Tennessee at Georgia Tech. Lean: Georgia Tech +12.5. Lean: Under 55.5.

Spread.Georgia Tech +12.5 is a modest price-based lean. EL makes Tennessee an 11.4-point favorite, only 1.1 points below this spread; that is not an established edge. The dossier shows Tennessee moving from -8.5/-9.5 before the openers to -12.5, but retreating from -13.5 at two later-opening books. Much of the contrasting Week 1 impression is therefore already priced. Tennessee now asks freshman Faizon Brandon to cover a substantial road number after facing Furman. Tech has two additional preparation days and Mendoza's efficient passing debut offers a route to staying competitive. The counterargument is substantial: Tech struggled badly running the ball, and Tennessee gets linebacker Arion Carter back. That return reduces my enthusiasm for the underdog and is more relevant than simply repeating the Harmon injury. Reported betting splits lack a sufficiently clear sampling basis to establish sharp action. I see no verified, unpriced personnel advantage; this is a small lean on the home cushion.

Total.Under 55.5 is my stronger lean despite EL's 58.4 projection. Tech's opener produced only six offensive points, with its touchdown coming on a kickoff return, and its running game struggled. Carter's return and Knowles' defensive installation add reasons to discount inherited offensive expectations. The over case deserves weight: Mendoza passed efficiently, Brandon supplies explosive upside, and Tennessee's scoring ceiling could carry this total. Defensive dominance against Furman is weak evidence, while rain can create short fields as well as suppress execution. I therefore apply only a small weather discount, with no verified strong-wind forecast. The dossier's largely stable 55 to 55.5 market also supplies no convincing under steam. Offensive uncertainty and Tennessee's defensive personnel lead me below the market, but not to a strong position.

Where our number looks wrong.The spread estimate is too precise for final-margin ratings after one game, especially Tennessee's FCS opener. It misses Brandon's first road-start uncertainty, Carter's return and the defensive coordinator change. The 58.4 total looks potentially high because inherited scoring ratings may overstate Tech's rebuilt offense and understate Tennessee's defensive improvement. Also, raw_margin_home=-9 conflicts with the dossier's stated home-perspective sign convention; audit that field before interpreting it.

Spread.Our number is Tennessee -11.4, the market is -12.5, a 1.2-point gap our own system reads as a pass. SP+ is the outlier at Tennessee 35-20 (SI, 9/8). So no model here has a real edge, and the question is what the four-point move actually priced. It came off the two noisiest samples in football: a 56-9 win over an FCS team and a 14-13 home loss decided by a blocked 45-yard field goal at the gun. Action Network shows 78% of tickets on Tennessee (9/9), so that move is public, not sharp. What is underpriced is the spot: Faizon Brandon's first career road start (SI, 9/8), at night, in a hostile stadium, with an 80% chance of showers and thunderstorms at kickoff (NWS Atlanta, 9/9), against a defense that held Colorado to 14. Tennessee also arrives short: Dodson out for the year, Duarte hurt, Harmon's ACL, Farooq out at safety. Two touchdowns is a lot to lay into that script. I want the points.

Total.I am fading our own totals model, which says 58.4. Its documented blind spots are weather and roster turnover, and those are the whole story here. NWS Atlanta (forecast 9/9) has showers and possibly a thunderstorm before 8pm Saturday with an 80% chance of precipitation; the 7:00pm ET kickoff sits inside the wettest window. Georgia Tech's offense produced six points at home in week 1 on 18-of-38 passing and 51 rushing yards, and Brent Key's stated identity is physical, clock-eating, run-first football against a Tennessee front that he will keep testing. Tennessee's 56 points came against an FCS team with a freshman quarterback now making his first road start in the rain. Rain, a broken home offense and a freshman road quarterback all point the same way.

Where our number looks wrong.Two places. First, the in-season rating reads Georgia Tech as having scored 13 points; it cannot see that 7 of those came on a 105-yard kickoff return, so the offense that actually has to move the ball on Saturday is worse than our number thinks. Second, Georgia Tech's rating is still mostly its preseason prior, and that prior (2.926, 6.55 expected wins) is anchored on a 9-4 2025 with 12.6% returning production, a new offensive coordinator in George Godsey and a first-year transfer quarterback - the exact roster profile the prior handles worst, and it is flattering Tech here. Net, our -11.4 is probably a touch too kind to Georgia Tech, which is why SP+ sits at -15 and why I hold this at a lean rather than a strong position. I am taking +12.5 on the spot, the weather and the source of the four-point move, not because our number says the market is off - by our own gap rule this is a pass.

Spread.The lean comes from context the model can't see. Georgia Tech's offense was legitimately broken in the opener: 51 rushing yards and 13 points in a 14-13 home loss to Colorado, 115th nationally in total offense, and HC Brent Key called the performance 'poorly coached, soft' (AJC, 9/3/26). Two starting WRs (Collins, Plummer) are Questionable with undisclosed injuries since 9/4. Tennessee's 56-9 opener came against FCS Furman, garbage-time inflated, not a reliable signal against a real defense, and GT's own defense actually held Colorado to just 14. The line has drifted from -8.5/-9.5 to -12.5, past the expensive 10-14 key-number cluster; a wounded-but-competitive home dog gets the points here, and one independent handicapper (Doc's Sports, 9/8) landed on the same side.

Total.It carries a wind term but nothing for rain: the fresh NWS forecast (checked 9/9/26) has Saturday's Atlanta kickoff at an 80% chance of showers/thunderstorm, wetter than our own captured forecast. Combined with a Georgia Tech offense that managed just 51 rushing yards and 13 points in the opener, and a Tennessee scoring line inflated by an FCS blowout rather than a real defense, the under is the more defensible side of this total.

Where our number looks wrong.The 1.2-point gap is inside our own no-edge zone, so the model isn't asserting a real signal either way. Its bigger blind spot: it can't see that GT's offense (51 rush yards, 13 points, HC calling it 'soft') was the actual problem while GT's defense held Colorado to 14 - the in-season refit is still mostly last year's priors plus a one-game sample, and it can't distinguish a broken offense behind a competent defense from a broken team overall. It also has no visibility into the two questionable GT WRs or Tennessee's linebacker churn (Harmon out for the year, Carter back from suspension), and it weights a 56-9 FCS blowout the same as a P4 result when that game is mostly noise for grading Tennessee's real quality.