Season records
Every completed Colts season from 1998 through 2025, including regular-season and postseason outcomes.
REPORT 001 · PUBLIC RESEARCH AUDIT
Methodology without giving away the vault.
Colts Era Lab is backed by a private 28-tab research workbook. This page publishes the scope, record counts, formulas, source families, scoring rules and limitations needed to audit the conclusion without exposing the full organized research database.
28research tabs
401core structured evidence records
28completed seasons, 1998–2025
300,000random-weight robustness simulations
Evidence inventory
What was actually researched.
The 401-record count is intentionally conservative. It counts the principal row-level evidence registers once and excludes dashboards, repeated summaries, derived era tables, charts and narrative conclusions.
QB decisions
Major acquisition and succession choices graded by cost, intended role, durability and outcome.
Draft classes
Every class from 1998–2025; 25 mature classes scored and 2023–2025 withheld from final class grades.
Notable draft picks
Representative hits, misses, contributors and developing selections with year-level source links.
Major resource moves
Trades, free agents, extensions and premium resource decisions; the unresolved Sauce Gardner trade remains unscored.
Coaching records
28 completed coaching-season records plus 14 major coaching and governance decisions.
Elite player-season events
AP first-team All-Pro and Pro Bowl player-seasons used to measure peak talent, origin and durability.
Adversity cases
Major injury and availability shock seasons evaluated for preparedness, replacement execution and result preservation.
Research phases
Results, quarterback, drafting, resources, coaching, player peaks, resilience, then final scoring and stress testing.
Published scoring model
Seven dimensions. Results matter most.
The final score is a weighted organizational model, not a win percentage and not an opinion poll. The underlying workbook keeps full-precision values; the public report displays one decimal place.
| Dimension | Construction | Weight |
|---|---|---|
| Team results | 70% win percentage + 30% winning-season rate | 25% |
| Postseason achievement | Playoff, division-title, deep-run, Super Bowl trip and championship rates | 20% |
| Quarterback management | Continuity, intended-starter share and multi-year succession stability | 15% |
| Draft and development | 65% mature draft index + 35% player peak/homegrown-core index | 15% |
| Roster-resource strategy | Move quality, positive rate, failure avoidance and premium-resource hit rate | 10% |
| Coaching ecosystem | 40% continuity + 60% coaching-phase quality | 10% |
| Injury resilience and depth | Preparedness, replacement execution, result preservation and adjustment | 5% |
Published result
- Bill Polian — 70.5
- Ryan Grigson — 61.1
- Chris Ballard — 47.8
Stress testing
- 8 weighting philosophies
- 7 leave-one-dimension-out tests
- 200,000 unrestricted random-weight draws
- 100,000 bounded 5%–35% draws
Robustness result
Polian won 97.7% of unrestricted random-weight draws, 100% of bounded draws and all seven leave-one-dimension-out tests.
Counterexamples retained
Ballard wins the deliberately narrow drafting-plus-transactions stress test. Grigson wins the roster-plus-resilience-only stress test. The system can produce another winner when the philosophy genuinely changes.
Source policy
Primary facts first. Judgment stays visible.
Leadership transitions and current organizational facts are anchored to official Colts material. Historical season, player and draft evidence is cross-checked against established statistical databases. Every scored register keeps a row-level source field in the private workbook.
Indianapolis Colts
Official transaction releases, coaching announcements, historical retrospectives, roster records and injury/season reporting.
colts.com ↗Pro Football Reference
Year-by-year team records, drafts, coaching/staff context and player honor reference tables.
Colts historical index ↗nflverse
Master draft-pick data used as a second structured reference for selection history.
nflverse data ↗Secondary record checks
Reputable historical record tables are used where helpful, while official or primary sources control organizational facts whenever available.
Private workbook policy
The complete workbook is intentionally not published as an unrestricted public file. Public transparency comes from publishing the methodology, formulas, evidence inventory, source families, exclusions and reproducible model behavior while protecting the organized research asset itself.
Known limitations
What the model does not pretend to know.
Public information only
No private scouting boards, medical files, ownership discussions, contract negotiations or internal coaching evaluations are available.
Recent classes withheld
The 2023–2025 draft classes are tracked but excluded from mature class scoring until enough NFL seasons exist.
Unresolved future capital
The Sauce Gardner trade is recorded but excluded from the resource index until the on-field return and future first-round opportunity cost can be judged responsibly.
2026 remains live
The final historical verdict is locked to completed seasons through 2025. Active 2026 results do not enter the published score.
Judgment is not hidden
Some draft, transaction, coaching and resilience components use evidence-coded judgment. Component rules remain explicit so readers can challenge them.
Robustness is not truth probability
The 300,000 random-weight runs show whether the ranking is sensitive to reasonable priorities; they are not a probability that one executive is objectively “correct.”