Same Opponent, Different Team: The Numbers Behind UNC's Revenge Win Over TCU

An OpponentIQ Deep Dive — August 31, 2026

Last September, TCU walked into Chapel Hill and spoiled Bill Belichick's college debut, 48–14, in a game that was never close. Fifty-two weeks later, the same two programs met at Aviva Stadium in Dublin for the Aer Lingus College Football Classic — UNC's first international game ever — and the Tar Heels, an 8.5-point underdog, walked off with a 15–10 win. TCU had scored 28+ points in 30 straight non-conference regular-season games. That streak died in Ireland, where the Horned Frogs didn't score at all over the final 49 minutes.

The final score says "ugly win." The play-by-play data says something much more interesting: this was one of the most complete one-year defensive transformations we've charted. We ran both meetings through the OpponentIQ engine — every snap, tagged by down, distance, field zone, direction, and depth. Here's what actually changed.

The Headline Flip

Success rate is our schedule-keeping stat: a play "succeeds" if it gains 40% of yards-to-go on 1st down, 60% on 2nd, or converts on 3rd/4th. It strips out garbage yards and tells you who's actually winning snaps.

In 2025, TCU won 62% of its offensive snaps against this defense — a video-game number. In 2026, UNC flipped a 29-point success-rate deficit into a 7-point advantage. And TCU's points per possession collapsed from 4.80 to 1.00. Same opponent, same coordinator's system on the other sideline — a 79% drop in scoring efficiency per drive.

The Coolest Stat in the File: 15 Carries, Minus-4 Yards

TCU gashed UNC for 258 rushing yards at 7.4 a carry in 2025, including Kevorian Barnes' 75-yard touchdown. In 2026, the Frogs ran 39 times for 106 yards — 2.7 per carry.

But the direction matrix is where it gets fun. TCU ran between the tackles 15 times in Dublin and gained minus-4 yards. That's -0.27 yards per carry up the middle, with a 60% stop rate. A year after getting bulldozed, UNC's interior — led by Xavier Lewis (7 tackles, 4.5 TFL, 2 sacks) — turned the middle of the field into a wall. UNC stuffed 19.4% of all TCU runs for zero or negative yardage and posted an 18.3% havoc rate (TFLs + sacks + forced turnovers + breakups per snap). For reference: in 2025, UNC recorded zero sacks against this same program. In 2026: three, plus Melkart Abou Jaoude's third-quarter strip-sack that set up points.

Fourth Down Was a Graveyard

Here's a stat that decided the game and won't show up in a box-score tweet: TCU went 1-for-5 on fourth down. Four turnovers on downs — each one functionally a takeaway. UNC's fourth-down defense created negative plays on four of the five attempts, including a stop that went for minus yardage on 4th & 7-10 (average: -15.5 yards per play in that bucket). In a five-point game, an 80% fourth-down stop rate is the margin.

The red zone tells the same story. In 2025, TCU converted red-zone trips into touchdowns at will — 65.4% success rate inside the 35, three TDs. In 2026: 34.6% success, a 12.5% TD rate per trip, and one touchdown all night. Bend, don't break, formalized.

The Offense Fixed Its Worst Habitat

The 2025 loss had a hidden killer: field position. Backed up inside their own 35, UNC's offense succeeded on just 22.2% of snaps — which is how you end up punting six times and handing a great offense short fields. In Dublin, UNC's backed-up success rate was 64.3%, nearly triple. First-down success jumped from 38.1% to 57.1%, which meant the Heels lived in 2nd-and-manageable instead of 3rd-and-forever.

The passing profile change is just as deliberate. In 2025, UNC threw deep on 7 of 21 attempts (a desperate 33% deep rate) and completed 28.6% of them. In 2026, Billy Edwards Jr. went deep just twice in 28 attempts — and still averaged 14.5 yards per completion, because the catch-and-run game did the work. Jelani Thurman turned 8 targets into 94 yards; Jordan Shipp, the lone holdover headliner from the 2025 file, added 71 more. Explosive play rate actually rose (18.4% → 22.8%) while the deep-ball diet shrank. That's efficiency, not conservatism.

What the Model Still Flags

We chart weaknesses too, and Dublin wasn't spotless: UNC converted just 3 of 11 third downs (27.3%), succeeded on just 2 of 5 third-and-short snaps, and coughed up two fumbles. TCU's backed-up offense still succeeded 57.1% of the time, and the right-side run defense allowed 8.3 yards a carry on 11 attempts. Clemson on September 19 will find those rows in the matrix too.

But the through-line from the 2025 file to the 2026 file is unmistakable. Stop rate: 38.0% → 57.7%. Explosives allowed: cut in half. Rush yards per carry allowed: cut by two-thirds. Points per possession allowed: cut by 79%. Belichick said last year's roster "came here for somebody else." The 2026 data reads like a team that came to play for him.

Every number in this post comes from OpponentIQ's head-to-head engine — snap-level success rates, direction matrices, field-zone splits, and havoc metrics for every FBS matchup. Want your team's file? Visit OpponentIQ.

Sources for game details: GoHeels recap, Bleacher Report, ESPN