BYU at Colorado State. Lean: Colorado State +17.5. Lean: Under 52.5.

Spread.Colorado State +17.5 is a modest lean. EL makes BYU a 15.4-point favorite, leaving only 2.1 points of disagreement with the latest independently located quote, insufficient evidence from a model without demonstrated ATS value. The stronger argument is personnel: Jim Mora’s rebuilt offense offers more scoring potential than CSU’s 2025 baseline suggests, giving the Rams a plausible route to stay within three scores. However, the market has already recognized improvement: the dossier shows DraftKings moving from +21 to +17.5. I would not characterize that move as confirmed sharp action. The counterargument is substantial: Wyoming’s loss included a +4 CSU turnover margin, and Southern Utah still gained 442 yards against the Rams. BYU’s efficient quarterback play could punish that defense. Damuni’s absence helps CSU’s passing matchup, while Moa’s reported knee surgery reduces BYU’s backfield depth; both developments predate the opening lines and deserve little additional adjustment. Take the remaining cushion, with restrained conviction.

Total.Over 52.5 has slightly stronger support than the spread. EL's 54.36 projection provides only a small corroborating signal. CSU's Southern Utah box score supplies the better argument: 678 yards gained and 442 allowed, suggesting scoring potential on both sides despite the opponent-quality discount. Bachmeier's efficient Arizona performance strengthens the case for BYU production. The under deserves respect: BYU held Arizona to 17, CSU's explosive gains and favorable turnovers can regress, and a comfortable BYU lead could produce clock-consuming possessions. Moa's absence also reduces offensive depth. Nevertheless, CSU need not reproduce its opening scoring pace to contribute enough alongside BYU. The total has remained steady while the spread shortened, and the available forecast does not establish disruptive kickoff weather. I favor the offensive evidence without assuming a verified tempo increase.

Where our number looks wrong.The model cannot distinguish CSU's genuine offensive rebuild from turnover-assisted margins against weaker opponents. Its 6.37-point rating upgrade may overstate overall improvement while its totals baseline understates the new offense. The raw_margin_home field also has a sign inconsistency: the supplied ratings plus home field imply a 12.321-point BYU advantage, which should be positive under the stated home-spread convention. Finally, the totals documentation says no weather input, yet its components explicitly subtract 1.908 points for wind.

Spread.Part of that is news - starting rover linebacker Siale Esera is 'more than likely' out, on top of safety Raider Damuni (a month-plus) and running back Sione Moa - and part is a real reassessment of Colorado State, which is 2-0 for the first time since 2011 with Hauss Hejny running Jim Mora's offense. This is Fort Collins at altitude, at night, in front of a near-capacity crowd, against the first ranked team Colorado State has seen. My own number is around BYU -15 to -16. That leaves a point and a half of value rather than a fat edge, and I am buying it late after 3.5 points of it has already been taken, which is why this is a lean and not a position.

Total.Both scoring averages are inflated by cupcakes: BYU's 45.5 includes 63 against Utah Tech and Colorado State's 46.5 includes 58 against FCS Southern Utah. Strip those out and the FBS evidence is BYU 28, Arizona 17 (45 points, with Arizona shut out after halftime) and Colorado State 35, Wyoming 13 (48). That points to the high 40s, not 52.5. BYU has allowed 12 points a game and is the best defense either offense has faced. Add a forecast that carries a chance of showers and thunderstorms at a 5:30pm mountain kickoff and the tilt is down.

Where our number looks wrong.The model reads this as a 'pass' at a 2.1-point gap, which understates how much has actually changed. Colorado State's rating has climbed from a -10.85 preseason prior to -4.48 on the strength of a 58-24 win over an FCS team and a win over a weak Wyoming; the refit is rewarding margin against opponents that tell you very little, and I would not trust the size of that jump. On the other side the model cannot see that BYU is down a starting safety, a starting linebacker and a back, and its flat 2.5-point home field badly undersells Fort Collins at altitude at night against a ranked visitor. So the two errors happen to point in opposite directions and the 15.4 lands near where I would put it by accident, not by insight. Do not treat the 'pass' label as neutral information.

Spread.Our model and the market are close here (15.4 vs 17.5, a 2.1-point gap that is a pass on its own), but the market has moved 3.5-4 points toward Colorado State since Wednesday (BYU opened around -21, some books as high as -20.5/-21, now settled at -17.5, per youwager.lv and Yahoo, both 2026-09-13/14), a real signal that sharp money respects what new coach Jim Mora has built faster than a rating model anchored to a 2-10 preseason prior could. CSU is 2-0 and outscoring opponents 46.5-18.5 behind transfer QB Hauss Henjny (65% completion, 9 TD, 1 INT through two games), and a first-year staff with 75% roster turnover is exactly the profile our rating is slowest to pick up since it leans on last year's results and returning production. BYU is the better overall team and I would not pick them to lose, but I want the extra points against a home underdog the market itself is warming to, especially with BYU missing a starting safety.

Total.Our totals model (54.36) and the market (52.5) are close, so there is no real model signal here; this is a game-flow and weather call. BYU's defense has been excellent, allowing just 12 points per game through two contests including a true road test at Arizona, and that unit projects to slow down a Colorado State offense that has feasted on Wyoming and Southern Utah but has not faced a defense of this caliber. The dossier's captured forecast shows a 50% precipitation chance and 9.7 mph wind at Canvas Stadium, conditions that would work against Henjny's efficient early-season passing numbers if they hold. Both offenses can score, so this is not a strong lean, but the combination of BYU's proven defense and possible weather nudges me under the market number.

Where our number looks wrong.The model's rating for Colorado State is still dragged down by a -10.85 preseason prior built on a 2-10 2025 team; even after two wins the in-season refit (-4.48) has not caught up to what a new coaching staff and a clearly upgraded transfer quarterback have actually produced on the field. This is precisely the blind spot the model's own documentation flags: it cannot see a coaching change's real impact or a portal quarterback's performance, only final margins bent through a preseason-heavy prior.