Prior seasons and winners

Investigated 2026-09-10. Full detail in data/history/ffpc_history_sources.md; parsed results in data/history/<season>/results.json; champions in data/history/champions.json.

Availability (short version)

• Prior-season draft boards: not obtainable. The draft-room API only answers for the current season (every 2015-2025 league ID returns an empty board), the Wayback Machine never captured draft boards, and the one archive that has every FFPC board since 2016 (fantasymojo.com's FFPC Data Warehouse, including "winners deep dives") is members-only.
• Championship-round leaderboards: partly. Final tables exist for 2015, 2017, 2018 and 2025 (679 rows), mid-round snapshots for 2016, 2020, 2022; nothing for 2019, 2021, 2023, 2024. Regular-season standings exist for 2017, 2018, 2024 (partial), 2025.
• Winning rosters: four seasons. FFPC's own blog posts give the champion's final-week lineup for 2017, 2020, 2021 and 2022. Everything else is images with usernames only.
• So a true ten-year roster study needs a fantasymojo membership (about the cost of one pizza; their winners deep-dives are exactly this). Without it, the strongest evidence is the leaderboards, and they are unambiguous.

What the leaderboards say (every season with data agrees)

• The $1M is a three-week variance sprint. Correlation between the three-week sum and final total is 0.95-0.99 every year; the regular-season carry-over explains 1-9% of the variance. Your PPG gets you in the door and buys a few points; it does not decide anything.
• Winners were about three standard deviations above the qualifier field over weeks 15-17, every single year. That is not a "solid team" outcome; it is a ceiling outcome. Every champion had at least one week of 207+ (2025: 220 and 212 back to back; 2017: 223; 2018: 226; 2022: 223), and most top-10 finishers had a 200+ week.
• Top-10 finishers come from everywhere in the regular-season distribution. Their regular-season percentile ranks among qualifiers in 2025 were 94, 75, 86, 78, 57, 24, 51, 61, 80, 94; in 2015 they included teams at the 7th and 12th percentiles. The 2024 champion was 10th in regular-season points overall; the 2020 champion came from far down the board with a 256.85 final week and won by 0.45.
• The roster you carry into week 15 is the one you finish with. No free agents in weeks 15-17; the 2022 winner (Mahomes + Kelce stack from slot 12) won on in-season waiver adds made before the cutoff.

Documented champion lineups (final week)

• 2022 (slot 12, $1M): Mahomes, Kelce, Lamb, DeVonta Smith, McKinnon, Stevenson, Fournette, Zay Jones, Tucker, PHI D. QB+TE stack, three waiver-wire RBs.
• 2021 ($500K): Burrow, Chase (55.6 that week), Kupp, Deebo, Tyreek, Jonathan Taylor, Javonte, Andrews, Bass, TB D. QB+WR1 stack.
• 2020 ($500K): Herbert, Kamara (56.2), Gaskin, Fournette, Diggs, Evans, Hopkins, Waller, Gay, KC D.
• 2017 ($250K): Prescott, Le'Veon Bell, Gurley (49.6), Kamara, Crowder, Golladay, Hopkins, Ben Watson, Tucker, LAC D.

Commonalities in those four: a QB stacked with at least one of his own pass catchers in three of four; one 50-point outlier game in the deciding week in three of four; elite RBs that were mid-round or waiver acquisitions rather than round-1 picks in most cases; a kicker and defense that mattered (Tucker twice).

What to take into tomorrow's draft

1. Draft weeks 1-12 for the qualification (top-2 seed or top-2% overall), then treat everything about weeks 15-17 as ceiling: stacks, dome/warm-weather week 15-17 slates, and players with a real 40-point-week outcome.
2. Do not pay for floor after round 8. A "safe" bench does nothing in the three-week sprint; a bench of contingent RBs and stack pieces does.
3. Keep the roster in a state where your best week-15-17 lineup can still be assembled on the week 14 waiver run; there are no moves after that.