Previous post:

Next post:

How much will Terrelle Pryor help Darren McFadden?

by Chase Stuart on September 1, 2013

in History, Random Perspective On, Rushing, Statistics

Will the read option come to Oakland?

Will the read option come to Oakland?

The 2012 Raiders weren’t very good. Oakland finished last season 4-12, with the third-worst SRS rating (-10.8) in the NFL. And the outlook isn’t any better for 2013, either. Oakland has fifty million dollars of dead cap space; as a result, the players on the current roster cost the team only 67 million salary cap dollars, nearly $30M less than the next-lowest team. According to Pro Football Focus, Oakland’s five best defensive players last season were Lamarr Houston, Desmond Bryant, Philip Wheeler, Richard Seymour, and Rolando McClain. Only Houston returns in 2013, as Bryant is in Cleveland, Wheeler signed with the Dolphins, Seymour remains a free agent, and McClain has gracefully retired.

The team’s top two offensive players in 2012 were left tackle Jared Veldheer and quarterback Carson Palmer. Palmer is now in Arizona, while Veldheer is out indefinitely with a torn triceps. Brandon Myers (Giants) and Darrius Heyward-Bey (Colts) are also gone, and they combined for 29 starts last season. The big free agent signings were S Charles Woodson (Green Bay), LB Kevin Burnett (Miami), QB Matt Flynn (Seattle), CB Mike Jenkins (Dallas), CB Tracy Porter (Denver).

As a result, there’s little optimism in Oakland entering the season. The Raiders are one of the favorites to land the first pick in the 2014 draft, so the 2013 season will likely be used to see what building blocks actually exist in Oakland. After Terrelle Pryor outplayed Flynn in the preseason, many now think the Raiders going to start Pryor in week 1 because, well, why not? If that’s the case, we’ll have another example to test out a theory that’s widely-accepted by conventional analysts.

Most believe that running quarterbacks provide a boost to ther running backs; that effect appears to be magnified when you introduce elements like the read-option offense and the Pistol formation.  Last year, Alfred Morris went from unheralded sixth round pick to 1600-yard rusher; much of the credit goes to Morris, but the Shanaclan and Robert Griffin III made life easier for Morris, too.  Conceptually, it’s hard to argue with this philosophy, but isolated examples never prove much. Consider that in San Francisco last season, Frank Gore averaged 5.51 yards per carry in the first eight games of the season (with a healthy Alex Smith) and then just 4.31 YPC ten games he that Colin Kaepernick started.

Today, I want to take a step back and see if there are any general trends when a mobile quarterback replaces (or is replaced by) a non-mobile quarterback. The first question is how to define a mobile quarterback, and there’s no correct answer. One way that seems to make sense to me is to take a quarterback’s rushing yards and divide it by his number of pass attempts. The results seemed pretty good to me, and there are 53 quarterbacks who (1) had at least 200 pass attempts in a season since 1990, and (2) averaged at least 0.90 rushing yards per pass attempt, which was a good cut-off after scanning the results. If we limit it to just quarterbacks who started 5 games or missed at least 5 games, that drops the number to twenty-one names. I’m going to remove Jim Harbaugh (1994) and Rob Johnson (2000, 2001), who otherwise would qualify. Neither player was really a running quarterback, and in Johnson’s case, he was splitting time with a more mobile quarterback in Doug Flutie.

That leaves 18 examples. For each team, I recorded the rushing numbers produced by the top running back. Then, I noted the top quarterback (as measured by pass attempts) in each game, and whether he was a mobile or non-mobile quarterback. The table below shows a lot of data, so let me walk you through the first line. In 2012, Gore was the top running back for the 49ers. The non-mobile quarterback to play the most games was Smith (of course, Smith played in all 8 games, but for all teams, I only listed the top non-mobile QB), and in those eight games, Gore produced a stat line of 119-656-4. He averaged 5.51 YPC and averaged 82 rushing yards per game. In 2012, Gore also played with a mobile quarterback — Kaepernick. He played in 11 games (including playoffs), and produced 877 yards and 8 touchdowns on 202 carries. That means he averaged just 4.34 YPC in those 11 games, and averaged 79.7 rushing yards per game in those 11 games.

Team
Year
RB
Other QB
G
Rsh
Yds
TD
YPC
Yd/G
Mobile QB
G
Rsh
Yds
TD
YPC
Yd/G
SFO2012Frank GoreAlex Smith811965645.5182Colin Kaepernick1120287784.3479.7
PHI2012LeSean McCoyNick Foles34417203.9157.3Michael Vick915666824.2874.2
TEN2012Chris JohnsonMatt Hasselbeck611469136.06115.2Jake Locker1016255233.4155.2
DEN2011Willis McGaheeKyle Orton58538414.5276.8Tim Tebow1220095244.7679.3
PHI2010LeSean McCoyKevin Kolb47126813.7767Michael Vick1214885865.871.5
TEN2009Chris JohnsonKerry Collins712371345.8101.9Vince Young92351293105.5143.7
KAN2008Larry JohnsonDamon Huard35727424.8191.3Tyler Thigpen913660034.4166.7
JAX2006Fred TaylorByron Leftwich711547724.1568.1David Garrard811666935.7783.6
CHI2003Anthony ThomasChris Chandler716360553.7186.4Kordell Stewart68141915.1769.8
PHI2002Duce StaleyA.J. Feeley610733333.1155.5Donovan McNabb1219381734.2368.1
MIN2001Michael BennettSpergon Wynn57532624.3565.2Daunte Culpepper89735603.6744.5
PIT2000Jerome BettisKent Graham48131923.9479.8Kordell Stewart12274102263.7385.2
CHI2000James AllenShane Matthews817469714.0187.1Cade McNown811642313.6552.9
TEN1999Eddie GeorgeNeil O'Donnell612042223.5270.3Steve McNair143081331104.3295.1
PHI1999Duce StaleyDoug Pederson918469523.7877.2Donovan McNabb714157824.182.6
JAX1995James StewartSteve Beuerlein42911704.0329.3Mark Brunell1010840823.7840.8
SFO1991Keith HendersonSteve Bono54214413.4328.8Steve Young99541714.3946.3
DET1990Barry SandersBob Gagliano69451135.4485.2Rodney Peete10161793104.9379.3
--Average--5.71004342.14.3273.6--9.81637244.24.4673.2

In 2009, Chris Johnson rushed for over 2,000 yards, but he was even better once Vince Young replaced Kerry Collins at quarterback. That season, CJ2K averaged 102 rushing yards per game with Collins but 144 yards per game with Young. Since then, this has been a frequently-cited example of why it helps your running back to have a mobile quarterback.

But there are nearly as many examples going the other way, including Johnson performing better last year with a stationary Matt Hasselbeck versus a mobile Jake Locker. Seven of the eighteen running backs averaged more rushing yards per game playing alongside the stationary quarterback, while half of the sample fared better in yards per carry. The sample size is not very large, so we should always remember that sometimes, splits happen. Each situation is unique, of course, which is why I’ve presented the full list so you can add in your thoughts. Intuitively, it certainly makes sense that a rushing quarterback would help his running back, but the results hardly confirm that proposition. And this counter-intuitive result jives with Doug’s unexpected findings when he examined Michael Vick, Warrick Dunn, and T.J. Duckett a few years ago.

On the other hand, back in 2006, Football Outsiders found some evidence in favor of conventional wisdom using a different methodology.

To find out if the numbers support this theory, we put together a list of quarterbacks since 1978 who rushed for over 300 yards, then either left the team or lost at least eight games to injury. We added in one more quarterback who had 290 yards in half a season (Kordell Stewart in Chicago, 2003) and took out quarterbacks who were replaced by other running quarterbacks (for example, Steve Young replaced by Jeff Garcia).

The resulting list had 16 teams. 14 of those teams got fewer yards per carry from their running backs in the year where the running quarterback was either injured or gone. The average drop was .46 yards per carry.

Aaron Schatz noted that one of the outliers was Vick’s Falcons teams, and that the effect wasn’t as strong in the other direction (when teams switched from non-mobile to mobile quarterbacks).

Previous “Random Perspective On” Articles:
AFC East: Buffalo Bills, Miami Dolphins, New England Patriots, New York Jets
AFC North: Baltimore Ravens, Cincinnati Bengals, Cleveland Browns, Pittsburgh Steelers
AFC South: Houston Texans, Indianapolis Colts, Jacksonville Jaguars, Tennessee Titans
AFC West: Denver Broncos, Kansas City Chiefs, Oakland Raiders, San Diego Chargers
NFC East: Dallas Cowboys, New York Giants, Philadelphia Eagles, Washington Redskins
NFC North: Chicago Bears, Detroit Lions, Green Bay Packers, Minnesota Vikings
NFC South: Atlanta Falcons, Carolina Panthers, New Orleans Saints, Tampa Bay Buccaneers
NFC West: Arizona Cardinals, San Francisco 49ers, Seattle Seahawks, St. Louis Rams

{ 5 comments… read them below or add one }

Tom September 1, 2013 at 3:53 pm

While I agree that a mobile QB plays a role, I don’t think it is a definable variable in the process in determining benefit or cost to a RB. For example, maybe the replacement QB was a more dynamic athlete who didn’t play from the start because he was a more raw talent. Adding such a talent to the offense could help stretch fields, spread defenders out, or cause them to react a fraction of a second slower. We also must consider strategy changes. With Kaep in, for example, they ran some pistol offense. Things changed in scheme from one QB to the other. We also are assuming constant variables elsewhere — when there aren’t. Maybe a big injury to a lineman occurred (look what happened when Jason Peters wasn’t there for PHI last year). Maybe a playmaker got hurt (look at Garcon in vs out in terms of offensive success in WAS). Also, we’re not accounting for random variability in the process as well within those data points. The bottom-line is that we can’t analyze any of it conclusively or even remotely. It’s a case by case situation. One team could be better suited to zone blocking, while another might be more equipped for a power game, etc etc.

Reply

Tom September 1, 2013 at 3:55 pm

On the Pryor note, I’d like to say it helps DMC. The reasons are because no defense would be scared of Flynn stretching the field with his weak arm. They could take away a lot of the field for OAK. With Pryor, the defense has to account for more variables. When you have to account for more variables, you can’t as easily key on any one component. They could try to shut down the running game and dare Pryor to beat them — but that’s a guess and possibly a foolish one. Either way, OAK’s line is decimated and there is a severe lack of talent. While I think it helps DMC, I think we’re in for a terribly long season.

Reply

Wade Iuele September 2, 2013 at 10:44 am

I am curious to see what kind of offense the Raiders will run with Pryor as the QB. I’m confident that for as long as Pryor is in there, the Raiders will rely upon the running game to move the ball. One thing that will not help McFadden is Pryor’s poor decision making which will lead to turnovers and playing from behind. Combine that with perhaps the worst defense in the NFL and McFadden may not be a factor in the second half of football games.

Also, watching Pryor in the preseason, I saw no evidence that he could (or was allowed to) change the play at the line to get McFadden into better running situations, or get him out of bad ones. I’m not sure Pryor can read the defense before the snap.

I think Pryor is good for the Raiders in terms of big plays and avoiding being sacked every series. But I don’t see him boosting McFadden’s value beyond the OC calling a running play 60+% of the time because the QB is a throwing liability.

Full disclosure, I am a life long Raider fan, so I am emotionally unstable on these issues.

Reply

Jef September 3, 2013 at 2:52 pm

Being a huge Raider fan I totally agree. Altho I think Pryor will win us a couple games on pure athleticism (I was afraid we’d be 0-16 with Flynn), I’ll be interested to see how smart he is and if he can read defenses. He’s horrible about driving down field like a champ, then forcing an interception in the end zone. Those are situations where I hope McFadden is the DMC of old and they depend on the run.

Hopefully I’m wrong and Pryor will be the best QB we’ve taken in the draft since Stabler. He will have more time to pass now that we got rid of Barron on the left side. I’m not sure if that’s a good thing or bad thing.

Reply

Ike September 4, 2013 at 2:39 pm

You can use all the metrics you want. Pryor’s not even an average QB. Any team that has to start him is doomed to failure with or without him.

I don’t know what preseason games you were watching but Pryor clearly did NOT out play Flynn. Pryor didn’t even out play McGloin. Flynn and McGloin out played Pryor in every passing stat, except sacks and passing yards for Flynn and the yardage was only the result of more attempts. Almost all of Pryor’s rushing yards were the result of one or more of the following: not being able to check off all of his receivers; not being able to recognize a blitz and audible a new play; holding onto the ball; and in all those situations, having to bolt. That’s not a good QB.

As far as dividing anything into yards gained, so what? How many dual purpose QBs ever played for a team that won a conference championship or Super Bowl?

BTW, re your chart. You fail to differentiate between a mobile QB, able to escape a collapsing pocket, and a running QB, a QB who thinks run first and/or has running plays called for him. McNabb, Peete, McNown, were not “running” QBs. The style and impact on a RB of QBs such as Vick, Tebow can not be compared to QBs like Peete and the others I mentioned.

Prior to the 90s almost every NFL QB was mobile. In the 90s teams had QBs bulk up thinking it would prevent as many injuries as in previous years. It didn’t. What it did was make QBs less mobile. The introduction of the Spread Offense in colleges around 2000 created contrasts and misconceptions about the needed essentials and role of a QB in the minds of today’s younger fans.

Reply

Leave a Comment

You can use these HTML tags and attributes: <strong> <em> <pre> <code> <a href="" title="">

{ 2 trackbacks }