Wagner at California. Pick: Wagner +58.5. Pick: Under 65.5.

Spread.Edge Labs has no spread number because Wagner is unrated; that is a model failure here, not evidence for either side. The best independent benchmark found makes California -35.7, while the market is -58.5; even Dimers’ much more pessimistic 60-4 projection covers Wagner. The market has clearly priced Wagner’s 87-3 loss at James Madison, but that rout included two punt-return touchdowns and a blocked-punt touchdown. Treating the entire 84-point margin as repeatable is aggressive. California produced only 249 yards at Syracuse, and its offense scored 14 points, the winning touchdown was a pick-six. Laying 58.5 also requires continued scoring after the outcome is settled, despite Clemson visiting six days later. Wagner’s cross-country trip and disastrous punt protection are serious risks, preventing a higher-confidence play. Still, +58.5 gives too much weight to one catastrophic result and too little to California’s modest offensive baseline. No reliable betting splits or official game-specific availability report could be verified.

Total.Under 65.5 is the stronger lean. The market equation is approximately California 62, Wagner 3.5, so the over largely requires California to reach a scoring level its offense has not approached: 24 points against UCLA and 14 offensive points at Syracuse. Wagner managed only 204 yards and three points at James Madison, but its poor offense alone does not create an over. The danger is non-offensive scoring after Wagner surrendered three special-teams touchdowns last week. Benign weather also helps execution. Even so, expected backup usage, Clemson six days later and California’s limited early production make outcomes such as 52-3 or 56-3 more plausible than the 62-plus California performance embedded in this total.

Where our number looks wrong.The spread model produces no number because Wagner is unrated, precisely where a usable FBS-versus-FCS adjustment is needed. Its California rating also learns from final margins and can overvalue the Syracuse win despite a 436-249 yardage deficit. On the total, assigning roughly a 1.3-point deduction to 6.6-mph wind looks excessive; the under case instead rests on game script and California’s offensive production.

Spread.Our model produces no ATS number here because Wagner (FCS) is unrated, so -58.5 is entirely the market's. The number demands a 59+ point win, and the case against it is Cal's own offense, not Wagner's competence. Cal has scored just 21 and 24 in its two games (22.5 ppg) under a first-year head coach, with QB Sagapolutele efficient but modest (career-low 182 yds vs Syracuse, 4 TD in 2 games). Wagner is genuinely awful (3-87 loss at JMU, 13.7 ppg, starting QB Jack Steven out with a neck injury), so a big Cal win is coming. But spreads this large are historically hard to hit: starters get pulled, scoring stalls, and the favorite must keep pressing. Local beat projections landed at 38-7 and 42-6 (roughly 31-36 point wins), both short of the number. Covering 58.5 AND staying under a 65.5 total is a narrow window that requires Cal to score in the high 50s while Wagner is shut out. I take the points.

Total.The market's 65.5 leans on Wagner's porous defense (87 allowed to JMU), but the ceiling is capped by Cal's own modest output: 21 and 24 points in its two games. Beat-writer projections (38-7, 42-6) imply totals of 45-48, well under. For the over to hit, Cal would need roughly 55+ while Wagner adds a couple scores against Cal's defense; Cal has not shown that kind of firepower under a new staff, and Wagner's offense (3 and 7 points in its two toughest games, backup QB) is unlikely to contribute much. Garbage-time scoring and Wagner's leaky defense add upside variance, which is why I stop at moderate conviction rather than a strong bet.

Where our number looks wrong.The spread model has no ATS number at all here (Wagner unrated), so it offers no read on -58.5 and cannot flag that Cal's demonstrated offense is far too weak to reliably clear a 59-point margin. It also cannot see Wagner's QB injury or Cal's questionable skill players.

Spread.Our rating model has no number here at all, Wagner is unrated, which the dossier flags as the least reliable case for the model. The market (Cal -58.5 to -57.5 across books, Covers.com 9/18) is effectively the only price signal. Wagner is dreadful: 1-2, outscored 87-3 by James Madison in week 2, allowing 40.7 ppg, and starting a backup QB (Benjamin Newman, in for injured Jack Steven since week 1) all season. That argues Cal could win by a lot. But at 58.5 points, the number already prices in a near-total wipeout, and two factors push toward the underdog covering: Cal is a rebuilding, first-year-coach roster (blue-chip ratio 6.7%, no proven blowout pedigree, only 22.5 ppg through two games against real competition) and Cal's staff has clear incentive to pull starters early and manage injury risk with Clemson on deck next week, capping garbage-time scoring. Extreme (55+) FCS spreads are historically inefficient in the market's favor once the outcome is decided. I take Wagner and the points.

Total.What tips it further under for me: Wagner has started a backup redshirt-freshman QB all season and its offense has managed only 13.7 ppg across three games against uneven competition; it is unlikely to sustain scoring against Cal. On the other side, Cal's own offense has been middling (22.5 ppg through two games) under a first-year coaching staff, and with Clemson looming next week the coaching staff has real incentive to shorten drives, run the ball, and pull starters once the game is decided rather than chase style points. Weather (68F, calm wind, no precipitation) is a non-factor either way.

Where our number looks wrong.The model produces literally no number for this game because Wagner has no preseason rating profile, its stated blind spot for FCS/unrated opponents is total, not partial. It also has zero visibility into Wagner's backup-QB situation or Cal's incentive to manage snaps ahead of Clemson, both of which matter more here than any rating would.