FanPost

Advanced Metrics suggest only marginal anxiety--but I am still nervous

So, uh, last weekend wasn’t great. Just about everything that could have gone wrong against the Nittany Lions did, and PJ pretty much followed the exact blueprint he needed to beat us. I have been irrationally irritated about the loss ever since, in no small part because of the inconsistencies with which B1G refs have called pass interference all year long. In fact, given my general busyness this week—I’ll be watching the PSU-IU game from my deer blind, actually—I considered not doing an Advanced Metrics post this week. But like any good addiction, after about 3 days away from the stuff, I needed my fix.

So, here we are! Let’s dig in. As a reminder, you can navigate to these two previous posts, in order to refresh yourself on the metrics we’ll be examining.

The Data: Overall Trends

The TL;DR version of this post is pretty simple: the metrics suggest we shouldn’t be too worried about this game. But for a few match-up based reasons—as well as, ahem, some "intangible" ones—I’d be lying if I said I didn’t still have some butterflies. So, what do the data actually say?

When Penn State is on Offense:

Before discussing specific advantages and disadvantages for our Nittany Lions when we possess the ball, a few words are in order about some improvements I have made to my data presentation here. Per some great suggestions in last week’s comment section, I have switched away from presenting raw rankings differences to using percentile differences. The reason for this is straightforward: presenting percentiles avoid crediting/punishing a team for extremely small differences in the metric in question. That is, if the distribution of scores looks like a bell curve (which it normally does), the difference between the 40th and 80th-place team in any given metric could, in real terms, be exceedingly small. Using percentiles, rather than rankings, better addresses this problem (though, admittedly, does not fully alleviate it). Additionally, I tried to delineate which differences between PSU and IU were statistically discernible from one another. Unfortunately, because I am only privy to the season-long averages for each of these metrics (and not the game-specific totals), I can’t calculate standard errors. Without standard errors, no t tests. In lieu of such tests, though, I have now arranged the graphs, top to bottom, to range from the biggest advantages to biggest disadvantages for Penn State. Obviously, the largest discrepancies should be where we focus most of our attention.

Okay, so with that out of the way, here is what the data look like when PSU trots out on offense:

psu-iu-offense


On paper, things look fairly even. However, it is worth noting that Indiana hasn’t played a ton of good offenses. In their nine games, they have surrendered 34 to MSU, 24 to UMD, 31 to Nebraska (who played its 3rd string QB for probably 2/3rds of that game), 24 to Ball State, and 51 to OSU. Those are, frankly, lousy numbers. For perspective, Minnesota gave up 10 to UMD and 7 to Nebraska. And you saw what MSU’s offense looked like against us. Nevertheless, their defensive statistics are bolstered by 3-point outings against jNW and UConn and a shutouts of and Eastern Illinois and rutg.

This caveat notwithstanding, I think there are some fairly strong observations one can make from these data.
First, IU can be had on the big play. Why? Because they play a fair amount of man to man coverage, which surprised me. When Nebraska’s 2nd string guy was in (they rarely passed with the 3rd stringer in there), IU couldn’t cover Spielman—a poor man’s version of KJ Hamler. If they insist on man-to-man, I think we see Rahne taking shots early and often, with hopes of getting KJ on a safety. Same was true for MSU, by the way: their LBs couldn’t cover MSU’s TE, and their CBs couldn’t cover White and Stewart. I don’t think we have the jump-ball guys to challenge them on the edges right now, but I do think Moose could have another nice outing ahead of him.

Second, the above is especially true on standard downs. As you can see, standard-down metrics are clustered near the top of the graphic. We need to take our shots on rushing downs, and we need to run some play action to loosen up that man coverage. If we do, we can hit on some big plays for sure. However, if we get behind the sticks, our lousy passing-down statistics can easily catch up with us.

Finally, one note on the running game that doesn’t show up so cleanly in the metrics. As you can see from the metrics, the running game statistics are pretty even across the board. However, I can tell you that a major area where the IU defense struggles is against QB scrambles. Lewerke ran all over them, and usually not on designed runs. Moreover, both of Nebraska’s QBs ran like crazy on the IU defense—though many of these were zone reads. The reasons for these struggles are many, actually: poor gap integrity on pass rushes, the general vulnerabilities of man-to-man defense, and bad eye discipline on the zone-read plays. But I am bringing this up for one main reason: do not complain about the QB runs when Rahne inevitably calls a bunch of them, particularly early in the game. If we keep running them unsuccessfully late in a close game, then yes, complain. But trust me, there is very good reason for us to want to try these early, and then RPO off of them later in the game.

When Penn State is on Defense:

Perhaps more than any game we have witnessed this season, Penn State exhibits a bevy of advantages on the defensive side of the ball.
psu-iu-defense
Among the largest of these are our rushing-play advantages: we completely outclass IU in just about every rushing metric. Simply put, if IU is breaking off big runs against us, or if they are muscling their ways into conversions in "power" situations, we might as well hang it up. But I really don’t think that’ll happen. This isn’t your older brother’s IU offensive line, fortunately.

So, how do we think IU will respond? Well, pretty clearly, they are going to air it out. I would not be the least bit surprised if Ramsey drops back north of 50 times on Saturday. Now, mind you, he might not pass in a lot of those situations. Indeed, he is a very mobile guy and scrambles with great effectiveness. Given Pry’s stubborn non-use of the spy, you can be sure we will cuss at least once when IU converts a 3rd & 10 on a QB scramble. In any case, expect a ton of passing.

What kind of passing should we expect? Well, I think we will see a ton of short stuff, mixed up with a bunch of deep stuff. Given IU’s rushing struggles, they have been using a lot of short passing and screens as an extension of the run game. They did a nice job attacking the edges of the MSU defense with this strategy, for example. Now, as you may know, the way to counter that plan is to play man-to-man. But, that is where the big play comes in: IU wants to hit enough big plays to keep you out of man coverage. Additionally, they want to run the QB just enough to frustrate you into playing soft zone. That keeps their short passing game open.

I’m not sure which approach is best for us this week: do we sell out against those short passes, and make IU beat us deep? Or do we allow them to dink and dunk down the field? I guess, on balance, I’d say take away the easy stuff, since we don’t want yet another QB to get in a rhythm against us. Nevertheless, unfortunately, our record against running QBs isn’t stellar; and TCF & Co. didn’t exactly acquit themselves well last weekend. Still, given that teams have had some success getting to Ramsey, perhaps the man coverage will allow us to get to the QB before he lets it fly. Remember that this team scored 38 on Nebraska (with Ramsey), so they can move the ball. I think they’ll probably end up with more than 20 points, and people will be pissed. But I also think that, if we don’t score more than 30, we don’t deserve to win this ballgame regardless.

Keys to the game:

PSU wins if…

  1. We get a running game going. If we are breaking off big runs, their man defense is going to be susceptible to the RPO and PA. Even the slightest bit of hesitation means 6 for our man KJ.
  2. Related to (1), we absolutely need to say ahead of the sticks. We definitely should take shots on first down, as this is when IU has shown to be most susceptible. But if it isn’t successful, we need to do everything in our power to get it to 3rd & 5 or shorter. Please, Coach Rahne, one shot per set of downs. Just one.
  3. Clifford doesn’t throw a pick. Ultimately, I think our defense is good enough to win us this game, but we can’t be giving the opponent short fields. IU is a capable offense, and they will definitely make us pay.

IU wins if…

  1. They protect the QB. If we do not pressure Ramsey, it is going to be a long day defensively. First of all, their WRs are quite good. You are going to get sick of hearing the name Whop Philyor (or something like that). Dude is similar to KJ and will probably be the main reason we won’t be able to play man consistently. If we don’t get to Ramsey early, Whop will get open. I actually think Wade can run with him for a few seconds, but GT will get beat like a drum if IU draws that match-up.
  2. They prevent the big play. I have my doubts about Cliff’s ability to carefully plod the team down the field. Hopefully a healthy Cain aids in this area, but the longer the drive, the more opportunities for mistakes. And we can’t afford those.
  3. IU hits on some big plays. If IU hits on these early and gets us out of that man coverage and into soft zone, they can more easily run their base dink-and-dunk stuff. That kind of stuff shortens the game and keeps our offense off the field, which is exactly the opposite of what we want.

You created a Fanpost! Good for you! Any content from a premium site will be deleted once we catch wind of it--as will any inappropriate content. If you simply want to share a link, quote, or video, please consider using Fanshots instead.