Pat -- I think Ainge should be +0.7, if my artsy brain is doing the math correct.
Paul | Email | Homepage | 06.21.06 - 7:25 pm | #

Really interesting stuff. This is why this site is my home.

Pat, without demanding too much of you, is there any way to cross-reference this data with D-line play, such as sacks, pressures, knockdowns? I doubt statistics that detailed can easily be found if at all, but even with some "quick and easy" quantitative data it would be really interesting to see if you could find an statistical link between QB mobility > D-line play > secondary play. We've been talking about this for a week or two.

I briefly considered attempting to look up some of this stuff, but to be honest numbers and tables make my head hurt. Either way, great post.
Brad | Email | Homepage | 06.21.06 - 7:26 pm | #

Too simple. Remember that patsies can skew numbers. In order words, Henne's number this year against Central Michigan, Ball State, Vandy and otehr simple wins will make his average yards per attempt high, etc. So strength of schedule can make it look like we slowed someone down, whenever in fact, putting 50 points up on Central Michigan makes the average way too high. Maybe, perhaps, your analysis is not simple enough. We lost against most mobile quarterbacks, period. Analysis done. Good defensive line play will make the difference so the corners and safeties do not have to cover while the d-line counts one-Mississippi, two...
Yago | Email | Homepage | 06.21.06 - 8:22 pm | #

Paul - thanks, fixed.

Brad - that's what we're going to look into. Did the so-called mobile QB's actually hurt us with their legs (i.e. higher percentage of passes completed outside of the pocket) or was it merely the threat of the run? Was it poor contain by some of our young defensive ends? How did we change our blitz strategies for better or worse? That's the kind of stuff we're going to look into.

Yago - Agree about the patsies, but Henne also had to face the PSU and OSU defense. And we probably could make it simple and just say "mobile QBs hurt ND", but it's a long off-season and we figured it might turn up something interesting.
Pat | Email | Homepage | 06.21.06 - 9:55 pm | #

OK, first of all, this is great! I can appreciate a well-thought analysis, and this is a prime example.

Next, for what it's worth (not much), I think that it would be a valuable exercise to focus on the exceptions rather than the data that confirm your hypothesis. Here's what I mean:

Assume that we won because we held the opposing QB to less than the season average yds/pass attempt, and assume that we lost when we didn't. I believe that will raise a number of good questions that will be more telling than drilling into the statistics of Troy Smith and Drew Stanton. For example:
Why did Lamar Owens (mobile quarterback) perform so poorly compared to his average, other than the obvious reason that we were Navy's toughest opponent? I'm going to guess that completion percentage and increased attempts in "trash time" will assist the explanation...but how did we prepare differently for this game than others?
Why didn't we lose to Isaiah Stanbeck with the phenomenal improvement over his average that he made that game? Could this be a case of ND being one of the poorer secondaries that UW faced? (Oh, wait, Stanbeck's number should be +2.3, just noticed that.)
Also, why did Ainge have a better day with regard to his season average? Again, an indication of the defenses they faced???
I am willing to write-off Edwards and Beck as statistically insignificant.

As I write this I'm starting to think that this measure is more indicative of our opponents' opponents' defenses than anything in particular about ND. Maybe it would be better to just rank our opponents by their yds/pass attempt and draw a line between the wins and loses. There's a very clear line at the 9 yds/pass attempt mark. This will highlight Smith, Stanton, Stanbeck and Leinert, so the enigma is still around Stanbeck, but if we identify the key to that win, then we figure out the "recipe" for beating a mobile quarterback, even if he has success throwing the rock.

Again, great analysis! Looking forward to the follow-up...
Roke | Email | Homepage | 06.21.06 - 11:34 pm | #

I disagree with doubting this.. who cares if these "mobile" qbs torched vandy etc... they torched us and it was infuriating to watch. Fast DEs are a neccesity as well as versatile LBs...


Great Post.
1000steps | Email | Homepage | 06.22.06 - 12:00 am | #

Roke - As for Stanbeck, I think what really hurt him/Washington was turnovers. The first drive he passed the Huskies down to the 1 yard line before the receiver fumbled and we recovered. Later he drove them down to around the 20 before throwing an interception in the endzone. I think there was a fumble later on as well. That and the fact that their offense wasn't very good. My guess is ND tried to take away the running game (they only had 41 net rushing yards) and force Stanbeck to beat ND with his arm. That led to a few deep passes being completed, which helped to bump up Stanbeck's numbers.
Pat | Email | Homepage | 06.22.06 - 8:16 am | #

Pat I agree that the turnovers hurt Stanbeck. I think that same reasoning can be used for Ainge (Tommy Z's INT TD). Also I suspect that Wash and Tenn had difficult years on D as well.

As for Owens, I suspect that he had the least amount of pass yards for the season compared to the rest of the folks on the list. Also notice that his YpPA for the season is the higest by over a yard and he still had the 2nd best game (YpPA wise) vs ND amoung teams that ND beat. I suspect that his numbers are at least a bit skewed by the fact that a lot of defenses are surprised when Navy throws and when they do throw it's looking for a long ball either because they're behind, it's 3rd and a mile, or it's a trick play going for 6.

While I'd suspect that ND had fewer sacks against the more mobile QB's, I'd agrue that the sacks may also indicate the quality of the O-Line that ND faced. I think that would be and interesting look too. How did ND's D-Line do for sacks/QB hurries compared to the team's average? That might be the measure of how good/bad the D-line's pass rush play was.
OITLinebacker | Email | Homepage | 06.22.06 - 8:36 am | #

To be honest, I really think that Stanbeck just flat out had a great game (other than the turnovers). From 20 yard line to the other 20 he was pretty phenomenal in that particular game, and the fact that ND got up ahead gave him a whole second half to air it out.
Brad | Email | Homepage | 06.22.06 - 8:40 am | #

P.S. Excellent idea for a statistical approach to d-line play OITLinebacker.
Brad | Email | Homepage | 06.22.06 - 8:42 am | #

We should be happy we faced the QB's we did, and not Vince Young...he would have had a field day with our defense as he did with almost any defense he played last year.
Anonymous | Email | Homepage | 06.22.06 - 9:00 am | #

You mean to tell me that Vince Young is really, really, really good?
Phil K. | Email | Homepage | 06.22.06 - 10:01 am | #

Apparently he's a pretty decent runner, but has a weird throwing motion. A team with a QB like that shouldn't be more than a .500 team, right? How'd Texas do last year anyway?
IrishHoffy | Email | Homepage | 06.22.06 - 10:18 am | #

I hate to say it, but the mobile QB's remind me of my techmo bowl style EA NCAA 2006 teams.

In other words, I try to pass as much as possible bc its fun, but the game is a pain in the arse and doesn't allow for easy passing (how come I average 55-60% completion, and the computer averages 75-80% wtf?). Anyways, in the video game, bad passing is never a problem because I know I can always run for a first down when I'm stuck in a 3rd and 10 or 4th and 10 situation.

I got the same feeling seeing these mobile QB's play against ND. They seemed to consistently pick up the yards they needed on 3rd down to keep drives alive. Sometimes they did it by running; sometimes they did it by passing, when the threat of their running left LBs spying & frozen and receivers wide open.

At least that's how it seemed. Not sure how you could analyze to prove that. . . .
atepesm | Email | Homepage | 06.22.06 - 10:45 am | #

Pat, excellent point on the Washington turnovers. I did some research and found that for the QBs that improved their YpPA despite ND winning, like UW and UT, ND was able to create a positive turnover ratio of +3 and +2, respectively. Now, against Stanfod, which I did deem statistically insignificant, we actually had a negative turnover ratio of -2 (Stanford played a turnover-free game), but we were still able to pull off the win. I think we can all agree that the Stanford game could have gone either way...

Bottom Line: Turnovers can offset offensive statistics in ultimately determining the outcome of a game. (Nothing revolutionary here...)
Roke | Email | Homepage | 06.22.06 - 10:47 am | #

Good analysis. It is nice to see you back at number crunching!

Now, you need to plot yards per pass vs. season rushing total for their Notre Dame game and the yearly average with points, not lines. The visual will establish if there was a difference better, and allow to detect anomalies from your assertion.

Sean
Sean | Email | Homepage | 06.22.06 - 11:24 am | #

Uh, yeah. Just like he said.
Brad | Email | Homepage | 06.22.06 - 11:55 am | #

I think you are missing one important piece. 1) Troy Smith threw for some good yards, but how many times did the defensive line chase him out of the pocket? 2) How many times did the secondary have solid coverage and Troy would scamper down the field for 15 yard?
matt bose | Email | Homepage | 06.22.06 - 12:16 pm | #

I was actually stunned that Troy Smith's YPA was less than 25. I'm still having flashbacks.

One serious point: Remember all of the dropped passes by the Navy receivers? (There may have been some bad throws in there too.) I think Owens might have had a much better day if their execution was a little better.

Great job Pat! I am looking forward to the follow-up.
Ben | Email | Homepage | 06.22.06 - 1:09 pm | #

Another metric you may want to chart: With rushing yardage as the vertical axis you could plot the number of touchdowns thrown and run for vs. ND versus average touchdowns thrown for or run for per game, and this is the most telling description of the quarterback's effect upon a game.
Sean | Email | Homepage | 06.22.06 - 1:29 pm | #

I think the thing with Stnabeck wasn't so much the turnovers, it was the fact that he had about 48% completions. So his YPA was way up there because he was constantly making 30 yard passes from their own side of the field, only to have that followed up by a dive into the line for no gain and two incompletions. Much better to be able to string three or four 7 yard passes in a row than to go 1/3 for 32 yards. Plus, he hasn't had 6 years in the system to learn the offense...
gozer the traveller | Email | Homepage | 06.22.06 - 2:11 pm | #

Give UDub time, my friends. It takes a while to install the West Coast Offense. Jeez.
Brad | Email | Homepage | 06.22.06 - 2:28 pm | #

Matt Bose's comment, "How many times did the secondary have solid coverage and Troy would scamper down the field for 15 yard?" reminds me of CWs theory about big plays. . . limit the 10+ yard runs and limit the 20+ yard pass plays.

Well maybe there's a further extrapolation, limit the 5+ yard QB runs or limit the 3rd down conversions or something . . .

Did the mobile QB's have a better than average 3rd down conversion rate? I'm not sure, but it sure felt like it . . .
atepesm | Email | Homepage | 06.22.06 - 6:37 pm | #

You'll never find a thread like this on a tosu website.
Matt | Email | Homepage | 06.23.06 - 2:16 am | #

This post is cruel. Reminds me too much of mid-season analysis.

Only 71 days, 17 hours, 16 minutes, 22 seconds until kickoff at Ga Tech... (HT: kgreen)
Joe | Email | Homepage | 06.23.06 - 2:48 am | #