$PPUFFINS
Heating UpSnapshot Window: 2026-08-31 11:40 UTC · ← Back to Crypto Overview
Social Momentum Summary
Total Engagement - Comments: 19, Retweets: 3, Likes: 35, Impressions: 1914
Verbatim Community Citations & Social Evidence 1 source posts analyzed
There are millions of regressions you can choose based on data up through a date in the past, like 2016. One of them will fit the data from 2017-2026 better than the others. It still does not predict what comes after 2026. We proved this back in 2022 to debunk S2F modelloooors
![]()
AI visual note: This chart, titled "BITCOIN POWER LAW: THE OUT-OF-SAMPLE TEST," displays a phase median metric comparing out-of-sample R² validation against random walks across different lookback windows (from 90 days to 7 years). It shows mostly positive predictive values (ranging from +0.478 to +0.995) for windows between 1-7 years, contrasted with a negative value of -1.181 at 90 days, supporting the post's argument that choosing regressions based on historical fit doesn't necessarily translate to genuine predictive power for future Bitcoin price movements.