TCU vs North Carolina: previewing a game with no games played yet

Previewing a season opener is an honesty test: there is no 2026 film, no efficiency numbers, no form to lean on. So our engine does what the sharpest preseason analysts do — it builds a projection from what actually predicts the next season: last year's opponent-adjusted rating, the transfer portal, the projected starting quarterback, and recruiting talent, all with deliberately wide error bars. With those inputs, this opener between TCU and a heavily rebuilt North Carolina — a rematch of last year's Week 1, which TCU won 48-14 — comes out as a clear-but-not-safe TCU edge, and, encouragingly, our number lands right on the opening line.

What the preseason model keys on here:

• Quarterback certainty is the cleanest edge here: TCU returns Josh Hoover (a proven starter), while North Carolina's job is unsettled — a portal committee with no clear returning starter. Preseason, a known quarterback is worth more than a higher-ceiling guess.
• North Carolina rebuilt through the portal: 20 in, 31 out, and on star value it lost more than it added (net -36). This is roster turnover, not continuity — project it with wider error bars.
• Encouraging sign for the read: our model and the opening market agree on the favorite (TCU by about 6), and our projected margin lands close to the line — a preseason number we and the market both buy.
• Our model projects TCU for about 6 wins over its 11-game schedule.
• TCU's schedule isn't kind: its toughest test projects to be Texas Tech, one of several ranked opponents on the slate.

Two teams, two philosophies

These rosters were built in opposite directions. TCU is continuity: a 9-4 team last season bringing back its quarterback and running it back. North Carolina is the most dramatic teardown in the sport — it moved 31 players out and brought 20 in, a net loss of 36 stars of talent, and is effectively fielding a new team. Roster turnover that severe is exciting and almost impossible to project precisely, which is exactly why the model widens its uncertainty on the Tar Heels — and why last year's 48-14 result tells you very little about this one.

The quarterback edge

Preseason, the single most valuable thing a team can have is a known quarterback — and that's TCU's biggest advantage. Josh Hoover returns as a proven starter who posted a strong +0.39 EPA per play. North Carolina, by contrast, walks into Week 1 with an unsettled room: a six-way competition among returners and transfers with no established FBS starter. A higher-ceiling unknown can absolutely pop, but a known commodity is worth more on the day you have to project the game — and it's why the model leans TCU even though the two rosters recruit almost identically (No. 32 and No. 31 in talent).

What the model sees — and how sure it is

The simulator makes TCU about a 66% favorite, projecting roughly 28-22 — a margin of about a touchdown. That is essentially the opening market (TCU -6.5), and our total of 50 sits right on the market's 49.5. When a from-scratch preseason model and the market independently land on the same number, that's a number worth trusting. The honest caveat is the uncertainty: with zero games played, every rating here carries a ±6-point error band, far wider than a midseason read. This projection will sharpen every week once real snaps exist.

What the game turns on

When we perturb the simulation one factor at a time, the turnover battle moves the result more than anything else — about 26 points of win probability swing. That matters here specifically: a brand-new quarterback in a brand-new system is exactly the profile that gives the ball away early in the season. If North Carolina protects it, this is a coin flip; if the Tar Heels are loose with it, TCU pulls away. In the simulations where each team wins, the winner is the one that wins the turnover margin.