints: 9585

Double-D

September 20th, 2016 at 10:21 AM ^

He can't however cover the guys our safety's are supposed to cover...unless you put him on the slot and take out Thomas. Our CB play has been pretty solid so far this year. Safety play needs to get fixed in a hurry.

Joined: 12/21/2013

MGoPoints: 87270

Mongo

September 20th, 2016 at 9:37 AM ^

If I am getting this data ... and 3 game cumulative average run/pass plays for each team ... This would be the point prediction: UM vs PSU Run points = 19.8 (40R, 41.3% success, 1.21 pts per) Pass points = 18.5 (30P, 41.2% success, 1.5 pts per) Turnover/Special Team Points = 9.3 (UM's average this year per game) Total Points = 47.6 PSU vs UM Run Points = 10.4 (35R, 29.6% success, 1.08 pts per) Pass Points = 18.6 (30P, 34.4% success, 1.8 pts per) Turnover/Special Team Points = 2.3 (PSU's average this year per game) Total Points = 31.3 There is your 16 point spread by the numbers. If JL is healthy, no way PSU scores 18.6 points off pass plays. JL is likely our missing link and why our points off pass plays have been so high. Would not shock me to see Lewis taking that high-twitch wiggle guy whether from the nickel or corner spot. Clark playing safety could happen as well. Most of the explosive pass points have come from safeties having to cover faster WRs. Thomas has been great in run support, but a liability in pass cover.

Joined: 06/25/2016

MGoPoints: 9585

jdemille9

September 20th, 2016 at 10:58 AM ^

Three games is a small sample size, especially against the quality of opponents played, but it's still very interesting to look at.

Either way, I think any weaknesses can be attributed in part to the injuries (three high value starters out is tough to overcome) and in part to some kinks in working out the new system. We all probably figured there'd be some kinks and some busted plays while they master Brown's defense.

It's one thing to know the system on paper or even in practice but to perfect it in a game takes time. Even incredibly high IQ guys like Jabrill Peppers haven't got all the nuances down yet. It wasn't gonna happen overnight.

Come end of October this defense should be clicking on all cylinders and, unless we suffer some more significant injuries, I don't see why they don't roll into The Game undefeated and with a murderous defense no one wants to face. 

Joined: 01/29/2011

MGoPoints: 8525