Louisiana Tech at LSU. Lean: Louisiana Tech +35.5. Lean: Under 55.5.

Spread.Louisiana Tech +35.5 is a modest price-sensitive lean. Our LSU -25.7 creates a 9.8-point discrepancy, but that is not an established betting edge. More concerning, the dossier incorrectly marks LSU as having no coaching change: Lane Kiffin and Sam Leavitt materially weaken comparisons with last year's offense. LSU's 644 yards against Clemson support a substantial upgrade, while Tech's 80-point FCS opener says little about its ability to compete here. Nevertheless, the captured market has already moved LSU from -31.5 to -35.5 at DraftKings, crossing the five-touchdown margin. That adjustment makes chasing LSU less attractive. Tech's Blake Baker showed enough passing production to make a late scoring drive plausible, although the competition caveat is enormous. Wright and Harris remaining off LSU's roster supplies no fresh downgrade relative to its opener. I take the extra half-point above 35, with low conviction because LSU's talent advantage could overwhelm Tech before substitutions matter.

Total.Over 55.5 at FanDuel is my slightly stronger lean. The model's 49.24 deserves consideration, but its historical offensive baseline misses LSU's coaching and quarterback transition. LSU's 644-yard opener provides a concrete reason to question that baseline; Tech also demonstrated passing efficiency, albeit against FCS opposition. The under case is substantial: LSU could suppress Tech almost completely, then shorten the game with reserves, and thunderstorms could disrupt offensive execution. Unconfirmed receiver availability adds uncertainty. Still, the forecast does not establish sustained heavy rain at kickoff, and the dossier's light wind is insufficient grounds for a large weather deduction. LSU scoring in the upper 40s would leave a manageable contribution required from Tech.

Where our number looks wrong.The first_year_head_coach=false input is demonstrably wrong: LSU is beginning the Kiffin era. The spread likely understates LSU's offensive ceiling, while Tech's FCS rout is weak evidence for its rating. The total's reliance on LSU's 2025 offense likely biases 49.24 downward. Neither model gap justifies strong conviction.

Spread.Our number (LSU -25.7) says take the points by nearly ten, but most of that gap is model error, not market error. The preseason prior carries LSU at 14.1 with first_year_head_coach=false; Lane Kiffin is in fact in his first LSU season (Wikipedia 2026 LSU team page, read 9/9/26) with 21.2% returning production, so the rating is grading a 7-6 2025 roster that no longer exists - one that just beat Clemson 51-10 (9/5/26). On the other side LA Tech's 0.198 is fed by an 80-6 win over FCS Northwestern State (9/5/26), exactly the input the model warns is least reliable. Re-priced honestly, LSU -33 to -38 is defensible and the market is not obviously wrong. I still take the dog, on price rather than power ratings: all five books opened -31.5 to -33.5 on 9/6 and sit -35/-35.5 now, a two-to-four point run with no injury news behind it, and 35.5 is a five-touchdown number a Kiffin team stops chasing once the third string is in. Weak lean, low slate priority.

Total.Both models point under and I agree, mildly. The market (55.5-56.5) implies LSU near 46 and Louisiana Tech near 10; LSU would essentially have to repeat its 51-point Clemson night while emptying the bench. The real under driver is game script - once LSU is up 35 in the third, the clock runs and the backups take over. Two LA Tech receivers listed questionable 9/6 and still unconfirmed trim the dog's scoring further.

Where our number looks wrong.Two concrete errors. (1) LSU's preseason_v1 has first_year_head_coach=false. That is factually wrong - Kiffin is in year one (Wikipedia, 9/9/26) - so whatever coaching-change term the prior applies never fired, and the rating is instead anchored to a 7-6 2025 team with 21.2% returning production, i.e. almost none of the roster that hung 51 on Clemson. The 19.5 in-season value after one game is still mostly that stale prior, which is the entire reason our number reads -25.7 while the market reads -35.3. (2) The gap is inflated from the other end too: LA Tech's rating rose off an 80-6 win over FCS Northwestern State, and the model's own weakness note says the number is least reliable against an FCS opponent - yet nothing discounts that game before it feeds the Elo refit. Both errors push the same direction, so the '9.6-point lean away, moderate' should be read as roughly zero real edge, not as a priced one. The totals model has the milder version of the same problem: its early-season weight is thin, so it is pricing 2025 LSU's offense, not Kiffin's.

Spread.Our number (LSU -25.7) disagrees with the market (-35.3) by 9.6 points, flagged 'lean away' at moderate. I'm fading that lean. The gap comes from a rating that, after one game, is still mostly LSU's modest preseason prior; it cannot see that LSU just throttled Clemson 51-10 (644 yards, dual-threat QB Sam Leavitt: 230 pass/113 rush/2 TD) while Louisiana Tech's 80-6 win came against FCS Northwestern State. Our own backtest shows zero proven ATS edge at any gap size (46-50% in every bucket across 9,927 held-out games), so a 9.6-point disagreement here isn't evidence of value, just model staleness. Louisiana Tech's only reported absences are two questionable WRs; LSU has no injuries listed. I'll lay the number, but keep confidence moderate since backdoor-cover risk on a 35.5-point spread is real if LSU rests starters early.

Total.Real factors lean under: the Baton Rouge forecast shows a 70% chance of showers/thunderstorms right around and after kickoff (weather.gov, retrieved 2026-09-09), and games with 35+ point spreads routinely see the favorite run the ball and empty the bench once the lead is safe, capping second-half scoring. LSU's defense held Clemson, a real opponent, to just 10 points in Week 1, and Louisiana Tech's offense now has to sustain drives against a defense of that caliber rather than an FCS one. Taking the under at the lowest number currently available.

Where our number looks wrong.Our rating is still mostly the preseason prior after one game and can't see that LSU just posted one of the more decisive results of the young season (51-10 over a power-conference opponent, 644 yards) while Louisiana Tech's margin came against an FCS bottom-feeder. The market has already moved 2-4 points further toward LSU since the weekend, sharp book included, and our own backtests show zero proven ATS edge at any gap size, so the model's 'lean away' likely reflects stale priors rather than a real mispriced number. I'm fading it, though a -35.5 lay still carries genuine backdoor-cover risk if LSU empties the bench early or the rain slows the game down.