Maryland +17.5 Leads a Flat Week 12 Board

Fifty-four games priced, zero games graded above the model's bottom tier. Here's what the numbers say anyway, record and all.

By Oddsnyper CFB Desk ·

A Board With No Star Player

Fifty-four games got priced for Week 12. Zero of them cleared the model's bottom confidence tier. That's the headline whether we like it or not — every play below is a C, which in our grading scheme is the floor, not a foundation. We're not going to dress these up as conviction bets because the market happened to leave a gap on the number.

The widest gap on the board belongs to Maryland at USC. The model has USC winning by 14.7. The market has USC by 17.5. That's a 2.8-point split, the largest of the week, and it lands on Maryland. Colorado State at Fresno State is right behind it — model says Fresno State by 16.9, the line says 19.5, a 2.6-point gap that puts Colorado State on our side of the ledger. Pittsburgh at Louisville checks in with the same 2.6-point split: model has Louisville by 10.9 against a market number of 13.5, so Pittsburgh is where the model lands.

The Rest of the Gaps

Northern Illinois at North Dakota State is the oddball of the group — a game where the model actually projects the favorite to cover harder than the market thinks. North Dakota State is projected to win by 27.9, the market has them at 25.5, and the 2.4-point gap sides with North Dakota State, not the dog. That's worth noting only because it cuts against the pattern of every other play here, all of which lean toward the side getting more points than the model wants to give away.

Oklahoma State at Arizona State is close to a coin flip by the model's math — Oklahoma State by just 0.1 — against a market line of Oklahoma State by 2.5. That 2.4-point gap sides with Arizona State. Iowa at Illinois has the model backing Illinois off an 8.5 market number against a 6.3 projection, a 2.2-point difference. NC State at Florida State and Arizona at Kansas State round out the board at 2.1 and 2.0 points respectively — NC State getting the number against Florida State's projected 8.4 versus the market's 10.5, and Kansas State getting the model's nod against a market that has Arizona covering 7.5 when the model sees 5.5.

What the Record Actually Says

The season-to-date spread record sits at 260-289-7. Totals are 266-314. Across 631 graded games, that's a losing record on spreads and a losing record on totals. We're printing that plainly because it's the entire context for everything above it. A handful of 2-3 point divergences on C-grade plays doesn't mean much when the underlying track record hasn't cleared .500.

None of this is a demonstrated long-run edge, and none of it is advice. A 2.8-point gap on Maryland is a bigger number than a 2.0-point gap on Kansas State, but bigger doesn't mean better when the whole tier is sitting at the bottom of the model's confidence scale. Treat the board as a list of places where the model and the market disagree — not a list of winners. The sample is still small, the record still proves nothing, and Week 12's best divergence is still just a C.

See the model’s graded picks →