Week 1 of the college football season brings 10 picks to the board — 5 against the spread and 5 on totals, with the first kicking off Thursday. Colorado +6.5 at Georgia Tech is the strongest read on it, and the top 3 are broken down below. The sharpest disagreement is on Liberty +6.5 at James Madison, where 16% of the bets account for 70% of the money. Florida Atlantic +27.5 at Florida has moved 1 point against us since we locked it, so the number on the card is better than what is available now.
Colorado +6.5 at Georgia Tech — Thu, Sep 3 · 8:00 pm ET. Four angles line up here: sharp money, a public fade, a reverse line move and a revenge spot. Our side has 13% of the bets but 26% of the money. Public betting splits are the other way, with 75% of tickets on the other side.
UMass +29.5 at Rutgers — Thu, Sep 3 · 6:00 pm ET. Three angles line up here: sharp money, a public fade and a reverse line move. Our side has 33% of the bets but 73% of the money. Public betting splits are the other way, with 66% of tickets on the other side.
Baylor-Auburn under 57.5 — Sat, Sep 5 · 3:30 pm ET. Two angles line up here: sharp money on the total and a total on the move. Our side has 17% of the bets but 89% of the money. We locked 57.5 and it is 59.5 now, so the entry is worse than what is on the screen.
The other 7 picks are on the board below, each with the line we locked and what it is doing now. For the full slate, see college football odds, line movement, SP+ rankings and every CFB pick.
This board refreshes every week. Entry lines are locked when a pick is made and are not updated as the market moves; the line shown beside each entry does move. Every pick here is graded and carried in the record below.
Colorado at Georgia Tech
UMass at Rutgers
Clemson at LSU
Liberty at James Madison
Florida Atlantic at Florida
West Georgia at Kennesaw State
Albany NY at Buffalo
Kent State at South Carolina
Baylor at Auburn
Marshall at Penn State
Colorado at Georgia Tech
How these picks are made
Every pick starts from betting splits — how many tickets are on each side, and how much money. We compare that against how the line has moved since it opened, plus the weather and the history between the two teams. Nothing here is a model prediction of who wins; it is a read on where the money is going and whether the market agrees.
What we look for
- Sharp money
- The share of dollars on a side runs at least 15 points ahead of the share of bets. Lots of small tickets on one team and fewer, larger ones on the other is the pattern we are looking for; it is the single strongest input we have.
- Fade the public
- At least 70% of tickets are on one side. On its own that means nothing — the crowd is often right — so it only counts when the money or the line is going the other way.
- Steam / RLM
- The spread has moved at least 1 point since it opened. When it moves toward the side the public is NOT on, that is a reverse line move, and it is weighted higher than an ordinary move.
- Total on the move
- The total has moved at least 1.5 points since open.
- Weather under
- Forecast at kickoff shows sustained wind of 15 mph or more, precipitation odds of 60% or higher, or a temperature at or below 20°F.
- Revenge spot
- The team we are backing lost the last meeting in this series. It is the weakest input on the board and can never carry a pick by itself.
How a pick gets rated
A pick only rates above the base tier if sharp money is present on the market being picked — sharp money on the total does not strengthen a spread pick. From there it is a question of how much else agrees:
- Lean
- Cleared the bar, but without sharp money behind it or without anything corroborating.
- Strong Lean
- Sharp money on this market, plus one other signal agreeing.
- Strongest Signal
- Sharp money on this market, plus two or more others agreeing.
Entry lines and grading
The entry line is the number recorded the moment a pick first qualified, and it never changes afterwards. The line beside it does move, and when the two differ we show the gap — including when it has moved against us. Results are graded from final scores after the game, and every pick is carried in the record whether it won or lost.
A pick marked first sighting was already qualified the first time we saw the game, so we cannot say what the number was when the signal actually fired. Those grade win or loss like any other pick, but they are left out of the closing-line average, because that figure would not mean anything for them.
None of this is a guarantee, and no one should bet more than they can afford to lose.
Jason Ziernicki is the founder of CLEATZ, where he analyzes sports betting data, public betting percentages, alt-line trends, and prediction markets across the NFL, NBA, MLB, and college sports.
He is based in Jackson Hole, Wyoming, where he routinely trades on Kalshi each month, hoping to win on weather markets like snowfall, as well as sports and politics.
His work focuses on turning sportsbook data and betting market trends into actionable insights for bettors/traders.
