Colorado vs Georgia Tech Prediction, Picks & Odds — Sep 3, 2026

Georgia Tech opened last season in Boulder and left with a 27-20 win, and Haynes King carried it out of the building himself: 19 rushes, 156 yards, three touchdowns on an afternoon he barely had to throw. Twelve months later the sites flip, and King is gone. So is Colorado's starter. Bobby Dodd hosts two quarterbacks who have started a combined two college games, and most of the players catching and carrying the ball around them are new as well.

The records point one way — Georgia Tech went 9-4 and scored 32.2 points a game, Colorado went 3-9 and scored 20.9 — and both rosters behind those numbers have been stripped. Georgia Tech returns 6% of last season's passing production; Colorado returns none of its rushing production. What arrived to replace all of it is the game.

◄ColoradofavorsGeorgia Tech►

Last season's body of worklast year's power rating, regressed toward the averageGeorgia Tech +19

Roster talentoverall talent level of the roster — recruiting pedigree plus an independent preseason roster gradeColorado +18

Transfer portalnet proven talent gained vs lost in transfers — it raises or lowers a roster's projected strengthColorado +4

Each ingredient of our projection and the side it favors — bars toward Georgia Tech are why it's the pick (home field is counted separately).

GRIDPEX.COM

Last season's body of work is what favors Georgia Tech, and it is the single biggest thing separating these teams — bigger than the home field that comes next. Everything else tilts the other way: Colorado's roster is the more decorated, 30th nationally in talent to Georgia Tech's 39th, and it churned harder in the portal, 42 in and 38 out against Tech's 19 and 20. The distance between last year's ledger and this year's personnel is what makes this close.

Two quarterbacks, two kinds of unknown

Alberto Mendoza has the job. Brent Key said in fall camp that the Indiana transfer would start "if we played this weekend." Mendoza has also never started a college game. His career is 25 pass attempts — 24 of them in relief at Indiana last season, for 286 yards and five touchdowns, nearly 12 yards every time he threw — plus 13 carries for 190 yards. Real ability, in a very small sample.

What he inherits is the issue. King threw 361 of Georgia Tech's 394 passes for 2,951 yards and 14 touchdowns, and he also took 185 of the team's 462 carries for 953 yards and 15 scores. Georgia Tech ran for 31 rushing touchdowns last season; the quarterback kept it on half of them.

The combination has a track record. Since 2013, 49 teams lost a starter who ran nearly 40% of their rushing plays and handed the job to a quarterback who had barely played (n=49, 2013-2024). Those offenses slid from 30.6 points a game to 27.8, and 73% of them scored less than the year before. Openers were the exception: that group averaged 30.3 points in game one and won 62.5% of them. The erosion arrives over months.

Colorado's situation is quieter. Julian Lewis started two games as a redshirt freshman, threw for 589 yards with four touchdowns and no interceptions, and enters locked in. He is not King — 23 carries for minus-36 yards tells you the quarterback run game is not part of the plan — but he has taken FBS snaps, and Kaidon Salter was nowhere near as central to Colorado as King was to Georgia Tech.

When Georgia Tech has the ball

Georgia Tech's rushing offense ranked 28th, and may be better this fall despite losing its leading rusher. Jamal Haynes carried 124 times for 531 yards, 4.3 a pop, the least productive significant runner on the roster. In comes Justice Haynes, who ran for 857 yards and 10 touchdowns on 121 carries at Michigan — 7.1 a carry, fourth-best among the 220 FBS backs with at least 100 attempts. Malachi Hosley is back after 98 carries for 697 yards and seven scores, following 1,192 yards at Penn. Four of the five projected linemen are holdovers; only center Joseph Ionata, from Alabama, is new.

They run at a Colorado front that has been emptied and refilled. Linebacker Johnny Chaney Jr. (99 tackles) and safety Tawfiq Byard (79 tackles, now at Texas A&M) are gone, and the Buffaloes return about 6% of the run-defense production that ranked 71st. The replacements came by portal: linebacker Liona Lefau from Texas, edge rushers Vili Taufatofua (San José State), Immanuel Ezeogu (James Madison) and Toby Anene (North Dakota State), interior men Santana Hopper (Tulane) and Dylan Manuel (App State). Real pedigree, not one snap together.

Through the air, Georgia Tech's 15th-ranked passing attack is a memory. Eric Rivers (47 catches), Malik Rutherford (46), Isiah Canion (to Georgia) and Bailey Stockton (to Florida) all left alongside King. The new room is Elon's Isaiah Fuhrmann (46-907-9 in FCS), New Mexico State tight end Gavin Harris (36-514-1), holdover Jordan Allen (22-304) and Cal's Jaiven Plummer — roughly 1,800 combined receiving yards, half of it below this level. Colorado's pass defense ranked 125th and returns 5% of it after losing corners Preston Hodge (13 breakups) and Colton Hood; Boo Carter (Tennessee), Randon Fontenette (Vanderbilt), Cree Thomas (Notre Dame) and Jason Stokes Jr. (Utah) replace them.

When Colorado has the ball— its offense vs Georgia Tech’s defense

Passing: Colorado O #81 vs Georgia Tech D #85Colorado · slight edge

Rushing: Colorado O #99 vs Georgia Tech D #65Colorado · slight edge

◄ Georgia Tech defense winsColorado offense wins ►

When Georgia Tech has the ball— its offense vs Colorado’s defense

Passing: Georgia Tech O #15 vs Colorado D #125Georgia Tech · slight edge

Rushing: Georgia Tech O #28 vs Colorado D #71Georgia Tech · slight edge

◄ Colorado defense winsGeorgia Tech offense wins ►

Read it by possession: each side shows what happens when that team has the ball — its offense against the other team's defense, unit by unit (national rank in parentheses). Last season's grades, shrunk toward average by how much of each unit's production actually returns. Colorado returns 19% of last season's passing production and 0% of its rushing production. Georgia Tech returns 6% of last season's passing production and 35% of its rushing production. Colorado's projected line: 2 of 5 held over, 3 portal; Georgia Tech's projected line: 4 of 5 held over, 1 portal.

GRIDPEX.COM

When Colorado has the ball

On rank alone this is the closest crossing on the field — Colorado's 81st passing offense against Georgia Tech's 85th pass defense. The personnel makes it lopsided. Danny Scudero led all of FBS with 1,297 yards on 88 catches and 10 touchdowns at San José State, with four games over 150 including a 215 and a 189, a year after 650 yards at FCS Sacramento State. He projects into the slot. Outside him: DeAndre Moore Jr., 38-532-4 at Texas, and Kam Perry, 43-976-6 at Miami (Ohio) at nearly 23 yards a catch. Joseph Williams (37-489-4) actually returns. Those four produced 3,294 receiving yards last season — more than Colorado's entire team did (2,433), and Omarion Miller's 807 left for Arizona State on top of it.

The secondary that has to handle all of that does not exist yet. Georgia Tech intercepted three passes all season, 128th of 136 teams, and that was with corners Nehemiah Chandler (two picks, 13 breakups) and Ahmari Harvey playing. Neither is back. Not one defensive back is. The projected corners are Alabama transfer Jaylen Mbakwe — a four-star who was catching passes in Tuscaloosa a year ago — and South Florida's Jonas Duclona, with Kelvin Hill at nickel. Tech lists 17 defensive backs, nine of them true freshmen, and the recruiting class leaned the same way, headlined by four-star safety Traeviss Stevenson and corner Jaedyn Terry. Scudero working the slot against Hill, with safety help that has never played a snap together, is the most exploitable matchup in the stadium.

The ground offers Colorado nothing. The Buffaloes averaged 3.58 yards a carry, 99th nationally, and return none of it — Dallan Hayden and Simeon Price are gone, and Salter took 104 carries with him. Micah Welch (96-384-4), DeKalon Taylor and Sacramento State's Damian Henderson II (91-565-5 at 6.2 a carry) share a lead job that isn't settled, and they run at the one Georgia Tech unit that held together: Kyle Efford, Cayman Spaulding and Brayden Manley, plus new nose tackle Tawfiq Thomas, who transferred in from Colorado.

Nobody in this game gets to the quarterback

Mendoza's debut comes with unusual cover. Colorado produced 13 sacks in 12 games last season, 127th nationally, and the edge room that managed them has been swapped out entirely — so the classic first-start disaster, hit early and rattled into a giveaway, is the least likely thing on this field. Georgia Tech's rush is no scarier. Romello Height (9.0), Akelo Stone (4.0) and Jordan van den Berg (3.0) accounted for 16 of its 26 sacks and are all gone, leaving 11 spread across Manley (3.0), Spaulding (2.5), Amontrae Bradford (2.5), Efford (1.5) and Christian Garrett (1.5), with Alabama four-star Noah Carter the headline addition. Neither of these defenses has shown it can force the mistake, and the team that wins the giveaway count takes roughly 77% of games priced like this one (n=2,898 since 2014) against 75% leaguewide.

The floor, the ceiling, and the swing

Expect roughly 50 points and a game alive in the fourth quarter. Better than four times in ten this finishes inside one score; about one in five becomes a 20-point win for either side; eight times in ten the margin lands between Colorado by 18 and Georgia Tech by 22. The most likely single scoreline is 24-17. The winning scripts are nearly identical: Georgia Tech's wins average 31.9-18.2, Colorado's 30.9-18.5. Whoever takes this scores about 31, so the rock fight the visitor is supposed to want is a losing script for both sides — Colorado has to throw its way to the mid-20s, because nothing else gets it there.

The swing is first-and-goal from the four. Last September that snap was a King keeper; this year it is a handoff to Justice Haynes or Hosley against a front seven of strangers, and whether those drives end in touchdowns or field goals separates a comfortable afternoon from a nervous one.

1%

-38

2%

-32

5%

-24

8%

-18

13%

-10

17%

-4

18%

+4

15%

+10

11%

+18

7%

+24

3%

+32

1%

+38

0%

+46

Colorado winsGeorgia Tech wins

Market: Georgia Tech −6Projected: Georgia Tech by 1 (53%)

The range of outcomes for this game. Each bar is how often the final margin lands in that range — the more it leans toward Georgia Tech, the more likely Georgia Tech wins.

GRIDPEX.COM

Keys to the number

Georgia Tech is laying 6.5. What has to be true on the field for the Yellow Jackets to clear it:

• Score on nearly every possession. Georgia Tech finished 78th of 136 in points per scoring chance last season, and only about a fifth of the offense behind that grade is back. Clearing a number like this is a drive-efficiency day — points on almost every trip, nothing wasted — not one long touchdown. Coin-flip-ish.
• Win the turnover battle by two. Short fields are the fast route past 6.5, and Georgia Tech finishes plus-two in the giveaway count in fewer than one game in five. A long shot.

And what Colorado needs to stay inside it:

• Control the ball through the air. Colorado's one genuine edge is its passing game against an 85th-ranked defense; leaning on Scudero and Moore shortens the game and limits Georgia Tech's possessions. Coin-flip-ish.
• Force the low-scoring version. Colorado's covering games are low-scoring only about a third of the time, and the total sits near 50. Unlikely.
• Win or tie the giveaway count. Two-thirds of the time this game finishes even there, which is the world Colorado survives in; one lost fumble in Georgia Tech territory ends it. Likely.
• Steal a possession. A fourth-down conversion or a special-teams play worth an extra drive is the cheap way to manufacture the scoring chance the math requires. Unlikely.

Colorado's checklist is the more achievable one. Protecting the football and staying on the field with a veteran receiving corps are things this team can plausibly do; Georgia Tech's route to the number asks for a near-flawless scoring day from an offense that has to introduce itself first. Winning and covering are different jobs — the Yellow Jackets are the more likely winner, but a methodical favorite tends to land short of a number like this, and Tech clears 6.5 about four times in ten.

The bottom Line

Georgia Tech 25.9, Colorado 23.8 — the Yellow Jackets by about two, with the win itself close to a coin flip even with a 9-4 team hosting a 3-9 team. A home side with the better backfield and a quarterback nobody has film on, against a visitor carrying the best receiver on the field and no ground game at all. Colorado's path is narrow and completely live: Lewis to the mid-20s through Scudero, Moore and Perry, with a clean ledger. If Justice Haynes finishes the drives Haynes King used to finish, this stops being close in the third quarter.