It’s not exactly a secret that quarterback Nathan Elliott was abysmal for UNC’s first two games. Here’s what our positional reviews said, because they do it better than I could here:

Cal:

I don’t think it’s a huge stretch to say that Elliott singlehandedly lost this game for the Heels. His offensive line didn’t help him much (more on this later), but even taking that into account, Elliott did nearly everything he was supposed to not do: turned the ball over (four interceptions!!), made bad reads on option plays for much of the first half, stared down receivers, threw up prayers instead of scanning the defense, got several balls batted down at the line, and generally looked like somebody seeing his first college action...

ECU:

Elliott continued to throw into blanket coverage and nearly every ball seemed like a floater off his back foot...As the game started to slip away, Elliott went back to floating the ball and forcing it into tight [windows]. This was exemplified by a pass into triple coverage on a third and 20 less than eight minutes in the third quarter and a jump ball in the middle of the field with around 11 minutes in the fourth.

Against Pittsburgh, the stat sheet says that he turned it around: 22/31 for 322 yards and two touchdowns looks like a great line. And that’s what a lot of the press around Elliott has been this week: that he’s put his struggles behind him, he’s worked hard to improve, and is ready to be who he was against Pittsburgh for the rest of the season. Anything beyond a cursory glance, though, tells you how flimsy this narrative is: Twelve out of those 22 completions were behind the line of scrimmage, and 217 of his 322 yards came after the catch. And the tape is even less kind to Elliott. With the help of this condensed game replay from the ACC Youtube page, let’s take a look:

The (Kinda) Good:

Elliott made 8 throws on the day that I can reasonably call “good” and “requiring actual quarterbacking talent.” Of those, three were intermediate-deep downfield, three were behind the line of scrimmage, and two were short passes beyond the line.

Here’s an example of each:

ACC Digital Network

ACC Digital Network

ACC Digital Network

That first throw was easily the best play of Elliott’s career: in short order, he steps up to avoid pressure from the edge, gets out of the pocket to buy time, and fires a dime on the run to Newsome, who does a good job of working back and to the outside to give his quarterback a target. If Elliott could make throws like this regularly, Carolina would be undefeated. But I’ve never seen the ball come out of his hand with that much zip, and his accuracy on the run is historically scattershot. There’s a reason you only see a play like this out of structure, because he’s failed often enough that he’s no longer trusted to make this play within the offense.

The other two throws are more within Elliott’s previously displayed skillset, but they’re still positive aberrations, as we’ll see below. The short curl to Anthony Ratliff-Williams is well identified, then accurately and quickly delivered. The tunnel screen to Thomas Jackson is an absolute beauty, as it allows Jackson to not break stride and puts him exactly where he needs to be to maximize his blocking. These two plays exemplify the offense’s ceiling as long as Elliott is at the helm. With little downfield threat, Elliott’s placement needs to be pinpoint on these plays to squeeze every possible bit of yardage out of them against defenses that are going to be playing up to the line. On these plays and a couple of others like them, it is.

Unfortunately...

The Rest

After the first two games, Elliott told reporters that he needed to work on his footwork in order to improve his accuracy. You can judge that for yourself in the clips I’ve provided. Regardless of that, his accuracy was better against the Panthers than it was in the first two games. But it still wasn’t good enough, particularly with his limitations in arm strength and field processing. His first completion of the game illustrates this:

ACC DN

The flea flicker doesn’t work, and the downfield receivers don’t have separation. Elliott looks to Ratliff-Williams in the flat, and the nearest defender is about 10 yards away and going the other direction. Elliott makes the right decision to go to Ratliff-Williams, but for some reason pump-fakes, fooling nobody because there isn’t anybody to fool. He then throws while hopping backwards, taking zip off his ball and allowing the defender to realize where the ball is going and recover to where he is about 5 yards from ARW and closing in when the ball arrives. The receiver’s elite acceleration gets him the first down, but Elliott makes it harder than it should be.

In the above case, Elliott’s miscue doesn’t change the result. A couple of plays later, it noticeably did:

Here, Elliott tosses up a moonball into Antonio Williams’ numbers on the swing pass, denying him from creating any forward momentum with the catch and ultimately getting him tackled short of the sticks on third down. This throw needs to be deeper and on the outside shoulder to keep Williams running. Ideally, actually, it leads Williams into space, but with Williams looking back, that’s not an option here.

This kind of scattershot accuracy that restricts plays is all too common in Elliott’s Pittsburgh tape; many of his apparently positive plays were held back from being even more because he slowed down a receiver or started him running the wrong way. On several others, he throws a ball that has no business getting caught, but his receiver makes an outstanding adjustment to not only catch the ball, but then make a play after. Here’s an example:

ACC DN

Dyami Brown has to make a near 270-degree turn here along with the one-handed catch behind him, but amazingly, is able to do so without really breaking stride at all, and destroys the linebacker’s pursuit angle en route to a first down. It’s a very good play by the freshman. The result is the same as if Elliott had properly led Brown with this throw (which is all the worse because it’s on Elliott’s strong side; a throw across the body to the inside of the receiver instead of the outside is understandable but there’s no excuse when it’s away from the body), but again, Elliott makes his receivers’ jobs harder. We said at the beginning of the season that Elliott’s job was just going to have to be putting his skill position guys in position to make plays, but he’s making this proposition entail much more than it has to.

But, you might interject, it looks like these plays are working just fine! Who cares if it’s difficult if it’s succeeding? Well, you have to keep in mind the opposition here. Pittsburgh is, put charitably, not good. Furthermore, Fedora has their number. Making these plays against the Pittsburgh defense is much different from making them against Virginia Tech, Duke, and particularly Miami, all of whom have All-ACC candidates on their defenses. It won’t be nearly as easy to make their defensive personnel look silly like UNC’s receivers consistently did to Pittsburgh. And, with tape of the lengths Larry Fedora has to go to give Elliott a chance at success , those plays get even less chance to happen. Elliott’s not going to be facing much more two-high safety looks. And when this is what the quarterback usually offers going deep:

ACC DN

And the only real hope for a chunk play in the passing game is incredible plays like this:

ACC DN

Well, you can see why serviceable won’t cut it when it comes to the lateral passing game that Elliott requires to succeed. Heck, I’m not sure that perfect will cut it against defenses that have actually prepared for the game, but anything less definitely won’t. Relying on Elliott any further is nothing less than hoping a miracle will lift the program to victory. I mean no disrespect to Nathan Elliott himself, who by all accounts is a good guy, very hard worker, and a great leader. However, he’s shown through three games, and even the three before that, that he just isn’t the answer at QB. The other options on the roster might not be either, but as of now, they certainly have more chance at being so than the guy currently starting. Larry Fedora, if you have any interest at making a bowl game this season, don’t start Elliott. I don’t really care whether it’s Chazz Surratt or Cade Fortin who gets the call instead, but the quarterback position desperately needs a change for UNC’s upcoming game against one of the best defenses in the nation.