Does Mick McCall's offense need a dual-threat quarterback?

Quarterback is the most important position on a football field. A QB doesn't just have the ability to make below average teams good, or good teams below average, he is also the player around whom an entire offense is structured. Other skill position players and offensive linemen must fine tune their skills to mesh with those of the quarterback, and a coordinator or head coach must mold his scheme to accommodate the strengths and weaknesses of his signal called.

And here arises the biggest question surrounding Northwestern's three-way quarterback battle in 2015. Pat Fitzgerald basically laid out the dilemma at Big Ten Media Days:

"Two of the three guys [Matt Alviti and Clayton Thorson] are very similar. Their qualities are very similar. And then Zack's a little bit more traditional. He can run around a little bit, but I don't think he runs as well as Trevor when Trevor was healthy. So we've got some things to work out that way."

Since the competition began back in the winter, this difference in style among the three quarterbacks has governed the way the competition has been viewed. More specifically, many believed Thorson (or Alviti) to be the favorite, in part because of his dual-threat ability. That makes sense due to the common idea that offensive coordinator Mick McCall's best offenses at Northwestern have been with a mobile quarterback at the helm — or at least sharing the helm.

But is that really the case? Does McCall's offense need a running threat from the quarterback position to reach its full potential?

McCall came to Northwestern from Bowling Green, where he had been the offensive coordinator in 2007, and the quarterbacks coach from 2003-2006. Here's a look back at McCall's offenses since taking that Bowling Green job in 2003:

Note: PPP is Points Per Play, a measure of explosiveness. YPP is Yards Per Play, a measure of efficiency. Adj means adjusted for strength of opponent. Rank is out of all FBS teams. QB Rush Yds account for a team's starter and backups, and count sacks as negative yards, as is standard in college football. All stats courtesy of adjustedstats.com, except for QB rushing yards, which were compiled from Bowling Green's official stat archives and sportsreference.com.

Mick McCall offenses, 2003-2014

Year	Team	Adj PPP	Adj PPP rank	Adj YPP	Adj YPP rank	Raw PPP	Raw PPP rank	Raw YPP	Raw YPP rank		QB Rush Yds
2003	BGSU	0.348	70	5.67	39	0.361	67	5.89	24		919
2004	BGSU	0.469	24	6.12	20	0.592	5	6.83	4		342
2005	BGSU	0.367	61	5.23	61	0.449	22	5.68	36		232
2006	BGSU	0.198	111	4.6	96	0.275	96	4.87	88		856
2007	BGSU	0.341	80	4.95	83	0.417	44	5.57	48		168
2008	NU	0.363	67	5.18	70	0.321	86	4.89	89		551
2009	NU	0.293	97	5.15	83	0.303	98	4.98	93		462
2010	NU	0.336	83	5.29	69	0.354	74	5.38	69		580
2011	NU	0.41	45	5.86	36	0.371	62	5.59	50		705
2012	NU	0.457	39	5.5	59	0.417	50	5.27	86		942
2013	NU	0.375	68	5.65	63	0.339	89	5.37	75		522
2014	NU	0.309	92	4.84	104	0.288	110	4.52	122		-94

If numbers scare you, don't worry, graphs are coming. But first, let's tell the extremely abbreviated story of those 12 years, because context is always necessary when dealing with numbers.

In 2003, McCall had Josh Harris, a dual-threat QB who earned All-America honors.

In 2004 and 2005, he had Omar Jacobs, who wasn't necessarily immobile, but was a pro-style QB. Jacobs was an All-American in 2004, but missed a few games in 2005, which explains the offense's regression.

Jacobs left for the NFL after 2005, and was replaced by the athletic Anthony Turner in 2006.

The 2006 offense really struggled though, and the following year, 2007, Tyler Sheehan, a pro-style QB, beat out Turner for the starting job. Turner moved to running back (and the following year, changed positions again to wide receiver); the offense improved.

In 2008, McCall's first year at Northwestern, he had C.J. Bacher, a pro-style QB whom he coached to a reasonably good season. After Bacher went down with injury though, McCall changed the offense to account for the running ability of backup Mike Kafka, and NU didn't miss a beat.

In 2009, the offense remained pass-heavy under Kafka, who was still mobile but didn't have as much success on the ground.

In 2010, Dan Persa was a true dual-threat playmaker. In 2011 though, Persa — slowed by his Achilles injury from the previous year— became the throwing quarterback in McCall's first true two-QB system, as Kain Colter entered the picture as the runner. The 2011 season featured McCall's best offense at Northwestern.

In 2012, Trevor Siemian — significantly less mobile than Persa — took Persa's spot as the passer in the platoon. The offense was explosive, but not efficient. In 2013, the same system returned, but injuries set the offense back. In 2014... well... you know that story by now. It was a complete disaster.

So back to the question at hand... does Mick McCall's offense need a running threat at the quarterback position? First of all, it's important to establish that McCall has coached offenses with statistically mobile quarterbacks in some years, and in other years he has not:

More importantly though, there seems to be little correlation between a McCall offense's success and its quarterback's success on the ground. First, a look at raw and adjusted efficiency:

And now a look at raw and adjusted explosiveness:

Now, since McCall made the move from the MAC to the Big Ten, there does seem to be some correlation between QB rushing yards and the other variable, especially when it comes to explosiveness. One potential explanation is that because McCall's early NU teams were burdened by a talent disadvantage — something he did not have to deal with at Bowling Green relative to his conference opposition — perhaps he did need a mobile QB to have success in spite of weak offensive lines and an absence of big play ability at the skill positions.

But we must be careful about assuming causation. It's just as possible that the QB rushing yards were down because the offense as a whole was bad. Or, even more possible, that both are explained by a third variable.

For example, in 2009, Northwestern's quarterback was Mike Kafka. Kafka was a very capable playmaker with his legs, exemplified by his record-breaking 217 rushing yards in a single game against Minnesota in 2008. But Kafka ran for just 295 yards in 2009, and the offense struggled. But the offense's struggles aren't solely explained by that rushing yards number. Both variables were likely explained by that third variable: talent. Kafka, despite being a mobile QB, didn't have the players around him to put up the rushing totals that he might've been able to achieve if he had been in charge of NU's offense in, say, 2011.

With a sample size of just 12 — 12 years, yes, but that is small because we are considering each season as a whole, not in terms of the individual plays or games that comprise it — much of the year-to-year fluctuation in efficiency and explosiveness can be explained by talent. It is based on the offensive line, wide receivers and running backs, and most importantly, the natural ability of the quarterback.

In all most cases, for example, an offense is going to be better under an NFL-caliber QB than it would be under a replacement level college QB, no matter the styles of the two players. (Save for Northwestern's 2014 offense, which was horrible even though Siemian looks likely to make the Denver Broncos as a seventh-round draft pick.)

But isn't that kind of the point? Doesn't that answer the question stated in the title of this article?

McCall's offense doesn't need a mobile quarterback. It needs a good quarterback. McCall will tailor his offense to that quarterback's abilities, just like he did within the two-QB systems of 2011-2013. And with Northwestern's improved recruiting, McCall is no longer operating at a major talent deficit, so even if that argument had some merit back in McCall's early NU years, it no longer holds.

At Bowling Green, the offense wasn't better in 2004 compared to 2006 because it had a pocket-passer instead of a runner; it was better because it had a future NFL Draft pick instead of a soon-to-be running back. At Northwestern, the 2011 offense wasn't better than the 2009 offense or the 2014 offense because it had two capable running QBs; it was better because it had two healthy, talented QBs, and, crucially, more weapons surrounding them. The 2012 offense wasn't necessarily the most explosive because it had that mobility at QB; it was so because it had a healthy Venric Mark.

The 2015 offense isn't going to be made or broken by whether or not it has a dual-threat quarterback. Its success will depend on whether or not it has a competent quarterback, and whether or not that quarterback has a good enough supporting cast.