A BCSCG statistical snippet: Why Notre Dame's Rushing Defense ranking is misleading..

[IH8Orange]

You have posters wishing that Alabama was destroyed by Katrina.

That's probably because most of northwest Indiana looks like a hurricane came off of Lake Michigan, made landfall at Gary, and then beat a path over to Lake Erie. :tongue:

 

[CRMSNtide]

It would be nice to see Bama fresh, rested and get some guys who were banged up in the Georgia game back. That's one thing people forget: Bama had not had a bye week since early October. We went through a gauntlet and couldn't hold up in the last leg of it against texas A&M. Georgia was easily the most intense and physically draining game I've seen Bama play that I can remember. The time off will help our guys get their edge back mentally and physically. Bama didn't play a great game in a lot of ways against UGA either. It very easily could have gone the other way. Thankfully it didn't. We have a lot we can and need to to improve on.

You know, Georgia came into the SECCG game on a roll, biggest game for that school in 30 years, everybody fired up.
Then we throw a pick in the endzone and have a kick blocked giving them a TD.....

And we still kicked their butts, this team has something special going for them, they don't quit.

I agree, we needed the rest and will get some guys 100% healthy, then look out golden domers, can you just imagine if we put it all together and play one of our best games of the year?

 

[Nolan]

I love stats as much as anyone. But stats can be presented in a number of ways to mean different things. Looking at ND overall, I come away impressed by their defense and how hard they play. Those guys are tenacious and they really play together.

Sure, they may win ugly, but still they win. I gotta respect that. ND is the kind of team that will make you pay for dumb mistakes and turnovers. If we don't turn it over and avoid penalties I like our chances a lot. I don't think they can stop our run game for an entire four quarters. But I also think their QB will be a headache at times with his scrambling ability.

 

[MBA_99]

I'd like to see an analysis involving the first team offenses and defenses. I have only watched ND vs Pitt and USC. They sure as heck kept starters in all day. Against Michigan we emptied the bench a good bit; did they? I seriously doubt it. Auburn's first string would've had like -50 rushing yards against our first string D if they weren't back at Bryant Hall shortly after halftime. In all but 3 games I cant recall a second half where Eli wasn't talking like it was late on A-Day.

 

[uafan4life]

I love stats as much as anyone. But stats can be presented in a number of ways to mean different things. Looking at ND overall, I come away impressed by their defense and how hard they play. Those guys are tenacious and they really play together.

Sure, they may win ugly, but still they win. I gotta respect that. ND is the kind of team that will make you pay for dumb mistakes and turnovers. If we don't turn it over and avoid penalties I like our chances a lot. I don't think they can stop our run game for an entire four quarters. But I also think their QB will be a headache at times with his scrambling ability.

I don't think anyone's really saying that they aren't a great team - certainly not me. Any team that makes it to the BCSCG - and undefeated especially - is, at the very least, a very good team.

It's just that they may not be as good as a lot of people - seemingly most of their fans - believe them to be.

As you said, stats can be deceiving. That's kind of my whole point with the differential analysis stuff, especially situational and targeted differentials. Sure, team A may score 43.2 points per game while team B scores 35.7 points per game. That doesn't necessarily mean that team A has a better overall, or even better scoring, offense - especially if the defenses that team A played averaged giving up 39.7 points per game while the defenses that team B faced averaged giving up 27.2 points per game.


The most important thing to do when looking at stats, and trying to compare two different teams by using them, is to use as many common denominators as possible. In some cases, such as comparing an SEC West team to an SEC East, that can be fairly easy and fairly accurate. Of course, you still have to take into account various unknown or unaccountable factors such as when the teams face(d) each other, playing styles, injuries, coaching adjustments, individual match-ups, and emotion just to name a few. Those are things that a statistical analysis cannot really account for, making any statistical analysis flawed from the outset.

That doesn't mean, though, that you can't learn anything from them.

 

[uafan4life]

I'd like to see an analysis involving the first team offenses and defenses. I have only watched ND vs Pitt and USC. They sure as heck kept starters in all day. Against Michigan we emptied the bench a good bit; did they? I seriously doubt it. Auburn's first string would've had like -50 rushing yards against our first string D if they weren't back at Bryant Hall shortly after halftime. In all but 3 games I cant recall a second half where Eli wasn't talking like it was late on A-Day.

That's not a bad idea but difficult to do, especially on a large scale. I prefer what I call "targeted" stats, e.g. looking at only the top ranked or rated teams, only at teams with similar styles, etc. And, on that note, here's another interesting snippet regarding the scoring offenses and defenses of these two clubs...


Everyone knows that both Notre Dame and Alabama have two of the highest rated Scoring Defenses in the Country. In fact, they're #1 and #2, respectively. Of course, that ranking is based upon the number of points allowed per game. While that is a very important stat, I place a bit more weight on Plays Per Point or - for easier reading - Plays Per 10 Points. This is the number of plays the defense is forcing their opposing offenses to run in order to score 10 points, on average. Interestingly enough, Notre Dame and Alabama again rank #1 and #2, respectively, in this stat. Alabama is forcing its opposing offensive units to average 56.26 plays to get 10 points while Notre Dame is forcing a very impressive 61.05 plays. The third ranked team in this stat, Florida, is requiring 50.90 plays - and those three teams are the only teams that top 50 plays. In fact, there are only 20 teams in the Country averaging forcing more than 35 plays for their opponents to net 10 points. Of course, that's just raw stats. What about the differentials?

I have several different categories that I like to break (or split, hence the term "splits") teams and units into for comparisons. One of my favorites is the Top 5 category, especially relating to unit performance. In this case, I took the Top 5 Scoring Defenses (overall - per game) that each offense faced and used those aggregate numbers to form the differentials for those units. Likewise, I also used Top 5 Scoring Offenses, Top 5 Scoring Efficient (Plays Per Point) Offenses & Defenses, Top 5 Total (Total Yards) Offenses & Defenses, Top 5 Efficient (Yards Per Play) Offenses & Defenses, and Top 5 Driving Efficient (Drives Per 10 Points) Offenses & Defenses. I won't overload the post with all of the stats from all of those breakdowns - I'll give a summary in a minute - but here are some of the interesting notes from those first four categories, the Top 5 Scoring / Scoring Efficient Offenses & Defenses...

Versus the Top 5 Scoring Defenses & Offenses They Played
	Alabama Offense	Notre Dame Offense	Alabama Defense	Notre Dame Defense
Points_Per_Game	31.2	19.8	18.2	9.6
Opponents' Avg.	19.9	17.4	35.7	32.8
Differential	11.3	2.4	-17.5	-23.2
Plays Per 10 Points	22	31	34	65
Opponents' Avg.	35	38	19	21
Differential	-13	-7	15	44
TD_Scoring_Percentage	80.0%	50.0%	80.0%	25.0%
Opponents' Avg.	69.7%	59.9%	81.9%	77.5%
Differential	10.3%	-9.9%	-1.9%	-52.5%

As you can see, Alabama's Offense performed significantly better than Notre Dame's Offense when facing each of their 5 toughest opponents to score against. Conversely, Notre Dame's Defense performed noticeably better than Alabama's Defense did when facing each of their 5 toughest opponents to keep from scoring. One thing you'll notice is that Bama's Offense far out-shined Notre Dame's Offense when facing the best teams they faced. Granted, the Defenses Notre Dame faced were allowing slightly fewer points per game and requiring slightly more plays per 10 points than the ones Bama faced but that difference is much, much smaller than the difference in the performance between those two units. Something you won't notice - since it isn't there, mostly because it won't fit - is that the combined Scoring Defense differential ratings for the Defenses that Alabama faced in this grouping is noticeably better than for the ones the Notre Dame faced.

A third thing you'll notice is that Notre Dame's Defense performed much better than Alabama's defense. What you may not notice, though, is that their numbers are a little erratic. For one, the best offense they faced was USC at the end of the year which had several key injuries including along the Offensive Line and at Quarterback, starting a Freshman backup getting his first start against the #1 team in the country. The really weird thing, though, is that the three worst games, from a points allowed standpoint, were against teams not in their Top 5 Scoring Offenses category. That can mean - or be due to - several things. As alluded to before, it could have a fair bit to do with luck, e.g. catching teams without their best offensive weapons. It could mean that Notre Dame is very "streaky" on Defense, i.e. when they're on they're very on and when they're not on they're more average. It could also mean that their Defense has a knack for "coming to play" against the better offenses while slacking off a bit in other games.

The biggest factor that you don't see in that little chart, though, is all about the match-ups. Neither offense has faced a defense ranked as highly as the other team's defense - because there aren't any. Both teams have faced 5 defenses ranked in the Top 30 in what I call the Total Scoring Defense Rating - a combination of the points per game, plays per 10 points, drives per 10 points, and TD percentage raw stats and differentials. Against those teams, Alabama's Offense measured up much better than Notre Dame's Offense. The Defenses, though, present an interesting twist. Alabama's Defense has faced 5 Offenses ranked in the Top 35 in the Total Scoring Offense Rating, with two of those in the Top 10. Notre Dame's Defense has only faced 2 Offenses ranked in the Top 35 in the Total Scoring Offense Rating and none in the Top 20. Notre Dame has, though, faced 7 Offenses in the Top 60 - the same number as Alabama. This category becomes very interesting, though, when you look at where each team's offense ranks.

Notre Dame's offense currently comes in at 82nd in the Country. While Notre Dame's offense ranks a very respectable 32nd in driving efficiency, they are 73rd in plays per 10 points, 75th in points per game, 84th in scoring differentials, and 119th in TD percentage. That ranking means that Alabama's Defense has faced 8 offenses that are more difficult to prevent from scoring than Notre Dame's Offense.

The number of offenses that Notre Dame's Defense has faced that are ranked higher than the Crimson Tide's Offense?

Zero.

Why?

Because, which will be a surprise to some, the strength of this year's Crimson Tide team is actually the Offense, not the Defense.

The Crimson Tide Offense comes in at #2 in the Total Scoring Offense Rating, trailing only Oregon. While Alabama is only 15th in the Country in Points Per Game and 37th in TD percentage, the Crimson Tide is 3rd in Plays per 10 points, 2nd in driving efficiency, and 1st in scoring differentials.


So, when you look at the raw stats and rankings for the Scoring Offenses and Defenses of Alabama and Notre Dame, there are four aspects to consider when weighing those numbers:
1) Alabama's Offense hasn't faced any Defenses better than Notre Dame's but has faced 4 Defenses in the same "zip code", i.e. within 25 spots, and has performed exceptionally well against those teams.
2) Notre Dame's Offense hasn't faced any Defenses better than Alabama's but has also faced 4 Defenses in the same "zip code", i.e. within 25 spots, and has performed exceptionally well against those teams.
3) Alabama's Defense has faced 8 Offenses that are better than Notre Dame's and has performed very well against those teams.
4) Notre Dame's Defense has faced 0, yes Zero, Offenses better than Alabama's and only faced 1 Defense in the same "zip code" and, while they performed very well against that Offense, that Offense was missing their starting Quarterback.


And, just for kicks, if you combine all of the Total Scoring Offense & Defense ratings and differentials and apply them to the Alabama - Notre Dame unit match-ups, the numbers add up to a predicted score of:

Alabama 24
Notre Dame 10

 

[crmsntd91]

Well I guess the only thing I can follow that post up with without looking dumb is ROLL TIDE ROLL!!!

 

[selmaborntidefan]

From an ND fan who has watched every snap this season, I think that is a pretty good assessment. To #2, that is kind of what we try and do. We like to (except early in the game) have long drives over 5 minutes and move the chains. Control possession. That will be much more difficult against your D I believe, unless Golson continues to keep plays alive with his feet.

I'm going to watch your other games that are on You Tube. I found several last night. Stanford next. I don't think it's fair to judge ND by Pitt any more than it is to judge Bama by aTm. In both cases they were follow-ons to emotional wins.

The one thing that surprised me was that the score was a tad misleading. Notre Dame unqeustionably beat OU and deserved to win, but the 30-13 final score was a tad misleading. That's not to take away from a great win in Norman - but the turning point was that bizarre interception by Teo.

One other point: OU was -3 in turnovers this year - they turn the ball over a LOT. Notre Dame was +9. That was a tiny bit misleading because Michigan turned it over six times. Still, Notre Dame got more than they gave up, which is why they're here.We're at +12, and even if you take away the game where Arkansas had five we're still at +7. (Note also we lost a turnover in that game when an interception return for a TD was wrongly ruled to have hit the ground).

Furthermore - MOST of our turnovers (we had 16 in 13 games) were irrelevant. Lacy's fumble against Michigan was on the last play of the half and meant nothing. We had two immediately after the lightning storm in Missouri halted the game that didn't bring any real harm. And the fumble in the Auburn game (that I attended) was in their red zone when we were up 49-0 by the backups. You'd have to tell me which incidental turnovers the Irish had.


Bottom line: whoever wins turnover battle wins.