Advanced Stats Matchup Analysis - 2017 Michigan at Penn State
Introduction
Behold, the latest installment of the new and improved Advanced Stats (S&P+) Matchup, this time featuring Michigan at Penn State.
This matchup analysis draws upon the Advanced Stats Profiles published weekly by Bill Connelly on Football Study Hall. The profiles feature Connelly’s well-known Five Factors, and also include the more detailed groups of S&P+ metrics that break down elements of the game such as Rushing and Passing, as well as the down-and-distance scenarios known as Standard Downs and Passing Downs. This new interpretation is a updated take on what you may recall from last season. It still reviews the same 26* advanced stats metrics, but with a new approach that displays the matchups graphically, in a way that more clearly distinguishes and gauges the significance of any net advantages. For more details regarding the definition of and concepts behind each of the metrics, the Advanced Stats Glossary is a handy reference to bookmark.
* The complete set of metrics now includes six more “plus” and other metrics that are adjusted for the level of competition. These were introduced for the 2017 season this week in the Advanced Stats Profiles. They are: Rushing S&P+, Passing S&P+, Rushing Adjusted Line Yards, Passing Adjusted Sack Rate, Standard Down S&P+ and Passing Down S&P+.
Methodology
If you’re interested in the approach to analyzing Bill Connelly’s base metrics, the formulation for deriving the matchup metrics and the data visualization concept for the charts, you can read more in the previous Michigan at Indiana diary. Nonetheless, nothing here is etched in stone, and certainly suggestions for improving any of the aspects of the methodology are welcome and appreciated!
Michigan at Penn State Matchup Analysis
So, on with the matchup analysis!
The Five Factors Matchups
Here are the matchups for the core Five Factors metrics that compose the actual S&P+ ratings from which the game scoring margin is derived. As of the beginning of this week, that margin stands in favor of Penn State, at 10.6 points. Keep in mind a couple of things: the weightings of the factors into the predicted scoring margin are not uniform and, a team has control of only the first four. Not that is matters, because Michigan has an advantage in none of them. Woof.
Efficiency
In Efficiency, the UM Offense is well below average, while the Penn State Defense is top ten, which pulls down the UM Offense a goodly amount. On the other side, the Penn State Offense is well above average, however, the UM Defense remains elite - down to #2 from #1 last week in this category! In the end, net matchup gives a slight advantage to Penn State in Efficiency.
Explosiveness
In Explosiveness, the UM Offense is about average, but the Penn State Defense is well above average, which pulls the UM Offense down to well below average. On the other side, the Penn State Offense is top ten, while the UM Defense is about average, which leaves the Penn State Offense about the same. The net matchup gives sizeable advantage for Penn State in Explosiveness.
Field Position
As for Field Position (measured in yards), the UM Offense is about average, while the Penn State Defense is elite – in fact, the elitist – ranked #1 in this category. The matchup, which is the average of the two (not the geometric scaling as with the first two), pulls the UM Offense down to well below average. On the other side, the Penn State Offense is also elite (ranked #3), and the UM Defense is also about average, which leaves the Penn State Offense, which leaves the PSU Offense still above average. The net matchup, as a result, is a considerable Field Position edge for Penn State.
Finishing Drives
In Finishing Drives (measured in points), the UM Offense is below average, while the Penn State Defense is elite (#3). This matchup, like Field Position, is also determined by a simple average, and leaves the UM Offense well below average. On the other side, the PSU Offense is about average, while the UM Defense is above average. This leaves the PSU Offense still about average. The net matchup is a considerable advantage for Penn State in Finishing Drives.
Turnover Margin
The Tale of Turnover Luck is like a Tale of Two Cities… It was the best of times, it was the worst of times. Looking at the expected turnover margins for Penn State and Michigan, ranked #3 and #6 respectively, you might think they’re twins. But alas, when looking at actual turnover margins, ranked #2 and #92 respectively, these twins were clearly separated at birth. The difference is attributable to luck, and if Michigan did not have bad luck, it would have none at all, whereas PSU’s luck has exceeded even its lofty expectations. Setting luck aside however, and just comparing the statistically-based expected turnover margins, the gulf is narrow. The net difference amounts to about 1.6 PPG, and gives a slight advantage for Penn State in Turnover Margin.
Rushing Matchups
In going up against PSU at least, UM manages to avoid the clean sweep that was the case against MSU. Indeed, Michigan captures a net advantage in half of the Rushing matchups, in addition to the overall metric. Here we see the artifacts of Michigan’s flickering hope, a flicker that needs to be given a gentle source of oxygen and ignitable fuel for it to be successful. So with that, on to the Rushing matchups.
Rushing S&P+
The aggregate Rushing S&P+ metric is an uninverted metric, meaning that higher values correspond to both higher-ranked offenses and defenses. Here, the UM Offense is slightly above average, but the PSU Defense is further above average, which pushes the UM Offense down. On the other side, the PSU Offense is well above average, but the UM Defense is nearly top ten, which pushes the PSU Offense to below average. The net matchup result is a slight advantage for Michigan in overall Rushing S&P+.
Rushing Success Rate
In Rushing Success, the UM Offense below average, while the PSU Defense is top ten, which pulls the UM Offense well below average. On the other side, the PSU Offense is well above average, but UM’s Defense is elite (ranked #4). In the end, the net matchup balance is a sizeable advantage in Rushing Success for Penn State.
Explosiveness (IsoPPP)
In Explosiveness, the UM Offense is above average, as is the Penn State Defense, which makes the UM Offense closer to average. On the other side, the Penn State Offense is above average, while the UM Defense is below average. In the end, Rushing IsoPPP (Explosiveness) favors Penn State by a slight margin.
Adjusted Line Yards
In Adjusted Line Yards, the UM Offense is about average, while the PSU Defense is above average, pushing the UM Offense to drop. On the other side, the PSU Offense is below average, while the UM Defense is elite (ranked #4), which knocks the PSU Offense to rock bottom. The net result is a considerable advantage for Michigan in Adjusted Line Yards.
Opportunity Rate
In Opportunity Rate, the UM Offense is well below average, while the Penn State Defense is nearly top ten, knocking the UM Offense down to rock bottom. On the other side, the PSU Offense is also nearly top ten, as is the UM Defense, which takes the PSU Offense down below average. The net result is still a sizeable advantage for Penn State in Opportunity Rate.
Power Success Rate
In Power Success Rate, the UM Offense is actually well above average, nearly top ten. Think “the Hammering Panda”, and that is what exemplifies Power Success. The PSU defense, meanwhile, is well below average, which vaults the UM Offense up even further. On the other side, the PSU Offense is nearly rock bottom, and to make matters worse for them, the UM Defense is the elitist - ranked #1 in this metric – which hurls the PSU Offense into a black hole. In the end, the matchup balance is an enormous advantage for Michigan in Power Success Rate.
Stuff Rate
Last is Stuff Rate (a contra-metric). In this case, the UM Offense is close to average, while the PSU Defense is well above average, which raises the UM Offense. On the other side, the PSU Offense is once again nearly rock bottom, while UM Defense is top ten. Think “space hogs” like Moe Hurst, with some Brian Mone sprinkled in. So, in the end, the matchup result is another considerable advantage for Michigan in Stuff Rate.
Passing Matchups
Once again UM manages to gain a split on the component metrics with PSU, actually capturing 2 out 3, including the one that is related to OL and DL performance.
Passing S&P+
The aggregate Pasing S&P+ metric is an uninverted metric, meaning that higher values correspond to both higher-ranked offenses and defenses. Here, the UM Offense is below average, while the PSU Defense is top ten, which pushes the UM Offense to well below average. On the other side, the PSU Offense is well above average, but the UM Defense is the elitest (ranked #1 in this metric), which pushes the PSU Offense to well below average. In the end, the net matchup result is a sizeable advantage for Penn State in overall Passing S&P+.
Passing Success Rate
In Passing Success Rate, the UM Offense is well below average, OK? Also, the PSU Defense is top ten, OK? OK, so that takes the UM Offense down a few notches further. No worries. Not a big deal. The reason is because, even though the PSU Offense is about average, the UM Defense is, once again, the elitist (ranked #1 in this metric). The scuttlebutt is that the PSU Offense is obliterated into nothingness, Saquon Barkley on a swing pass or not. The net is a negligible advantage for Michigan in Passing Success Rate.
Passing Explosiveness (IsoPPP)
In Passing IsoPPP (Explosiveness), the average UM Offense is brought down by the elite PSU Defense. On the other side, the slightly above average PSU Offense is unperturbed by the average UM Defense. The net matchup result is a considerable advantage for Penn State in Passing IsoPPP.
Adjusted Sack Rate
Adjusted Sack Rate is an uninverted metric, meaning that higher values correspond to both higher-ranked offenses and defenses. However, it is not a contra-metric as is typical with the other Sack Rate metrics. You can blame Bill Connelly…
In Adjusted Sack Rate, the nearly rock-bottom UM Offense gets a reprieve from the below average PSU Defense. On the other side, however, the well below average PSU Offense gets hammered by the top ten UM Defense. The net matchup result is a considerable advantage for Michigan in Adjusted Sack Rate.
Standard Down Matchups
Michigan captures only 1 of the 5 Standard Down matchups with Penn State, with UM’s only advantage coming in SD Line Yards per Carry. It’s been noted before but is worth repeating that UM’s defensive scheme under Harbaugh, and under Don Brown in particular, is typically weak in the Explosiveness metric - these results are no different - and it’s not a bad thing. What’s important is that it is usually offset by a strong Success Rate metric, which is born out in these results as well, in so far as it amounts to a push in SD Success Rate.
Standard Downs S&P+
The aggregate Standard Down S&P+ metric is an uninverted metric, meaning that higher values correspond to both higher-ranked offenses and defenses. Here, the UM Offense below average, and the PSU Defense is well-above average, which pulls the UM Offense down further. On the other side, the PSU Offense is also well-above average, but UM Defense is also, which pulls the PSU Offense down below average. The net matchup result is a sizeable advantage for Penn State in overall Standard Down S&P+.
SD Success Rate
In SD Success Rate, the UM Offense is well below average. Do you always have the sense that this offense is always “behind the chains”? Well, there you go. The PSU Defense is well above average, which pushes the UM Offense toward rock bottom. On the other side, the PSU Offense is well above average, but the UM Defense is elite (ranked #2), making the PSU Offense also look well below average. In the end net matchup result ends up being a negligible advantage for Penn State in SD Success Rate.
SD Explosiveness
In SD Explosiveness, the UM Offense remains below average, while the PSU Defense is well top ten, which pulls the UM Offense lower still. On the other side, the PSU Offense is above average, while the UM Defense is only average, leaving the PSU Offense about the same. The net matchup result is a significant advantage for Penn State in SD Explosiveness.
SD Line Yards per Carry
In SD LYPC, the UM Offense is below average, and the PSU Defense is above average, pulling the UM Offense down further. On the other side, the PSU Offense is about average, while the UM Defense is elite (ranked #3), pulling the PSU Offense down significantly. The net matchup result is a significant advantage for Michigan in SD Line Yards per Carry.
SD Sack Rate
In SD Sack Rate (a contra-metric), the UM Offense is well below average, while the PSU Defense is slightly above average, which worsens the UM Offense effective Sack Rate. On the other side, the PSU Offense is slightly above average, but the UM Defense is elite (ranked #3), pushing the PSU Offense up as well. In the end, the net matchup result is a sizeable advantage for Penn State in SD Sack Rate.
Passing Down Matchups
Last, but certainly not least, are the Passing Down matchups, which once again are split between Michigan and Penn State, with the net result being a push or near negligible advantage for Penn State. What is particularly promising for UM is there is some indications that UM will not need to be any more concerned about its pass protection in Happy Valley than in other locales, but the same does not hold true for PSU.
Passing Down S&P+
The aggregate Passing Down S&P+ metric is an uninverted metric, meaning that higher values correspond to both higher-ranked offenses and defenses. Here the UM Offense is above average, but the PSU Defense is top ten, which brings the UM Offense to below average. The converse hold true on the other side, where the PSU Offense is above average while the UM Defense is top ten. The net matchup result is a negligible advantage for Penn State in Passing Down S&P+.
PD Success Rate
In PD Success Rate, the UM Offense is about average, but the PSU Defense is elite (ranked #4), which pulls the UM Offense down considerably. On the other side, the PSU Offense is well above average, but the UM Defense is elite as well (#3), which pulls the PSU Offense down as well. The net matchup result is a sizeable advantage for Penn State in PD Success Rate.
PD Explosiveness
In PD Explosiveness (IsoPPP), the UM Offense is above average, while the PSU Defense is about average, which leaves the UM Offense about the same. On the other side, the PSU Offense is also above average, but so is the UM Defense, which pulls the PSU Offense down closer to average. The net matchup result is a slight advantage for Michigan in PD Explosiveness.
PD Line Yards per Carry
In PD Line Yards per Carry (LYPC), the UM Offense is perfectly average, while the PSU Defense is top ten, which drags the UM Offense down to well below average. On the other side, the PSU Offense is well below average, and the UM Defense is above average, which takes the PSU Offense down a notch. The net matchup result is a sizeable advantage for Penn State in PD LYPC.
PD Sack Rate
In PD Sack Rate (a contra-metric), as everyone should know by now, the UM Offense is well below average, but the PSU Defense is only about average, which leaves the UM Offense about the same. On the other side, the PSU Offense is bottom ten, while the UM Defense is above average, which makes it worse for PSU. The net matchup result is a significant advantage for Michigan in PD Sack Rate.
Conclusion
So at this point you may have a sinking feeling about these metrics matchups. Overall, UM has the advantage in only 10 of the 26 metrics, which sounds terrible, but is only 3 short of an even split with #3 overall S&P+ team. So, the margin stands at Michigan –10.1 points and a 28% win likelihood. What is going to be key is the strategic approach that Michigan takes in order to influence, if not impose its will, on the flow of the game, such that it will favor the niches in which Michigan holds its precious few competitive advantages.
So how does UM beat the highest caliber team it’s seen this season? On the road? Simple: don’t turn the ball over. One of the most significant characteristics of this team is in its expected turnover margin. These are turnovers that one would expect the average team to gain based on UM’s proficiency in defending passes, and inducing the other team to to put the ball on the ground. Alas, the variance of this metric is quite high, but it’s where the variance is high that the greatest upside can be captured. Also PSU has the stats, as well as actual turnovers (+12 on the season), to back a high expectation of being on the plus side of the turnover struggle themselves.
Other import considerations related to Turnover Margin are Field Position and Finishing Drives. Turnovers is one of the key reasons PSU ranks near the top in Field Position in both offense and defense, as well as Finishing Drives on defense. PSU has gained two turnovers this season from opponents in the Red Zone, which contributes to the low Finishing Drives mark on defense. In other situations, the only thing that flips field position better than a turnover is a long kick return.
Moving on though, one of the more important considerations is that the 3 Rushing metrics in which UM shows an advantage are all associated with Offensive Line / Defensive Line performance. The issues that the Wolverines have been having with its Offensive Line are not be so glaring when it comes to rushing, and more so lately due to the noticeable shift in schematic approach and rock-toting responsibilities. I would expect to see a blend of both Zone and Gap blocking schemes, which if executed effectively, will reap huge RPS benefits, when catching PSU guessing the wrong way.
What’s more, on the other side of the ball, the PSU OL remains a persistent weak spot. Yet, PSU is not a team that runs the ball to win games. If the Nits’ opponent packs the box to contain Saquon Barkley, that just enables them to take their shots down the field as they are wont to do. Of course, the fact that PSU will be facing the UM DL will mitigate much of that situation, and it may not be necessary for Michigan to crowd the box to contain Barkley – for the most part. Explosive plays can be expected, as the metrics as well as past history playing against Barkley bear out, and if PSU wins this game, it will be through that avenue.
Thus, in the end, a low scoring game benefits Michigan’s chances to escape with win, a game in which Michigan makes few costly mistakes, minimizes explosive plays, executes its run blocking with ruthless precision, works in some play action and otherwise solves its problems with aggression.
So, that concludes this week’s Five Factors Matchup Analysis!
Yours in football, and Go Blue!
October 19th, 2017 at 3:40 AM ^
October 19th, 2017 at 2:09 PM ^
...and let's be careful (with the ball) out there.
October 19th, 2017 at 10:06 AM ^
I'm irrationally confident UM will keep this one close.
October 19th, 2017 at 11:54 AM ^
I don't think it's that irrational. Every loss, except 2015 OSU, has been by 7 or fewer.
October 19th, 2017 at 1:02 PM ^
Michigan has been in every game since Harbaugh started except 2015 OSU. Harbaugh's first game, 2015 Utah, was probably the only other one where it felt that we were always playing from behind.
October 19th, 2017 at 3:32 PM ^
The FSU game felt out of hand at one point. No Peppers, no Butt, losing 20-6 at halftime, and 27-15 in the 4th quarter.
But besides those 3 game, 2015 OSU, Utah, and 2016 FSU, no game got out of hand. Only other game where I had no confidence we'd win was unfortunately 2017 MSU. Once they scored that 2nd TD, and the rain started to come down, you could just feel that our offense wasn't going to win us that game.
October 19th, 2017 at 11:26 AM ^
Basically, we're a 10(ish) point dog.
October 19th, 2017 at 11:29 AM ^
Barkley wherever he goes. If he splits out wide, Bush goes with him..etc..
October 19th, 2017 at 2:12 PM ^
That didn't work out so well with Sparty. They constantly got Bush out of the box by drawing him to the outside. I would rather have a safety or nickel on Barkley and keep Bush inside to rush or protect. Don Brown will figure it out.
October 19th, 2017 at 2:17 PM ^
To the contrary, I think it would be better to have Bush spy McSorely. I think if Don Brown can make sure the DE's key on Barkley when he reaches the mesh point with McSorley, it will force McSorley to keep it. Then Bush will handle McSorley once he breaks to the outside.
The trick will be more in handling Barkley on swing passes, screens and the dreaded wheel route. Since PSU runs a lot of RPO, I think Bush may be delaying his assault until he gets a read on whether it's going to be a run or pass.
October 19th, 2017 at 12:28 PM ^
Serious question...no snark, no trolling...a legitimately serious question wondering what you guys think:
(jokes about 2016 @ OSU aside) If Michigan wins at Penn State on Saturday, it will be the program's biggest road win since...
October 19th, 2017 at 1:09 PM ^
Reddit has a "fun" thread about Michigan not beating a team on the road that finished with less than 5 losses since 2007 at Illinois. As far as your question it would probably be 2010 at ND or 2006 at Penn State.
EDIT: I forgot to include at #2 ND in 2006 in this response. I believe that was the last road win against a ranked opponent.
October 19th, 2017 at 1:09 PM ^
2006 Penn State was a big win.
I was also pondering our biggest road win. In my memory I came up with 1996 at OSU. They were ranked #2, were undefeated at 10-0 and would go on to beat undefeated #2 Arizona State in the Rose Bowl. OSU finished #2 that year.
Others? Have we ever beaten a number 1 team well into the season?
October 19th, 2017 at 2:20 PM ^
But how many teams less than five loss teams have we faced on the road since 2007 and lost? Ohio State every other year?
October 20th, 2017 at 12:54 AM ^
But we are 0-14 against ranked teams on the road since Bo died in 2006,
October 19th, 2017 at 4:40 PM ^
This is no surprise seeing as how Michigan lost 5 or more games in 6 of the 9 intervening seasons since 2007. In the 7 combined RR & Hoke years, M had a record of 46-42 for an average of about 7-6. Not very competitive, to say the least. Michigan has only been competitive against stronger opponents since Harbaugh arrived, and opportunities against such teams on the road have been limited. The wins will come. If it doesn't happen Saturday in Happy Valley, the next opportunity will be at Camp Randall, at which point the Badgers will probably be an undefeated top 5 team already looking past Michigan to the B1GCG.
October 19th, 2017 at 11:40 PM ^
Also, this is kind of a strange stat to pick on. Since 2007, Ohio State has only had 4 opportunities for top 10 opponents on the road. Florida State has only played in 6 such opportunities. So to compound your point further, it makes no sense to cite a stat such as this when the occurence is rare, specifically here because of Michigan's success (lack thereof) over the last decade.
For a more current reference and since the head coach actually does factor into this heavily, Harbaugh has only had one such game in his tenure at Michigan - 2016 Ohio State. Odd that an Ohio State fan would bring this statistic up in this thread, state that he/she isn't trolling, but then again maybe it isn't coincidence. There is the opportunity for three such occurences this year - which is great for the B1G. This is also very likely to be the most down year we see for several years under Harbaugh.
October 19th, 2017 at 11:52 PM ^
Further data - since 2007 Michigan has only played in 4 such games. It isn't like we're biting off a huge sample size here, so I think a disclaimer should be made when someone chooses to cite this stat.
October 19th, 2017 at 1:03 PM ^
It will be the first win on the road against a ranked opponent since #11 Michigan beat #2 Notre Dame in September of 2006.
October 19th, 2017 at 1:16 PM ^
That is depressing. Holy smokes. I hope the boys change that stat on Saturday. It would make my weekend a lot more enjoyable.
October 20th, 2017 at 12:52 AM ^
How many times as a #1 or #2 team lost at home in the last ten years? (Well, it happened earlier this year...hmmm....who was that?)
I think a better question would be: when was the last time Michigan beat a top 25 team on the road? (EDIT: It was at ND in 2006...FFS...It's been over 10 years since Michigan beat a top 25 team in their house. Put another way, since Bo died Michigan is 0-14 against ranked teams on the road.)
You can take any team in the country and go decades between times they beat #1 or #2 on the road.
October 20th, 2017 at 9:10 AM ^
It wasn't the Swamp, but playing Florida in Florida counts as a road game in my book.
October 19th, 2017 at 1:35 PM ^
care about is the one showing... We Won !
October 19th, 2017 at 2:12 PM ^
Sounds like a tall order to not turn the ball over, execute the running game, limit big plays, keep it low scoring, not have a ton of mistakes, and be successful with play action. If just playing good defense was all we had to do then I'd be down with a win. But, I haven't seen us do all those things in one game yet, with the possible exception of Florida. But, PSU is better than FLA, and we have a different QB.
October 20th, 2017 at 12:13 AM ^
It may be a long night though our defense could reek havoc on PSU. If so, it will be a close win or a close loss. I'm hoping for a close win. I would hate to have to watch a complete blowout by PSU.
October 20th, 2017 at 6:35 AM ^
October 20th, 2017 at 7:51 AM ^
Call me an irrational optimist, but I think Michigan wins a fairly routine (if hard fought) B1G game without needing outrageous luck. Two reasons:
1) I just don't think the Wolverines' offense is as terrible as its last two games. O'Korn (I emphasize, a 5th year senior with three years in the program) has been playing like a bad 3* freshman, and Peters is apparently not pushing him for playing time. Is that really the state of the QB position at a top level CFB program with THIS coaching staff? Nope, I don't buy it; I think O'Korn improves to servicable this game, especially if Penn State sells out against the run. Similarly, are the WRs going to continue with the dropped balls (whether they get intercepted or fall incomplete)? Are the backs going to continue to put the ball on the ground (and not recover it)? Don't buy that either. I'm not saying Michigan is going to magically pull a +5 turnover ration out of its...nether regions; I'm just saying that if it stays +/- 1, as it is likely to do, and the WRs fix the dropsies, it will feel like we are getting 3-4 more productive drives per game.
2) I did not realize until reading the Conclusion of this post just how much turnovers have favored Penn State this year. As Ecky argued, that has driven much of their statistical success. Life will seem a whole lot harder for them if turnover luck falls back to the +/-1 peak of the bell curve.
October 20th, 2017 at 10:50 AM ^
you still have one of the best screen names on MGoBlog
October 20th, 2017 at 11:35 AM ^
The net matchup result is a considerable advantage for Indiana in Passing IsoPPP.
October 20th, 2017 at 9:41 PM ^
Thanks - corrected.
October 20th, 2017 at 12:40 PM ^
I'd rather see Michigan play at Penn State at night than Iowa. Just like MSU seems to have our number we have Penn States. We win 21 to 10.
October 20th, 2017 at 12:40 PM ^
I'd rather see Michigan play at Penn State at night than Iowa. Just like MSU seems to have our number we have Penn States. We win 21 to 10.
October 21st, 2017 at 3:23 PM ^
Nice analysis. You conclude, "a low scoring game benefits Michigan." How does UM not scoring a lot benefit it? Don't we just want a low scoring game by PSU?
October 22nd, 2017 at 12:50 AM ^
How did it go?
Comments