Fantasy Football Case Studies: Code, Data and Results

Code, league samples and result files for eight pre-registered studies of public fantasy football leagues (four dynasty, four redraft), published by Dynasty Dealer. The archive lets anyone check or rerun every number: the analysis code frozen before any results were seen, an independent re-implementation in a second language, the seeded league samples, and every headline and per-unit result file. Write-ups: Fantasy Football Case Studies (Dynasty Dealer) Dynasty studies (registered plan: 10.17605/OSF.IO/F8K59) Does trading win championships? How long does a dynasty rebuild take? How dynasty championships are built The superflex QB shortage Redraft studies (registered plan: 10.17605/OSF.IO/XZ7NB) Does draft position matter? Zero RB vs RB-heavy drafts Playoff odds after an 0-3 start How redraft championships are built Data source: public leagues on Sleeper, read through Sleeper's public read-only API. None of Dynasty Dealer's own data (values, trade database, users) is used. Each study ran on the 2025 season and was repeated on 2024 with the same frozen code; seven of the eight results replicated, and the eighth (redraft Study A) did not, which its page reports. Raw API responses are not included because they contain managers' display names; the included fetch scripts download them again from the league IDs. Reproduction check: rerunning this archive on the original downloads reproduced every 2025 result file and per-unit file byte for byte. Start with README.md. Licence: code MIT; data, results and documentation CC BY-NC 4.0 (non-commercial, in line with Sleeper's API terms). See LICENSE.md.