How the game projects

Georgia Tech has the stronger baseline. The path still runs through down-to-down execution.

The model makes Georgia Tech the neutral-field leader after blending opponent-adjusted efficiency, predictive ratings, roster quality, returning production, portal impact and special teams. Home field contributes to that position. The median score is Georgia Tech 31 · Colorado 23, with roughly 12.0 possessions per team.

This is a distribution, not a promise. The home team wins 67% of the model's outcome curve; the remaining probability captures turnovers, finishing-drive variance, injuries and the uncertainty that is highest early in a season.

The number bridge

Every point between neutral field and the final line.

The location calculation is no longer a blanket home-field number. It combines the learned national baseline with venue and conference history, capacity, altitude, travel, body clock and rest while the archived team ratings remain neutral-field numbers.

Matchup engine

Where the projection is coming from.

Georgia Tech's cleanest route is to keep the game in standard downs and make its stronger units appear repeatedly. Colorado's upset path is higher variance: create short fields, win explosives and force the favorite to finish long drives instead of allowing easy scoring opportunities.

Personnel file

Continuity, transfers and the people who can move the number.

Market translation

Over 50.5.

The independent line remains the 4M football opinion. The market-calibrated line shrinks that number only when uncertainty is elevated; it never changes the underlying team rating.

What needs to happen: a side position needs the projected unit advantages to survive first down and the turnover margin to stay near even. A total position needs the expected possession count and scoring efficiency to hold. If the market moves through the model number, the original edge is gone.