Looking at the Numbers Behind the UNC QB Rotation

Questions are mounting about Larry Fedora's use of his quarterbacks. But is the quarterback rotation actually having a negative impact on the Carolina offense? The stats and data are inconclusive at best.

Mitch Trubisky's back-breaking pick-six against Virginia Tech has brought the topic of the UNC quarterback rotation with Marquise Williams from the minds of fans to the lips of the local media, who have started to press coach Larry Fedora on the issue. Brian addressed the philosophical and practical issues behind this situation earlier this week. But just how much of an impact is the weekly quarterback switch actually having on the Carolina offense according to the statistics?

There is an old adage that if you have two quarterbacks, you don't really have one. Neither Williams nor Trubisky emerged from training camp with a clear edge but clearly Williams' on-field performance has earned him the unquestioned starting nod. Here are the season numbers so far for each quarterback:

So since there is no controversy over who the starter is, the question then comes to whether or not an actual "quarterback rotation" even exists. Here is a look at the number of series for each quarterback over the first five games:

Trubisky has only been on the field for 18 of UNC's 81 possessions this season (Kanler Coker played one series against Liberty). That means Williams has been the quarterback for 76.5% of UNC's offensive series overall, and 85.1% of the series in the current three-game losing streak. It is also worth pointing out that three of Trubisky's five series against ECU were the final three when the game was out of reach, as were his last three against Liberty, so fully one-third of the freshman's snaps have come in garbage time. There is clearly some room for debate on whether or not playing less than one possession in five even qualifies as a rotation.

Much has been made recently of Fedora inserting Trubisky in the third series of each game, but this has been the pattern all season long. Listed below are which series Trubisky has played in each game:

Since the third series of each game has been consistent over all five games, here are the results of the third series from each game:

Trubisky's numbers on the third series of each game are better than it would seem. He had the bad series against San Diego State (and did not play again), and a rough series against Virginia Tech (although Jack Tabb's dropped pass and John Ferranto's false start penalty put UNC behind the chains). But otherwise in the other three games Trubisky has two scoring drives and a 10-play drive against Clemson when the Tar Heels took over on their own four-yard line. Moreover, in the third series, Trubisky has racked up nine first downs and gone 5-8 on 3rd down conversions, while his only three-and-out was last Saturday against the Hokies.

One line of thinking out of the Virginia Tech game was that the insertion of Trubisky in the third series cost UNC some offensive momentum. Williams recovered from a disastrous first play fumble to lead the Tar Heels right down the field, only to have Tabb drop a sure touchdown, which forced UNC to settle for a field goal. After Trubisky's three-and-out, the UNC offense bogged down on its next three possessions, when Carolina ran 12 plays for only 17 yards. Fedora, looking for a spark to a lethargic offense, inserted Trubisky on the next series. The freshman hit Tabb for 11 yards and a first down, then ran a quarterback draw for 17 more yards and another first down before throwing the fateful pick-six. In two plays, Trubisky picked up two first downs and nearly twice as many yards as the previous three possessions combined before staring down a receiver and throwing a weak ball that was easily picked off for a touchdown.

But back to the suggested loss of momentum after the third series. Was this a one-time thing, or was it a pattern over the season? Here are the results of the second and fourth series from each game this season:

This chart is interesting in that it shows in the series after Trubisky plays, Carolina has punted all five times. But does this lend credence to the theory that playing Trubisky in the 3rd series each game disrupts the Tar Heels' offensive flow? It's hard to say. UNC scored touchdowns on its first and third possessions against Liberty, but did not even score an offensive touchdown until the second half against SDSU. And while UNC punted in the fourth possession against ECU, it did score a touchdown on its fifth series. Carolina opened the Clemson game with five straight punts, so the third series was not out of the ordinary. That leaves only the Virginia Tech game and the overarching question: did the quarterback change cost UNC offensive momentum, or did Bud Foster and the Hokies simply make adjustments that suppressed the Heels until the second half?

For his part, Larry Fedora has defended the manner in which he plays his quarterbacks, and it seems unlikely to change any time soon. And while the data may not indicate it is helping to play Trubisky, it doesn't seem that it is hurting either. The quantitative data seem to be ultimately inconclusive. Of course, my issue is less with trying to steal a series here and there for Trubisky as it is being bound to the third series, regardless of time, score, or momentum, or game situation. But unless Larry Fedora finds a compelling reason to change the pattern, Tar Heels fans can likely expect to see #10 trot out on Carolina's third possession each week and any other time Fedora wants to mix things up.