Rich Rodriguez, Quarterback Archetypes, and Success

Submitted by CollegeFootball13 on

[Ed.: Many excellent diaries of late that all but demand bumps. Larsonlo's Football Fundamentals 101 series is a must-read that will hit the front page at some point and the Mathlete's back, plus Misopogon is analyzing Cam Gordon in crushingly detailed fashion. Here's a quick one that hints at the QB duel winner.]

The quarterback is obviously the single most important position in Rich Rodriguez's offense, so I decided to take a look at some numbers from his time at West Virginia up until this past season at Michigan. What I found, while not exactly scientific, is pretty damn interesting.

Before I get to the summary, let's take a look at the charts... [ed: charts.]

West Virginia QB Charts

Year
Record
Player
Passing
Rushing
QB Rush%
Rush % Total
QB yds
Total Team
Offense Rush %
2001
3-8 (1-6)
Brad
Lewis
135/237
1,339yds
7TDs, 9INTs
54att,
41yds,
1 TD
18.6%
2.97%
52.4%
1,992 Rushing
1,811 Passing
2002
9-4 (6-1)
Rasheed
Marshall
139/259
1,616yds
9TDs, 5 INTs
173att,
666yds,
13 TDs
40.0%
29.2%
67.8%
3,687 Rushing
1,753 Passing
2003
8-5 (6-1)
Rasheed
Marshall
109/215
1,729yds
15 TDs, 8 INTs
101att,
303yds,
4 TDs
32.0%
14.9%
57.6%
2,762 Rushing
2,034 Passing
2004
8-4 (4-2)
Rasheed
Marshall
144/242
1,886yds
19 TDs, 9 INTs
169att,
861yds,
4 TDs
41.0%
31.3%
60.4%
3,034 Rushing
1,993 Passing
2005
11-1 (7-0)
Pat White
65/114
828yds
8 TDs, 5 INTs
131att,
952yds,
7 TDs
53.4%
53.5%
73.2%
3,269 Rushing
1,398 Passing
2006
11-2 (5-2)
Pat White
118/179
1,655yds
13 TDs, 7 INTs
165att,
1,219yds,
18 TDs
47.9%
42.4%
65.7%
3,939 Rushing
2,059 Passing
2007
11-2 (5-2)
Pat White
144/216
1,724yds
14 TDs, 4 INTs
197att,
1,335yds
14 TDs
47.7%
43.6%
65.1%
3,864 Rushing
2,067 Passing

Michigan QB Charts

Year
Record
Player
Passing
Rushing
QB Rush %
Rush % Total
QB yds
Total Team
Offense Rush %
2008
3-9 (2-6)
Steven
Sheridan
165/338
1,718yds
11 TDs, 12 INTs
118att,
293yds,
3 TDs
25.9%
14.6%
50.8%
1,771 Rushing
1,718 Passing
2009
5-7 (1-7)
Denard
Forcier
179/312
2,233yds
15 TDs, 14 INTs
187att,
591yds,
8 TDs
37.5%
21.0%
48.4%
2,234 Rushing
2,380 Passing
 
In Rodriguez's career at West Virginia there was a striking correlation between the amount of the quarterback's total yards came from rushing and the success of the team. When you sort by this statistic, from 2001-2007, this is what you get:
 
Year
Rush %
Total QB Yards
Record
2005 53.5% 11-1 (7-0)
2007 43.6% 11-2 (5-2)
2006 42.4% 11-2 (5-2)
2004 31.3% 8-4 (4-2)
2002 29.2% 9-4 (6-1)
2003 14.9% 8-5 (6-1)
2001 2.97% 3-8 (1-6)

Throwing the two seasons at Michigan into the mix wouldn't change much, either. The 2008 and 2009 campaigns would be two of the bottom four performances in terms of how many of the quarterback's yards came on the ground and two of the bottom three in terms of overall record.

Even beyond just the quarterback's performance in the offense, however, Rich Rod's offenses seem to thrive when the majority of the overall offense is primarily generated on the ground. From his last nine seasons:

Year
Total Team Offense Rush %
Record
2005
73.2%
11-1
(7-0)
2002
67.8%
9-4
(6-1)
2006
65.7%
11-2
(5-2)
2007
65.1%
11-2
(5-2)
2004
60.4%
8-4
(4-2)
2003
57.6%
8-5
(6-1)
2001
52.4%
3-8
(1-6)
2008
50.8%
3-9
(2-6)
2009
48.4%
5-7
(1-7)

And these aren't even necessarily based on the quality of the team. If this was ranked on team total rushing yards it would make sense that more yards would generally be correlated to a better record, but these are rushing percentages compared to the rest of the offense on a year-by-year basis, comparing one aspect of the team to another aspect of the same team.

Very interesting findings, if you ask me. Findings that point to Denard Robinson being the ticket to the promised land? Maybe. Clearly Rich Rodriguez's offense flourishes with a quarterback that can run, and run often. His best team had 73% of their total offense come on the ground. Its quarterback ran for 125 more yards than he passed for. Sounds like Denard to me.

Also notice that after White's first season--in which passing was sporadic and the team was heavily reliant on his rushing offense--his passing stats took off. As Denard's passing game develops, his rushing options will open up more, and as Steve Slaton got to experience, this opens the door for an unbelievable rushing attack from your running backs as well.

Agree? Disagree? Let me know what you see in these stats, or if they even matter to you at all. I look forward to hearing what you get out of these numbers. [Ed: /Ives.]

Thanks for reading,

CollegeFootball13

Comments

BiSB

August 16th, 2010 at 8:04 AM ^

Thanks.  I wonder about the effect of the intervening variable (i.e. both the good years and the running years were Pat White years, who was both a runner and good at football). Still, though, a stronger correlation than I would have expected.

PurpleStuff

August 17th, 2010 at 1:17 PM ^

I would like to see a similar look at what happened when Rodriguez was offensive coordinator at Clemson and Tulane (I remember Shaun King putting up a lot more yards through the air).  As is, this just looks to be a reflection of having a good team when Pat White was the quarterback (who just happened to be a great running quarterback) rather than a reflection of Rodriguez's preference or ability to succeed with more of a throwing QB.  Ryan Mallett would, of course, disagree with me.

Don

August 16th, 2010 at 8:03 AM ^

For example, in 2005, Pat White rushed for 952 yds, and the total team rushing is 3269 yds. If I'm interpreting those numbers correctly, that means that White's rushing yardage was just over 29% of the total WVU rushing yardage. What does the 53.5% refer to?

dakotapalm

August 16th, 2010 at 8:59 AM ^

I agree that your conclusions are true historically, but if the mental and leadership aspects are similar (between Denard and Tate), I don't see a reason why a healthy and less fumble-prone Forcier would not breed just as much success for a Rodriguez offense. Imagine the first few games  of 2009 with a competent defense? I think Rich would take that every time.

Personally, I like Denard as an individual and as a story, but if Tate learned to not wave the ball around, played with a healthy shoulder and David Molk were healthy- that seems like a recipe for a very potent offense.  I don't think that sort of QB was ever a choice for Rodriguez when he was at WVU.

I like the stats and the work that you have compiled. Good work.

jmblue

August 17th, 2010 at 4:01 PM ^

But that's also what made his play the last three weeks of the season encouraging.  He had a great game against a good Wisconsin defense and played well IMO for the first three quarters against OSU, despite having no running game to help him out.

Blue Blue Blue

August 17th, 2010 at 5:54 PM ^

seriously?  I thought his cough up for a TD to open the game was among the worst plays I have ever seen a Michigan QB make.  so much for the value of his "experience"!

 

these charts paint a picture of RichRod as a more ground based offense than I realized........I had images of all sorts of ninja slots taking quick slants and gouging secondaries, in additon of course, to the zone read.....and it is why I suspect we will see all three QBs early on.

Bronco Joe

August 17th, 2010 at 1:41 AM ^

Personally, I like Denard as an individual and as a story, but if Tate learned to not wave the ball around, played with a healthy shoulder and David Molk were healthy- that seems like a recipe for a very potent offense.

I definitely agree. Put another way, had Molk not gotten hurt, very possibly Forcier may not have gotten hurt as the offensive line would've given him much better protection overall. I think this would also have reduced his fumbles and INT's for the same reason.

I think there are exciting things ahead with Robinson and Gardner at QB, but I really worry about going to a relatively inexperienced QB again... The minutes Forcier played against the competition he was against week after week had to make a difference. Even with all the talent in the world, can DR or DG really make up for that? Or at the end of the season will we be asking, "What if Forcier had started instead? Where would the team have finished then?" Hopefully it will be a moot point.

michiganfanforlife

August 16th, 2010 at 9:03 AM ^

I would greatly welcome the idea of rushing all day, and only passing when you have to. This philosophy really tests the will of our opposition. The 3 yard gains up the middle in the 1st quarter turn into 8 yard gains in the 4th.  As long as ball security is less of an issue in regards to running the read option, this also will create less turnovers for our defense to deal with. Running gets your O-line in a nasty mind-set, and creates a rhythm for all the offense to follow. Please run the ball a ton this year...

cadmus2166

August 16th, 2010 at 9:41 AM ^

for this season, anyway.  I would not be surprised at all if DG is the starter for the 2011 season, barring a breakout season for either Denard or Forcier.

bluebyyou

August 16th, 2010 at 10:05 AM ^

I wouldn't be surprised either, but for another reason.  With Tate listed at 6'1" and 192 pounds and with Denard at 6' 0' and 193 pounds I live in fear of injury to one or both.  Gardner at 6'4' and 210 pounds will probably end up weighing 220 or 225 when it is all said and done and should be better equipped physically to take the pounding suffered by spread QB's.  Pat White was dinged up more than a few times at WVU.  Even Juice Williams and Pryor, both big QB's, were dinged up over the last few years.

Injuries to running QB's is the part of the spread that makes me nervous.  I like the way the NFL runs their wildcat offenses - keeps the QB alive.

MGlobules

August 16th, 2010 at 11:17 AM ^

that size alone correlates with injury-proneness, especially if a player works out, bulks up, and stretches. Seeing over and around those big guys coming at you is definitely an issue, and Tate was undersized, but I don't think it follows that because Gardner is taller he will be sturdier. DNA, bone density, all these things compound the issue. . . 

bluebyyou

August 16th, 2010 at 11:35 AM ^

That would make for a good diary, namely, size and weight of spread QB;s vs injury history.. As you noted, the equation becomes more complicated than what I outlined, but my guess is that smaller spread QB's are more prone to injury.  

Edward Khil

August 16th, 2010 at 12:22 PM ^

MGoCalibur posted a few enlightening posts last year on the subject of the risk of injuries in a spread offense.  His analysis showed that the risk for QBs in spread offenses was essentially the same as that for QBs in non-spread offenses, and also that the more athletic a QB, the less likely he would be injured when he left the pocket.

http://mgoblog.com/diaries/follow-qb-fragility

I'm not particularly concerned about QB injuries at the moment, because I think we've got three QBs who can run the offense at least moderately well.  I hope neither Tate or Denard get nicked, because I think that would necessitate burning DG's redshirt.

MGoStu

August 16th, 2010 at 5:54 PM ^

He said 2011, and since it sounds like he is looking really good in practice this year I don't think I'd be too shocked if he started next year. Especially if he gets some playing time this year.

Rasmus

August 16th, 2010 at 10:05 AM ^

Your charts simply show that the spread-option under Rodriguez has primarily been a run-based offense. When that offense is effective, the team will have a lead and continue to run the option. However, if the option component is not working, then the team is likely to be behind and start passing more (i.e., using the spread component)

This is common sense, but it is still interesting. It shows, I think, that while Rich's best teams were highly successful running, they were also able to pass when necessary. 2005 is an anomaly since White didn't start until the third game and, much like Denard last year (in a far, far more limited role), he wasn't ready to run the spread component effectively -- thus the high option (i.e., run) numbers for both team and quarterback. 2006 and 2007 are more indicative of what the spread-option looks like when it is at its best.

The "two-headed monster" at Michigan just adds another layer of complexity to the basic scheme -- Tate is somewhat more likely to use the spread component and pass, and Denard is somewhat more likely to use the option component and run, but it's all relative -- do you, opposing defender, really want to bet the house on which quarterback is going to do which?

OHbornUMfan

August 16th, 2010 at 12:32 PM ^

In my recollection, WVU stayed ground-based even when trailing.  Trailing didn't necessitate becoming a pass-happy team, since so many of their runs could a) go the distance, b) earn a first down, or c) attack the perimeter to allow for a quick step OB at the end of the run.

PeterKlima

August 17th, 2010 at 1:06 PM ^

I wanted to mention that teams throw less when winning.  That factor might really skew these statistics.  I appreciate the hardwork of the poster, but maybe 1st half stats would be a better way to look at it?

In reply to by PeterKlima

rlc

August 17th, 2010 at 1:48 PM ^

Exactly, teams that are not very good on a whole have to pass more. They have to do this because they are behind, or can't get first downs on the ground.

champswest

August 17th, 2010 at 1:38 PM ^

a QB with good speed and running ability (see Denard and Devin) because it opens up the passing game and makes it easier to throw.  IMO, Pryor's success throwing the ball at OSU is mainly due to the defense being so concerned with the threat of him running the ball the receivers are often left wide open.  Go back and watch some OSU tape (if you can stand it) and see how many times TP rolls out or gets flushed out of the pocket and finds a receiver running wide open.  Now, imagine DR or DG doing the same thing for us.

Not a Blue Fan

August 16th, 2010 at 11:18 AM ^

 

Year
Record
Player
Passing
Rushing
QB Rush %
Rush % Total
QB yds
Total Team
Offense Rush %
1997
7-4 (-)
Shaun
King
199/363
2,577 yds
24 TDs, 14 INTs
124 att,
511 yds,
5 TD
25.5%
16.55%
43.9%
2,038 Rushing
2,608 Passing
1998
12-0 (6-0)
Shaun
King
244/364
3,495 yds
38 TDs, 6 INTs
156 att,
633 yds,
11 TD
30.0%
15.33%
41.1%
2,523 Rushing
3,615 Passing
1999
6-6 (5-3)
Woody
Dantzler
112/201
1,506 yds
9 TDs, ? INTs
146 att,
723 yds,
4 TD
42.0%
32.44%
39.7%
1,580 Rushing
2,401 Passing
2000
9-3 (6-2)
Woody
Dantzler
122/212
1,691 yds
10 TDs, 6 INTs
172 att,
947 yds,
13 TD
44.8%
35.90%
53.1%
2,619 Rushing
2,311 Passing

 

I don't think I made any mistakes. I couldn't find Dantzler's INTs in 1999 or Tulane's conference record in 1997, but those are not terribly important. Dantzler split a lot of time at QB in 1999, but the other guy was a pocket-ish passer (and Dantzler had more yards, so...).

So add these data points:

 

Year
Rush %
Total QB Yards
Record
1997 16.5% 7-4 (?-?)
1998 15.3% 12-0 (6-0)
1999 32.4% 6-6 (5-3)
2000 35.9% 9-3 (6-2)

 

So, as far as I can tell, the only notable outlier is with Shaun King at Tulane.

wolfman81

August 16th, 2010 at 11:39 AM ^

that just shows that if RichRod has a more pass-friendly QB (as--iirc--Shaun King was) he can make the offense more pass friendly.  So if Tate wins the job, even though Tate can "scramble" and still run a bit, he might get a more pass-friendly section of the playbook to work with.  And therefore have another "outlier" year according to this idea.  [Disclaimer:  This is not a prediction about who the starting QB will be on opening day or any other gameday thereafter.]

Remembering some of the controversy when RichRod got here in 2008, some were frustrated that he didn't run a pro-style offense because he had a pro-style QB (Threet) and more pro-style talent (at least before they transferred or delcared for the draft).  I understand why RichRod did what he did from a scheme standpoint.  You always have to coach your strengths.  But he still accounted for the talent that he had.  He threw more than he did at WVU, and he ran his QB less (because, seriously, Threet doing his best Pat White impression was just comic relief, and not productive). 

In the end, I guess I've just taken a roundabout way of saying that the coaches will call the plays that best fit the personnel that they have on the field (regardless of who they are).  I don't think that they are thinking that they want the QB to rush for at least x% of the offensive yardage.  What they want is to have a very productive offense, putting their best athletes in positions to make plays, score touchdowns, and, in the end, win football games.  If they have to pass all over the field to do this, then fine, and if they can do this by running all over them then fine.  Take what the defense gives you and then exploit it should always be the mantra of an offensive coach.

Not a Blue Fan

August 16th, 2010 at 1:36 PM ^

A little post-lunch checking shows there's no real statistical evidence to support the "RR can't win without running" notion. The correlations between QB rushing yardage percentage and win percentage, QB rushing attempts percentage and win percentage, and team rushing yardage percentage (as calculated by the OP) and win percentage are negligible (R^2<0.4). Not surprisingly, the covariances are all positive (as one would expect), but fairly weak.

I'd really like to see more analysis, though, because this only appears to indicate that RR doesn't have to a disproportionate amount of success running the ball to succeed. I would really be keen to see some analysis of how his record compares to other coaches (does he need to rush disproportionately compared to the norm? etc).

maizenbluedevil

August 17th, 2010 at 6:23 PM ^

The past couple years I actually thought RR did a great job of tailoring his system to the personnel.

I remember watching WVU in RR's last year there, and, maybe my memory isn't real accurate but IIRC I remember them almost always running 4 WR sets, or a set with 2 RBs on either side of White, plus 3 WRs.

Since RR came to M, it seems like our default offensive formation is the one with 3 WRs, the TE in that sort of halfback position, plus an RB.  

We've also run a lot of I-form stuff.  (When I say "a lot", I mean in comparison to RR at WVU.)

So, of all the things people could criticize RR for, failure to adapt scheme to available personnel is IMO illegitimate.

Bodogblog

August 17th, 2010 at 4:00 PM ^

thank you for answering.  S. King is the outlier, but did he play before RR "invented" the zone read?  In HTTV, adding this ability to run the ball to the spread is what really blew up RR's offenses.  I don't know when that time period was

In any case, no worries.  Tate can run as well. 

Interesting data, but I think commitment (conditioning), and work/performance during this camp will determine the starter.  Tate's put himself behind if the wings/conditioning thing is true, and I think that matters more than anything else

HailGoBlue86

August 16th, 2010 at 11:33 AM ^

I'm all for this offense being like the West Virginia offenses from 2004-2007. If Denard gets the job and the stable of running backs pproduce we could have a lot of rushing yards just like those offenses. And might I add we have a lot more talented wide receivers to throw into the mix that West Virgina didn't have except for Chris Henry.

MGrad

August 16th, 2010 at 11:48 AM ^

Unless one views him as a product of the system, it might be difficult to prove the hypothesis when one considers that the QB driving the stellar numbers was Pat White.

If Denard can execute the zone read at the Big 10 level, and improve his defensive recognition and passing accuracy, then he will be a significant factor.  Those are all big areas to develop, however.  I think that Tate's main areas of development are decision making in both passing and the zone read, and getting healthy.  It will be interesting to see how the QB race unfolds.

UM Fan NY

August 17th, 2010 at 12:53 PM ^

all the wvu comparisons we've seen the last 3 years. completely different IMO. the spread was much more of a novelty at the time..... pat white was freaking amazing. ....and they played maybe 2 good teams a year in that crap conference. apples and oranges.

tacar

August 17th, 2010 at 3:32 PM ^

about Rodriguez's fit at Michigan. I'm 100% behind him, but if he's given enough time (another year after this one) and Michigan's still a .500 team at best, I think this will be the reason. Basically, defenses have adapted to this offense by now.

RoseBowlBound

August 17th, 2010 at 1:11 PM ^

This goes to the earlier comments on decision making.  RR will most likely go with the QB that he feels is least likely to turn the ball over especially early on in the season since these guys are still a couple years removed from HS.

Running with Denard and an improved offensive line over passing/scrambling/improvising with Tate might produce fewer turnovers and give us the best shot at winning football games since that is what this all about.

Wolv54

August 17th, 2010 at 1:46 PM ^

read option.  In it's infantile stages, the one read option was just killing defenses because of teh inherent mis match between 270 LB DE and quick and nimble QBs on one side and the numbers advantage with blockers when the ball goes to the RB.  Somewhere along that timeline there has been a shift in a defenses ability to counter the read option.  I guess to compare the success of the zone read option in 2008 and 2009 as compared to the early 2000's would offer some glimpse of how much is attributable to the players and how much is attributable to the system and it's effectiveness.

Interesting post.  We'll find out in 18 days.

massblue

August 17th, 2010 at 1:52 PM ^

To make this analysis meaningful, you need to compare the apparent correlation between the winning percentage and running QB with the same correlation for other programs.  That is, does the winning percentage of other teams increase as they run more.  I suspect the answer would be yes.  This means, the spread or its RR version is no different than other systems.  Teams that run more generally have a higher winning percentage.  

bronxblue

August 17th, 2010 at 1:53 PM ^

Nice data.  I do think that RR's offense is robust enough to take advantage of whatever skill set his QB possesses.  If the QB is a scrambler, then the offense has all of the delays, split handoffs, read options, etc. that benefit a runner.  If the QB is a better passer, then those little screen passes, bubbles, and short crossing routes are basically runs.  What I gather reading this chart is not so much that RR needs to run the ball to succeed, but that he needs to have an established QB who can do whatever he is "good at" well.  I still believe that if Threet had stuck around last year he would have succeeded because RR would have asked him to run at times but would have been fine throwing 20+ times provided they were shorter passes and the option was still a threat.  Hopefully Tate, Denard, or (shutter) DG can provide that type of consistency. 

Njia

August 17th, 2010 at 2:17 PM ^

I see a QB's runs or passes catching a defense completely flat-footed. They know he can run, and they know he can pass, but just when the DC thinks he's got the WVU offensive game plan dialed in, and calls the next defensive play accordingly - BOOM! - he finds his team playing Where's Waldo?

But, the biggest thing I see in that offense run by Pat White, is that it was a machine. Sure, most of the defenses it faced were pretty horrid. But, it also faced some pretty good teams in bowl games, and just shredded them.

slappy09

August 17th, 2010 at 2:32 PM ^

Are there any stats around o-line experience/performance tied to this data?  To me at a high level glance this seems like there is way too high an importance for that one strong player - in this case the QB.

If that really is the case and RR plans on running the ball 60% of the time with the non-sexy slot passes - no thanks - not exciting football.  I always remember watching the WVu games back when he was there and thought - great, but not exciting - seems like watching HS ball.