The Colts Built Their Roster Backward. The Data Shows the Way Out.

About 30% of the Colts’ payroll goes to running back, guard and cornerback. Those are three of the positions that do the least to win championships.

“Our standard is to compete for championships,” owner Carlie Irsay-Gordon said in a statement posted to the team’s social media accounts announcing the firing of general manager Chris Ballard, “and I believe a change in leadership is necessary to meet that standard.”

The standard is the right one. The formula isn’t. Championship teams pour their money and draft picks into the quarterback and the passing game. The Colts have done the opposite, and a new general manager will only meet that standard by changing how the team values positions.

Contending teams spend 21% on those three positions. Since 2017, championship teams got about 0.2 extra points a game, at most, from each of those positions. From quarterback, they got 2.3.

Expectations were high. The Colts opened 2025 at 7-1 behind Daniel Jones’s best season, when he ranked 6th among quarterbacks in passing efficiency. Then they spent like a contender. They traded two first-round picks for cornerback Sauce Gardner. This offseason they paid Jones $44 million a year, receiver Alec Pierce $28.5 million and running back Jonathan Taylor $22 million.

But the roster never pointed to a big jump. Before the season, a projection based on player ratings had the Colts at 8.9 wins. That ranked 17th in the league, less than one win better than last year’s 8-9.

Instead they are 2-4, losing by 3.3 points a game on average and on pace for about 7 wins. Jones has dropped from 6th to 26th of 32 starting quarterbacks.

Where wins come from

The chain is relatively simple. Wins come from outscoring opponents. An average team breaks even over a season, with a point differential of zero, and wins about 8.5 games. Every 36 extra points over a season, about 2 a game, adds roughly one more win.

A typical champion outscores opponents by 8.4 points a game, which works out to about four wins above average. The Colts are at −3.3 a game, which points to about 7 wins. The biggest driver of that gap is an efficient passing game. In any given game, the team that passes more efficiently (gains more per pass play, counting sacks and turnovers) wins 75% of the time. Quarterback play alone explains about 40% of the gap between contenders and average teams.

Not all positions are created equal

The table shows how many points per game each position added for championship teams compared with an average team (2017–2025), next to how much of their salary cap the last nine Super Bowl champions spent there and how much the Colts spend in 2026.

Cap shares in the table are of the full salary cap. The 30% and 21% figures at the top are shares of payroll, the money actually committed to players, so they run higher. Champions are the nine Super Bowl winners from 2017 to 2025, a small sample, and the interior defensive line figure rests most on it: across all contending teams, that position added only 0.15 points a game.

The Colts are strong in the wrong places. In 2025, their best groups after quarterback were running back and guard, two of the least valuable positions. Cornerback, where they spend 11% of the cap, was below average.

Guard stands out. Contending teams actually spend less there than other teams, and teams that pay their guards more tend to win by less.

Even an elite running back can’t move the needle much. In 2025, Jonathan Taylor led the NFL’s most productive backfield through a season that began 8-2, and it added only about 0.7 points a game. That is the best case at the position, and it’s less than a third of what champions get from their quarterback. Cornerback and linebacker add the least of all.

Money alone doesn’t buy wins. How much a team spends at a position barely tracks how well it plays. Contenders win by getting good players cheaply: 45% of their production comes from players still on rookie contracts.

The Colts are paying for the wrong things

The Colts spend more than contenders on guard, cornerback and running back, the positions that add the least. They spend less on receiver and on tackle, the one position where spending more tends to mean winning more. The table above shows the gaps.

It’s about to get worse. Gardner counts only $9.5 million against this year’s cap because part of his contract stayed on the Jets’ books. His cost jumps to $21 million in 2027, $26 million in 2028 and $36 million in 2029 and 2030.

At quarterback, the Colts pay about what champions do (10.3% of the cap vs. 10.6%) without getting top-10 play in return.

Case study: the Sauce Gardner trade

In November 2025, a 7-2 Colts team sent its 2026 and 2027 first-round picks plus young receiver Adonai Mitchell to the Jets for cornerback Sauce Gardner.

By the numbers, the cost was steep:

• What they gave: two first-round picks and a young receiver, together worth about $40–45 million in expected production.
• What they took on: $129 million owed through 2030, for a player whose play is worth about $14 million a year. That’s roughly $65 million too much.
• What position they bought: cornerback, which adds about 0.1 points a game for champions. Quarterback adds 2.3.
• What they gave up at quarterback: each first-round pick is about a 1-in-5 chance at a franchise quarterback. The trade used two.

As a bet on winning in 2025, the trade made sense. Instead the Colts went 1-7 afterward and lost their starting quarterback in Week 14. Gardner has also played worse, allowing 11.1 yards per pass thrown his way this season, up from 6.7 last year. He also suffered a mild left calf strain in the Week 6 loss to the Tennessee Titans and is considered week-to-week, according to reports.

Why the quarterback has to be drafted

The Colts won’t find a franchise quarterback in free agency or by trade, the approach they’ve used since Andrew Luck retired in 2019. Philip Rivers, Carson Wentz, Matt Ryan, Joe Flacco and Daniel Jones all came in as veterans. Only Rivers (2020) and Jones (2025) had a top-10 season, and neither did it twice. Their one first-round try, Anthony Richardson in 2023, struggled when he played: in 2024 he ranked 36th of 39 quarterbacks in passing efficiency. Moving on was the right call, and the answer is to try again, as laid out below.

Why aim for a top-10 quarterback? Because that’s what contending takes. Since 2016:

Eight of the last 10 Super Bowl champions had a top-10 quarterback that season, and so did 34 of the last 40 conference finalists.

Proven quarterbacks almost never become available. Of 100 top-10 quarterback seasons since 2016, only three were followed by a move to a new team. In each case the old team had already replaced him (Alex Smith, Case Keenum and Jimmy Garoppolo).

Contenders build their own. Since 2017, about half of contending teams’ starting quarterbacks were drafted by that team, and 39% were still on cheap rookie contracts. A cheap quarterback leaves 15–20% more of the cap for everyone else.

The catch: each first-round quarterback is a long shot.

Where a quarterback is picked barely matters. Patrick Mahomes went 10th, Josh Allen 7th and Lamar Jackson 32nd. What matters is how many tries a team takes. One first-round quarterback works out about 20% of the time. Two tries raise it to 36%, and three to 49%.

Losing on purpose for a high pick barely helps. Dropping from about 7 wins to 4 raises the odds of landing a franchise quarterback only from about 16% to 22%.

Does Shane Steichen stay?

The data supports keeping him through 2027, then judging him against clear targets.

*All of the 2025 shortfall came in games started by a backup quarterback.

His offenses do more with less. From 2023 to 2025 they won about one more game a season than their talent projected, 8th best of 59 head coaches since 2017. Jones had his best season under him in 2025, ranking 6th after years below average.

He has helped build a Super Bowl offense. With him as offensive coordinator, the Eagles’ offense rose from 29th to 11th in 2021, then 3rd in 2022, when they reached the Super Bowl. Their quarterback improved from 17th to 9th. The record isn’t perfect: the 2020 Chargers offense he ran ranked 15th.

Keeping the same coach helps young quarterbacks. First-round quarterbacks who kept their head coach through year three became top-10 starters 26% of the time, versus 10% when the coach changed. Struggling teams are also the ones that fire coaches, so the real effect is likely smaller than that gap.

The roster is the bigger problem. Even the best coach since 2017 added about 3 wins a season beyond his roster. Steichen adds less than half a win. The Colts are about 5 wins short of a typical champion.

What would cost him the job:

• A 2027 finish more than one win below what the roster should produce, with one of the league’s worst offenses for its talent.
• Two straight seasons below what the roster should produce.
• A 2028 rookie quarterback who gets worse in his second season.

The road to 2030

The Colts’ next first-round pick comes in 2028, the same year a lot of salary comes off their books. That makes 2028 the year to draft a quarterback, and 2030–31 a realistic window to contend.

Checkpoints to watch, based on what contenders look like:

• 2027: close to even with opponents; spending on running back, guard and cornerback under 26% of payroll.
• 2028: outscoring opponents slightly; a first-round quarterback starting by midseason; spending on those three positions under 21%, like contenders.
• 2029: winning by 4–6 points a game and making the playoffs; the quarterback in the league’s top half after two seasons.
• 2030: winning by 7–9 points a game and 12 or more wins, about what a typical champion does.

If the quarterback isn’t the one

Four out of five first-round quarterbacks never become top-10 starters. So the Colts need to know early whether they have the right one, and what to do if they don’t.

Give him two full seasons. Rookie years are misleading. Jared Goff and Josh Allen were below average as rookies, and Patrick Mahomes, Aaron Rodgers and Lamar Jackson barely played.

Judge him on year two. Here’s how first-round quarterbacks drafted 2002–2021 turned out, based on their second season:

The rule. If the 2028 quarterback ranks in the bottom third in 2029, or in the bottom half in both of his first two seasons, move on. Use the 2030 first-round pick on another quarterback. A second try raises the odds of finding the right one from 20% to 36%. That’s why the 2029 and 2030 first-round picks shouldn’t be traded.

The cost of a miss. Starting over in 2030 pushes the title window back to about 2032–33.

How the numbers were run