The Data Behind the Dysfunction
The Pattern Hiding in Plain Sight
On January 5, 2026, the Las Vegas Raiders fired Pete Carroll after a 3-14 season, one of the worst records the franchise has recorded since the AFL-NFL merger. Carroll, now 74, had been hired eleven months earlier alongside first-year GM John Spytek. The two had never worked together. They had no shared history to fall back on when things went wrong and things went wrong immediately.
Both men accepted the job. The argument is whether they truly chose the other. This isn’t about blame. It’s about structure.
That distinction matters. When a GM and head coach are paired by ownership rather than selecting each other, both parties retain an escape hatch when results disappoint. “I agreed to work with him, but I didn’t choose him.” The Raiders’ collapse was not an outlier. It was the latest data point in a pattern that has repeated itself across the NFL for two decades. And the data now allows us to see that pattern with uncomfortable clarity.
In Part 1 of this series, we synthesized conversations with more than 50 NFL industry professionals about what the GM role requires. The consensus was striking: the industry agrees on what makes GMs succeed but struggles to identify people who can do it. Respondents repeatedly cited “The Marriage Problem,” the relationship between GM and head coach, as the single most common predictor of failure. One put it simply: “The Head Coach and GM not being on the same page is a huge point of discrepancy.”
Part 2 asks a different question: What does the data say? Using a database of 175 GM-HC pairings from 2002-2025, ownership behavior analysis across 48 ownership eras, and documentation of 65+ NFL GM search processes, we can now quantify what the industry has long suspected. The NFL’s GM hiring process produces predictable failures because ownership groups repeat structural mistakes the data clearly identifies.
The findings should change how teams design their searches and how candidates evaluate which opportunities to pursue.
The Marriage Problem: What the Data Actually Shows
The industry talks about GM-HC alignment as if it were a soft skill, a matter of personality fit. The data tells a harder story. How a GM-HC pairing is formed, the governance structure at its origin, predicts tenure, performance, and playoff success more reliably than the individuals’ resumes.
We categorized 175 GM-HC pairings into five governance structures based on how the pairing originated:
- Hired together: GM and HC brought in simultaneously, through a coordinated search where both parties chose each other
- GM hired HC: GM in place first, selects the head coach
- HC in place first: New GM inherits an existing head coach
- HC hired GM: Head coach has authority over GM selection or significant influence
- Pairings without mutual selection: Both parties accepted the job, but neither selected the other. Often ownership installed the combination. Mutual selection requires genuine decision-rights, not simply being consulted. Where ownership retained decisive authority over both hires and either party lacked final say on the other’s hiring, the pairing was coded as lacking mutual selection.
Classification was based on documented hiring sequences, public statements about decision-making authority, and reporting structures at the time of hire. Borderline cases (such as pairings where both parties were hired within weeks but through separate processes) were categorized based on whether evidence suggested genuine mutual selection or parallel owner-driven decisions.
The results, drawn from 150 ended pairings, reveal a hierarchy of outcomes that should inform every search process:
| Category | # Pairings | Avg Years | 3+ Years | <2 Years |
|---|---|---|---|---|
| Hired together | 13 | 4.89 | 76.9% | 15.4% |
| GM hired HC | 82 | 3.49 | 48.8% | 30.5% |
| HC in place first | 23 | 2.21 | 21.7% | 52.2% |
| Without mutual selection | 10 | 1.62 | 10.0% | 70.0% |
| HC hired GM | 9 | 1.78 | 22.2% | 66.7% |
The spread is not subtle. Pairings formed synchronously, where GM and HC were hired together and chose each other, last nearly three times longer and produce playoff appearances nearly four times higher than pairings without mutual selection. More than three-quarters of synchronous hires reach the three-year threshold that typically indicates organizational stability. Pairings without mutual selection collapse within two years 70% of the time and produced zero playoff appearances. The sample is small enough that a single successful outlier would alter the percentage, but the consistency of the pattern across two decades is notable.
What explains the gap? Part 1’s respondents offered a clue: elite operators invest in the relationship before problems arise, building trust through productive disagreement rather than avoiding conflict. When GM and HC enter together, they share accountability from day one. Neither can blame the other. The structure itself creates alignment. Pairings without mutual selection invert this. Each arrives with a separate mandate and an escape hatch when results disappoint.
Not every candidate has leverage to pick their spot. Some take a job just to have one. But accepting a job is not the same as choosing your partner, and that distinction creates the mechanism for failure: you may have agreed to work with someone, but you didn’t choose to work with them.
When authority is ambiguous, decision-rights are split, or partners did not choose each other, leaders spend political energy defending lanes rather than solving problems. They stop telling each other the truth. Decisions slow. People protect themselves instead of the team. And the same people, in a coherent structure, would look far more competent.
This research is not a critique of individuals. It is a critique of the systems we place those individuals into.
The Successes
The successes in the “Hired together” category are not accidents. They are the predictable outcome of a structure that creates shared accountability and mutual selection: Schneider/Carroll in Seattle (14 years, Super Bowl champions), Beane/McDermott in Buffalo (8.6 years, perennial contender), Lynch/Shanahan in San Francisco (8.9 years, three Super Bowl appearances), and Holmes/Campbell in Detroit (4.9 years, back-to-back NFC North titles).
The Failures
The failures in pairings without mutual selection are equally consistent. The Raiders alone have produced three such pairings since 2018, each ending within a year. If you put good people in a bad structure, the structure wins.
The Prior Relationship Paradox
A common assumption is that strong prior relationships should predict success. It makes intuitive sense: people who already know and trust each other should collaborate more effectively once the pressure hits. If “The Marriage Problem” is about alignment, shouldn’t shared history predict success? The data complicates this assumption.
We scored each pairing on Prior Relationship depth using a 0-4 scale incorporating overlap years, multi-organization overlap, coaching tree connections, and reunion flags. The results were surprising:
| Relationship Level | # Pairings | Avg Years | 3+ Years | <2 Years |
|---|---|---|---|---|
| None (Score 0) | 93 | 3.25 | 40.9% | 40.9% |
| Low (0.25–1.5) | 7 | 3.21 | 57.1% | 14.3% |
| Moderate (2.0–2.5) | 17 | 3.60 | 41.2% | 41.2% |
| Strong (3.0+) | 30 | 2.98 | 40.0% | 30.0% |
Prior relationship alone does not predict longevity. Pairings with no prior relationship perform roughly as well as those with strong prior connections. The combination matters more than either factor alone. When we isolated “Hired together” pairings that also had prior relationships, the success rate jumped: 71.4% lasted three or more years, averaging 4.39 years together. The relationship provides a foundation; the synchronous hiring structure provides accountability. Together, they create conditions for success that neither achieves independently.
Ownership groups sometimes believe that hiring a GM and coach with shared history will replicate the benefits of synchronous hiring, that the relationship substitutes for the structure. It does not. A pairing without mutual selection between two people who know each other still fails at elevated rates because the accountability structure remains broken. Knowing someone is not the same as choosing them.
Owner Behavior as Predictor
If governance structure predicts pairing outcomes, what predicts governance structure? The answer, consistently, is ownership behavior.
We analyzed 48 ownership eras across all 32 franchises, tracking behavioral indicators including churn rate (measures combined GM and HC departures per decade), delegation patterns, confidence integrity (whether owners follow through on public commitments), and specific pathologies like asymmetric accountability and repeat pairings without mutual selection. These scores rely on publicly documented actions; private behavior may differ.
High-Stability Owners
High-stability owners (churn of two or fewer per decade, high delegation, high confidence integrity) include the Rooney family in Pittsburgh (0.8 churn rate), Benson family in New Orleans (0.9 churn rate), Allen Trust in Seattle (1.1 churn rate), and Bisciotti in Baltimore (1.4 churn rate). These ownership groups share common traits: they delegate genuine authority to football operations, they honor public commitments to their personnel, and more often, they resist the temptation to intervene when results disappoint.
High-Churn Owners
High-churn owners (churn exceeding five per decade, low delegation) tell a different story. The behavioral red flags that cluster together, including asymmetric accountability, repeat pairings without mutual selection, and low delegation, appear consistently in the same ownership groups. They may be more likely to bring in external consultants and consultants don’t fix broken structures. The recurrence of these indicators across the same ownership groups suggests this is not bad luck. It is a pattern.
The implication for GM candidates is direct: ownership behavior is knowable before you take the job. The patterns are public. Candidates who enter high-dysfunction environments believing they will be the exception are making a predictable error. Structural misalignment compounds costs. Cap inefficiency, staff churn, and roster resets follow predictably from governance failures.
The Search Process Problem
Part 1 identified flawed hiring processes as a recurring theme. The search data substantiates these concerns and reveals a counterintuitive finding about consultant involvement.
Professional search firms have been involved in some of the NFL’s most successful placements: Carroll/Schneider in Seattle, Reid in Kansas City, Arians in Tampa Bay. They have also been part of some of the least successful ones. The search firm does not predict the outcome. The governance structure does.
When we tracked placements by the same consultants across different governance structures, the pattern held. Candidates placed into “Hired together” structures succeeded at that category’s base rate. Candidates placed into “HC in place first” or pairings without mutual selection failed at those base rates. The consultant was constant; the structure varied; outcomes followed the structure.
Search firms serve multiple functions, and not all align with finding the best candidate. In some searches, consultants broaden the pool. In others, they validate a decision already made. The search database documents multiple instances where the eventual hire was the “prohibitive favorite” before interviews began. These searches produce defensible hires. They do not necessarily produce good ones.
The question is not “Did we find the right person?” but “Did we create a structure where the right person can succeed?” and, perhaps more critically, “Was this ever a real search?”
Implications for Current Openings
Two GM positions are open as of this publication: Miami and Atlanta. The data suggests they face very different challenges.
Miami Dolphins: The “HC in Place” Trap
The Dolphins fired Chris Grier on October 31, 2025 after nine years, but retained head coach Mike McDaniel. Their initial interview list draws from the 49ers, Packers, Eagles, and Rams, with several candidates connected to the Shanahan/McVay coaching trees.
But the structure matters more than the candidates. McDaniel remains in place, meaning any new GM enters an “HC in place first” governance category. The historical data on this structure is sobering: only 21.7% of such pairings last three or more years, while 52.2% end within two.
The data becomes starker when you isolate true stranger pairings within this category. When a GM inherits a head coach with whom they have no prior relationship, 80% of those pairings end within two years. Only two of fifteen such pairings (13%) in our database lasted three or more years, and both involved Hall of Fame-caliber coaches already established before the GM arrived.
Atlanta Falcons: The Structural Experiment
Atlanta’s situation is more novel. Arthur Blank fired both GM Terry Fontenot and head coach Raheem Morris on January 4, 2026, and announced a significant restructuring: the creation of a new “President of Football” position that will sit above both the GM and HC, with final decision-making authority.
Blank has retained Sportsology Group for the GM search and ZRG Partners for the HC search, the first documented use of two separate firms for coordinated football operations hires. The structure mirrors what Part 1 respondents called the “three-headed monster” model, which has emerged in various forms across the league.
The closest analogue is the synchronous “Hired together” model. If the President of Football genuinely integrates the searches, hiring people who complement each other and share a vision, Atlanta could replicate the conditions that produce 76.9% success rates.
Two variables introduce uncertainty.
First, Matt Ryan is expected to be named President of Football Operations. Ryan’s credentials are as a player, not an executive. The transition from franchise icon to front office leader has worked in some cases and failed in others; the variable is often how quickly the former player recognizes what he doesn’t know. Inexperienced executives often rely heavily on the consultants who guide them through searches. Second, Sportsology has former NFL executives on staff with prior connections to the Falcons organization. Blank’s historical behavior is encouraging: 2.6 churn rate, delegation score of 3.0, high confidence integrity. The structure is a gamble, but the owner behind it has earned benefit of the doubt.
Complexity Is Real. The Pattern Is Too.
Structure is not the only factor in GM-HC outcomes and does not guarantee success. Talent, judgment, roster quality, injury luck, and resource support all matter. And every category has exceptions, unstable structures that held together for a time, and stable ones that still failed.
But when we step back across twenty years of data, one pattern repeats with uncomfortable regularity: when the structure removes shared accountability at the origin of the relationship, the failure rate spikes, even for highly capable people. In an industry where very few leaders get a second chance, that distinction is not academic. It is existential.
Think of structure the way we think about turnover differential. It doesn’t determine every outcome. But across seasons and teams, the slope of the line is unmistakable.
Structure sets the baseline; execution determines variance. This research should be read as risk management, not prophecy.
Tailored Implications
For Ownership and Search Committees
The data points toward a consistent recommendation: run synchronous searches whenever possible. If you are hiring a GM, consider whether your current HC should be part of the evaluation, or whether a clean slate would produce better alignment. The short-term cost of coach turnover is real, but the long-term cost of structural misalignment is higher.
Avoid pairings without mutual selection. The pattern is remarkably consistent: installing a GM and HC who accept the job but do not choose each other produces a 70% failure rate within two years and zero playoff appearances across our sample. The sample size is limited, but the consistency is difficult to dismiss. Almost nobody overcomes that structure.
And if you hire a search firm, be honest about why. Consultants may be able to identify talent you would not otherwise access, or they can validate decisions you have already made. They rarely do both simultaneously. If the outcome is predetermined, own it. Performative searches waste everyone’s time and create false expectations.
For GM Candidates
Evaluate the structure, not just the opportunity. An open GM job is not automatically worth pursuing, even in light of the fact that your ‘stock’ is not completely in your control. You may be telling yourself you need to make the bad marriage work or you’ll never get your shot, but the governance category you are entering predicts your odds of success more reliably than your own resume.
Questions to ask: Is the head coach in place, and if so, what is the real power dynamic? What is the owner’s behavioral history: delegation, churn rate, confidence integrity? Has this ownership group installed pairings without mutual selection before? What happened?
The industry’s low rate of second-chance GM hires means your first opportunity may be your only one. Entering a high-dysfunction environment because it is available is a predictable path to a short tenure. Agreeing to work with someone is not the same as choosing to work with them. If you do not have genuine input into who your partner will be, or if your partner did not have genuine input into selecting you, the escape hatch exists for both of you. And one of you will use it.
For Scouts and Executives
The data confirms what Part 1 respondents observed: where you work shapes what you learn. Organizations with high-stability ownership develop executives differently than high-churn environments. The 49ers, Rams, Eagles, and Ravens have become GM pipelines not simply because they hire better people, but because they create better conditions where people can develop.
As you evaluate career moves, consider not just title progression but organizational stability. A lateral move to a high-functioning organization may be more valuable than a promotion in a dysfunctional one.
Structure Over Individuals
Part 1 concluded that the industry agrees on what the GM role requires but struggles to identify people who can do it. The data suggests that structure problems are just as impactful.
Even if a team identifies qualified GM candidates, if they repeat the pattern of placing qualified people into structures, they are designed to fail. Pairings without mutual selection produce predictable outcomes. Coach-first hierarchies produce GM turnover. High-churn ownership produces high-churn results.
The culture of NFL leadership rewards individual exceptionalism. “I can make it work” is the default posture. This piece challenges that mythology directly, and some will resist the implication that structure constrains what talent can accomplish.
The honest reality: those who need to read this most probably won’t. High-churn ownership groups trust their own judgment; that’s part of why they’re high-churn. If you have lived the pattern (and read this far), you already agree.
The value is in the middle. Candidates who might have taken a bad job instead give themselves permission to decline. Owners on the margin might pause before installing a pairing without mutual selection. Search consultants might push back on predetermined outcomes.
If this research changes one decision in one search cycle, it will be a step in the right direction. And over time, one decision at a time, that shifts the industry. Change in insular cultures doesn’t happen through sudden adoption. It happens through gradual absorption, until the framework becomes common sense that everyone claims they always understood.
The candidates entering Miami and Atlanta searches could be talented executives from successful organizations, but the question is whether the structures they enter will allow that capability to translate into results.
The data tells us that question may matter more than the names on the interview list.
Part 3 will profile specific candidates and share polling from NFL scouts on who should be strongly considered for these openings.