Recapping Notre Dame’s tough loss to Georgia with an review of the advanced stats. In a fairly even contest, where did the Irish come up short?
Bear with me for a quick trip into an alternate universe…
Notre Dame finally reversed their streak of losses in close games, gutting out a tough 22-20 win over a tough Georgia defense in a back and forth affair. The Irish defense was able to mostly contain the Bulldog running attack and force true freshman Jake Fromm to beat them. And while Fromm performed admirably and limited his mistakes, he wasn’t able to move the ball consistently against the Notre Dame defense. Kicker Justin Yoon bounced back in a big way with five field goals, including the game winner, after going 0-for-2 in the first game against Temple.
There are still many question marks for this team – Brandon Wimbush had an inconsistent day throwing the ball, and the Notre Dame offensive line went from unstoppable against Temple to unable to open up any running room against the Bulldogs. But there’s time for a young quarterback to improve, and the line won’t face many fronts with the combination of size, athleticism, and speed of UGA. The defense will have to pass much tougher tests against offenses with better passing attacks and lines, but looks to be head and shoulders above where it was in 2016 – it’s been a long time since the Irish have had competent units on both sides of the ball. The schedule now sets up nicely for ND, who will be favored in each game until a showdown with USC at home.
We weren’t all that far away from a write-up like that one after Saturday night. But how different does it feel this week? Our football fandom is constantly sitting on a seesaw, waiting for enormous athletes to jump on our side or the other and send our hopes diving toward a collision with the ground or catapulting up into the air. And if you’re a Notre Dame fan, even though there was a lot to like, it’s easy to be nauseous from the constant up and down, which lately has included a lot more landing with our butts on the ground than time airborne. There’s a strong desire to just get off the seesaw and go do something else (the Irish have played more one-possession games (42) than any FBS team since 2010), but I’m an optimist and a sucker, so here we go.
The Basics
This was an even contest that was mostly controlled by both defenses. The Dawgs outgained Notre Dame by a decent margin, but add in the penalty yardage and it’s almost dead even. UGA had a hard time keeping their hands away from Irish facemasks, and racked up 12 penalties for a whopping 126 yards. The Irish had a few dumb plays of their own, but account for penalty yardage and this contest was essentially dead even.
Explosiveness
A week after gashing Temple with one long run after another, the Irish were completely bottled up by the Georgia defense. The Bulldogs had the size and strength to clog runs up the middle and athleticism to contain attempts to attack with speed to the edge. Notre Dame’s longest runs of the day was an 8-yard Wimbush scramble and a Josh Adams 7-yard scamper. Those were the only rushing gains longer than 5 yards on the evening – Georgia completely shut down the explosive rushing game. I won’t beat a dead horse, but some carries for Dexter Williams – the guy sporting a 20.7 yards per carry average in this young season – may have been a good idea.
The Irish were able to find more success through the air – unfortunately just not on a regular basis. On the opportunities where Brandon Wimbush had time he was able to find some Alize Mack, Cam Smith, and Josh Adams downfield with some space to run. But in his second start Wimbush was constantly pressured and forced to make quick reads, and a few big opportunities downfield were barely missed.
Georgia faired a little worse breaking big plays through the air but better on the ground. The Notre Dame run defense was stout and disruptive for much of the game, but the few lapses they had went for 30+ yards and were able to tilt the field. The Bulldogs found success motioning backs across the formation and making the Irish cover the run to the edge and up the middle, and caught Notre Dame out of position enough to make a difference.
The passing defense continues to limit long gains downfield, although the secondary hasn’t faced a murderer’s row of opposing passing offenses. Still, Mike Elko has forced inexperienced quarterbacks and less threatening receivers to make plays, and to his credit and his players, opponents haven’t been able to make the Irish pay yet. There were a few drops by UGA receivers, but the majority of long gains came where Irish defensive backs were in the right position and simply had receivers make a better play on the ball. This is a defense that through two games that isn’t perfect, but looks much more fluid, cohesive, and fundamentally sound, and will be tested further in the coming weeks.
Efficiency
Sustaining drives and being efficient is difficult when the average distance to go for Notre Dame on 2nd down was 8.7 yards, and 7.7 yards to go on 3rd down. These were the Chinese finger traps the Irish offense had to avoid – passing downs with a young quarterback against a strong defense. It’s easy to criticize play-calling, and I agree that some more creativity and attempts to take advantage of Georgia’s aggressiveness were needed, but little was working regardless of how the Irish tried to move the ball.
Defensively, you couldn’t ask for a much better performance against a dangerous rushing attack. With a true freshman making his first start in a semi-hostile/semi-supportive environment, the Bulldogs were going to try to ride their talented backs and wear down the Notre Dame defense. And down to the final defensive possession, the Irish held their own against Georgia’s vaunted rushing attack. It remains to be seen just how strong that attack will be (is the Bulldog offensive line still bad, or now passable?) – but by success rate, this was Georgia’s 3rd worst rushing performance in their last 15 games.
Last week I predicted that leverage rate and the ability to avoid passing downs would be critical, and the Irish lost both of that battles. Both quarterbacks struggled on passing downs, but Notre Dame faced 10 more of them than Georgia. The Bulldog defense was simply the best unit on the field, and in the end controlled the game enough to scratch out a narrow win.
Finishing Drives, Field Position, & Turnovers
In an even matchup, the Irish lost a few more small but costly battles in the final three factors. Despite six scoring opportunities inside the Georgia 40, the Irish walked away with just one touchdown. This is the cost of inefficiency – even if you can break a big play, or string together a few successful plays, or the opponent commits a dumb penalty, it still requires additional successful plays to convert red zone visits into touchdowns.
Notre Dame barely lost the field position battle – starting drives on average on their own 28 while UGA’s average starting field position was its own 32. Tyler Newsome had another strong day, and Justin Yoon was perfect after his close misses in the opener. But the Irish coverage units gave some yards away, in particular on kickoffs.
The ND defense also did a strong job tightening up in the red zone, and added two takeaways to the ledger. One was a gift from a botched handoff, but Drue Tranquill had a fantastic pick that gave the Irish offense a chance to put a scoring threat together before halftime. Both Wimbush fumbles were tremendously costly, but it’s hard to fault him too much for them after the number of hits he took in this game. I don’t think ball security with fumbles will turn into a Golson-like issue for Wimbush, but after putting three on the ground last week it’s worth keeping an eye on.
Big Picture
This loss lowers Notre Dame’s ceiling for 2017 and leaves us frustrated, but mostly confirms many things we probably thought before the season. This was most likely going to be a top-20ish team, and against this schedule that projected to translate into something between 10-2 and 7-5 without some excessively good or bad luck or coaching blunders. This chalks up another one in “close loss” category and incrementally decreases the number of expected wins, putting some of the uglier scenarios in play. But the most likely outcome of 8-4 probably included a loss like this one.
And the way it unfolded wasn’t particularly surprising – we expected a close game, hoped for more against a talented yet incomplete team, and received a result that didn’t surprise anyone. The offense struggled more than expected, but the defense balanced that out by out-performing expectations. A Notre Dame quarterback in his second start struggled against a talented defense, and an offense under new leadership and playcalling had an early clunker. The defense is much improved but not to a place yet where it can (or should be expected) to be good enough to win games with the offense struggling.
The big picture piece of this that stings the most – that “we are who we thought we were” which is probably a good and not close to great team, according to this measuring stick. That’s been a constant theme of the Kelly era, and so have the inability to put together a strong offense and defense at the same time. I’m obviously a huge advocate for advanced stats, but we’re definitely to a point where BK may be an advanced stats darling – a coach good enough to avoid blowouts and win enough not to get fired, but not to beat strong teams.
On to Boston College?
Each game early in the season is an attempt to learn as much as we can about how good or bad this Notre Dame game truly is, but we’re still finding a lot out. Boston College will provide some more answers – mostly for the offense. The Eagles will have the usual solid defense, and probably try to replicate Georgia’s gameplan of forcing pressure and the passing game to beat them. Can the offensive line return to a better form? Can Wimbush and the receivers step up and find more consistency to make defenses pay for aggressiveness against the run?
The Eagles are playing faster on offense (9th in adjusted pace), and I applaud them for trying something new. But so far the results have been more of the same, and failing fast isn’t a great strategy. I’m still not sold that the Irish defense won’t be a liability against more complete offenses later this season, but this week provides another showcase opportunity for Elko and company to create havoc and turnovers. The defense should be able to attack without much fear of the Boston College passing game (125th in passing success rate) or rushing attack (111th in Rushing IsoPPP and 123rd in stuff rate). Hopefully we can continue to see young players improve, coordinators getting more and more comfortable, and mental toughness from a team and coaching staff that preached it over the entire offseason.
“I’m obviously a huge advocate for advanced stats, but we’re definitely to a point where BK may be an advanced stats darling – a coach good enough to avoid blowouts and win enough not to get fired, but not to beat strong teams.”
This is really interesting to think about. The basic question would be: is it just due to luck or is there something more behind it? Because the idea of being good with advanced stats is that you are pretty close in general to most of the teams you play but can’t quite win enough based on those stats right? With a large enough sample size, we’d expect it to generally even out (i.e. with certain stats/rankings we’d finish more or less with the expected wins) but in Kelly’s case it just never does (perhaps still too small sample size but perhaps not the only problem – hence the question: luck or other things?).
There’s been a lot of good writing about one-possession games, and the general consensus seems to be across both CFB and the NFL one-score games should be fairly random over time – i.e. close to .500. If it’s not, it could be attributed to some combination of coaching, QBs, and special teams (probably easier to see this in NFL where there’s more consistency in franchises over time). But generally the skill is in avoiding playing close games as opposed to record in 1-score games, and as noted, in the BK era no FBS team has played as many one possession games.
The sample size part is a challenge, but what I was getting towards is that advanced stats are a helpful judge to see beyond W-L record – i.e. if you get screwed with a tough schedule and play really good teams close, you’re probably better than a team that’s undefeated against cupcakes and maybe hasn’t even won by much against them. That’s overly simplistic, but BK has had several seasons now with a lot of these “second-order losses”.
Finally, it’s worth noting that for whatever reason the BK era has also been really streaky. At one point we won 15 of 18 one score games, and we’re now 5-12 since then and in the middle of a pretty awful stretch (1-9 in our last 10 I believe). But I think there’s enough evidence to have a compelling argument that advanced stats are no longer really defending the team’s quality as much as an indictment that a team of this quality is consistently under-performing in the wins column.
Tacking on to Mike’s point about frequency of one score games… As he noted above, since 2010 Notre Dame leads the country in the category with 42 games, going 22-20 in those games. The rest of the top 10 consists of Louisville, Pitt, Utah, Iowa, UNC, Va Tech, BYU, Navy, and Nevada. If you think about how those teams have performed overall, intuitively it seems to track pretty well with how we’ve done. Mostly middling seasons with an occasional burst.
On the flip side, of the teams that I would consider perennially elite since 2010 – Alabama, Ohio State, Oklahoma, Clemson, etc. – Ohio State is the highest on the list with 30 one-score games, which ties them for 50th. Alabama is lowest with 19, which ties them for 121st. If you’re wondering, OSU is 21-9 and Bama is 10-9. Very different win rates.
Just a few data points, but all that seems to support the idea that being better at winning one-score games is less important than staying out of one-score games.
I am not surprised that we have played more in the Kelly era than anyone. He seems to coach that way. The problem I have is that too many of the teams involved are teams we should not be playing one score games against. I agree with the analysis that over time you should be .500 in one score games. Looking at Bama and OSU provides the information. My assumption is that Bama is playing one score games against peers and as a result is 50/50. OSU, I would bet, is not. That is why they are 21-9. They are letting some bad, or perhaps poorer, teams hang with them. Kelly’s problem is that we are letting teams that should be poorer than us hang with us, and we are still 50/50. That would mean that we are not as good as we want to be. Over 7 years, regrettably, that falls on Kelly.
I begin by acknowledging that my emotions very likely interfere with my rational self after a painful one point loss, so it’s possible everything I write going forward is just crying out in anguish. Still. When talking about losing one possession games, and particularly addressing BK’s 1-9 record in the last ten such games, I think I see a pattern – or at least some commonality – among them. In most, perhaps all, of those games, I’ve almost tangibly sensed the tightening of BK’s sphincter (please go with the metaphor, not the visual, and we’ll get through this unscathed) during the game. Sometimes you can see him on the sidelines looking dazed, distant or removed as the game gets very stressful in the fourth quarter, as if he has no answer and is just willing to let the coin flip of circumstance decide the final plays. I have on many occasions wondered if, during the last few series of an unexpectedly tight game, he is privately processing what a loss here would mean to his job, his career, his legacy. Sometimes you see it in increasingly conservative and frightened playcalling. Sometimes you see it in perplexing clock management. Sometimes you see it in the way he projects stress to the very players who should be finding a voice to focus them positively on the immediate task. I think BK is a stress machine. I think he gets a little paralyzed under its weight. I think I’ve seen his gears seize up a bunch of times during that 1-9 run. And I think he transmits it across the organization. It’s not a new knock on him that he coaches not to lose rather than to win. To be fair, there is a way to do this with confidence and optimism, sure, Saban can do it because he has the athletes, three deep. But BK’s way isn’t due to confidence that he can out-athlete the other team. He instead is conservative because of fear and doubt. It’s a deadly combination and that approach gets terribly exacerbated by bad circumstances during the course of a game. You might ask why he used to have success in those one-score games and now is 1-9 in the last ten? I submit that the pressure on him is much more than it once was. I think he is coaching as a guy who is driven by his increasing stress. To win under this environment, it will take an overmatched opponent, or really good breaks, or a rare set of players who can overcome BK’s stress-induced propensity for frightened and conservative schemes and play calling. We know that stress and expectations drove Ara from coaching. It’s a real thing. It drove Meyer from coaching for a time. I’m afraid Kelly is there, but too stubborn to leave under these circumstances, but equally too stubborn to be the leader ND needs to get the most out of the athletes we have. I hope I’m dead wrong, but my… Read more »
Did you not read the articles about BK doing yoga? Were you living under a rock this offseason?
Good one. I have a feeling his yoga instructor’s choice of music trends more to Metallica than Enya.
I’ll step around all the sphincter talk – never thought I’d have to say that in the comments here – to address what I think are some pretty good points.
One thing I’ll disagree slightly on is clock management – I don’t think that’s a recent development, I think it’s something that’s been an issue for his entire tenure. I’m not sure what to make of this, since clock management seems to pop up for a lot of coaches at a lot of programs. Even the acknowledged “elite” coaches sometimes have head-scratchingly bad clock management moments. My main conclusion here is that clock management is likely much harder and more chaotic on the sideline than it seems from the stands, and one of those things that will break for you sometimes and against you sometimes.
I also don’t think he projects personal stress on the players more or less than any other head coach does. I think it just gets highlighted when he’s losing. Seriously, go to YouTube and look for clips of Saban and Stoops berating players – they’ve done it with impressive frequency over the years, and occasionally with a ferocity that would make purple-faced Kelly seem as threatening as Barney the Dinosaur in comparison.
As far as playing conservatively, though, I definitely think you’re on to something there. Looking at last year, the guy who called for the late punt against MSU or the late field goal against Navy is not the same guy who had true freshman Tommy Rees chuck it to the end zone against Tulsa. It’s actually a 180-degree change from the “get used to it” aggressiveness that he had back then. That I think speaks to a larger concern I’ve had for a while, that Kelly has lost the edge that got him to where he is. I don’t want him to be the riverboat gambler that Chuck was, but I don’t want him to turtle either. He needs to find that balance again.
I think it was the second to last drive we were up three and we went:
Adams to the right for ~ 4
Adams to the right for ~ 2
Wimbush pulls it and goes left for ~ 2
Kelly said if Wimbush had put his foot down and gotten upfield then he’d of gotten the 1st. I watched the play several times and I don’t know but the play calling bothered me. There was plenty of time left. We weren’t gonna run out the clock; not without a long drive.
Worse was that he punted. It worked out ok with the stop, but there were two options if we went for it:
1. Get the first down and continue the drive.
2. Fail, give them the ball on the 20ish. You still probably have to make a stop, and might give them an easy field goal with the stop. If you don’t make the stop, they get in the endzone quicker, and you get the ball back quicker. If you make the stop they might get the field goal. Either way, you get the ball back quicker, and they can’t just run the clock out by getting a couple firsts.
If you are behind with that little time, don’t give up the ball voluntarily. But this radical strategy would probably make you look like crap if you fail, and just feed the FIRE KELLY crowd more than a more conservative loss. That said, I’d rather watch the aggression than the conservative failure.
Doesn’t matter, Kelly punted, and got the ball back, so maybe I’m completely off-base, but it seems like picking up a couple of yards to keep the ball is a risk worth taking.
I didn’t have much of a problem with that possession. They had already had a couple of near-picks and the first Wimbush strip sack in the second half, and it’s not like passing the ball was working well either. With a lead and a defense playing well I don’t mind it.
I think the punt was without a doubt the right call. 4th and 10 is super low probability with how ND had been playing this year and this game on passing downs, and you’re only down two. So true, you still need a stop, but then you need a touchdown once you get the ball back instead of a just a field goal.
I think there were definitely some overly conservative decisions last year (punting vs MSU), not playing aggressively to avoid OT vs Texas, kicking the field goal vs Navy when you still need another score and they’re monopolizing the football. This game I don’t see fitting into that category.
This, so much. I think the combination of losing to Alabama in 2012, the Golson suspension in 2013, and the cratering of 2014 has left a mark on Kelly that he can’t get past. He’s done as a coach.
“I won’t beat a dead horse, but…”
Look, if the dead horse is the need to give the ball to Dexter Williams, then we should be kicking the crap out of that dead horse.
What’s the over/under for Dexter touches on Saturday – 5.5? I think I’d still take the under despite BK’s comments
If he gives him the ball 10 times and he goes for 110, he’ll get more post game questions about last Saturday.
The most reasonable explanation I heard for Dexter not getting any carries. If you put Dexter in, the defense knows it’s a run. They load the box and force a pass. Dexter doesn’t pass block as well as the other two backs and UGA likes to use crossing blitz schemes. They are very confusing, we saw the confusion they created with the o-line. So you would end up with your worst blocker being forced to stay home and block these confusing blitzes.
This makes a certain amount of sense to me. On the other hand, trying the same thing over and over and it continues to fail, doesn’t make much sense either.
Well, we did run roughly 1,000 screen plays without any of them going to our fastest player. If having him on the field forecasts ‘run’, then I can think of no better play than getting a screen out to him.
And I’ll also throw in that while I’m not at all MANBALL-obsessed, if protection is such a problem, then perhaps employing one of our 50 TEs who can’t get open or catch the ball anyway as a blocking FB and sending out our insanely fast RB to go see what he could do out there wouldn’t be the worst idea, either.
It…just…it’s not rocket surgery. It’s football. It requires the minor-est amounts of creativity and flexibility to find something that broadly works. The coaching staff just didn’t appear to have any on Saturday night. Maybe someone who charted each play could argue otherwise, but like you said, it just seemed like they kept doing the same unsuccessful thing over and over until the final whistle blew.
Took the words right out of my mouth. Just because Williams isn’t a great blocker (which IIRC we announced to the world!! WTF would we put that out there?) doesn’t mean it’s automatically a run when he’s in there. Even if it IS true, that is just an indictment on Kelly or whomever is calling the plays.
The best way to counter a bunch of blitzes is by hitting the RB on a few easy catches over the middle. The will back off the LBs and open things up.
Sigh. The one time I can remember Kelly being inventive on offense was moving to a power spread with Zaire against LSU in the bowl game. Apart from that it’s mostly been spread stuff that was effective in 2009.
God forbid we run a two back formation or keep a tight end in on pass protection. Or motion him in or out of the backfield. There are ways to mitigate his pass blocking, but the staff has shown absolutely no imagination trying to get the ball to him.
These numbers make me feel better about the loss (we kept it close despite losing in basically every statistical category), but not necessarily about the season. Is Georgia actually any good? Won’t know until halfway through the season, at which point it won’t matter.
We’ll learn more every week, that’s just how it goes. I agree – last year after Texas we thought “eh Texas is actually alright so this maybe isn’t such a bad loss” and we were very wrong. But BC for example should give us another good barometer for the offense – if we can run effectively again on a pretty solid front then that’d lead me to believe UGA is really exceptional defensively. If we struggle moving the ball again, it might mean that we overestimated our offense a little bit and teams with less talent can still limit us like the Dawgs did (this seems less likely, but would be a really bad sign)
But is BC even good defensively still? They gave up 34 to Wake and 20 to NIU. Doesn’t sound any better than our D. Basically, you don’t really know anything until it’s too late to be used as predictors. And what good does it do to even know how good your team is? Whatever I think I know doesn’t impact the games. I guess for gamblers it matters, or maybe for general optimism or pessimism.
I think I have just become so jaded by ND football. I want to know all the information I can, but purely out of wanting to be familiar with the team I love. I am totally done trying to predict anything, or have opinions on what should or shouldn’t be done, or trying to interpret things in a greater context.
What good does it do to think (know???) your team is actually good and competitive when all that really matters is the W/L. And even then, if W/L really mattered, would I still be an ND fan?
SIIIIIGH
Ha I get the frustration man. It’s imperfect but mostly right more than wrong, but there’s a reason why guys who consistently pick 55% of games right are considered geniuses. On the BC note, even in giving up a bunch of points to Wake they still held them under 4 YPP – four turnovers by the offense were the main culprit, the defense was solid.
I’m not arguing with the advanced stats, or any stats, as being accurate or anything like that. I just question the purpose of it all. What good does it do to actually know anything about the quality of football being played?
Maybe the knowledge really ties the room together.
This is an odd quote, after Saturday – https://twitter.com/PeteSampson_/status/908449937905655808
The replies are funny. Why everyone gotta bring up old stuff?