Bitcoin Models Are Wrong… Here’s the Better Way

In this episode, Matthew sits down with Plan C for a deep dive into the evolution of Bitcoin power law modeling and why many commonly used assumptions may be flawed. The conversation centers on a critical distinction between calendar time vs protocol time, showing how Bitcoin’s early block production delays still distort traditional models today.

The discussion expands into quantile regression vs OLS, where Plan C explains why Bitcoin’s skewed data makes median-based modeling more reliable than averages. He also introduces a newer concept around data weighting, suggesting that current models may overemphasize recent price action due to density bias, leading to unstable projections.

One of the most impactful takeaways is how small modeling changes can lead to massive differences over time. Plan C demonstrates that long-term fair value projections could differ by as much as 50% depending on the methodology used, challenging widely accepted Bitcoin price expectations. Overall, this conversation reframes how serious analysts should approach Bitcoin’s long-term valuation.

Follow Plan C on X: @TheRealPlanC

Crypto Voices: Episode 198

Chapters:

00:00:00 - Refining Bitcoin Power Law Models

00:03:42 - Calendar Time vs Protocol Time in Bitcoin

00:06:18 - Calendar Time vs Protocol Time Regression

00:10:45 - Comparing Bitcoin Start Dates and Projections

00:12:56 - Protocol Time vs Calendar Time Analysis

00:25:44 - Protocol Time vs Calendar Time Analysis

00:35:15 - Protocol Time vs Calendar Time Analysis

00:38:55 - Quantile Regression Superior to OLS for Bitcoin

00:42:04 - Measuring Self-Similarity: Weighting and Time Scales

00:44:15 - Log-Time Weighting for Stable Power Law Regression

00:48:48 - Bitcoin's Scale Invariant Power Law Properties

00:52:34 - Comparing Regression Methods and Stability

00:56:18 - Protocol Time vs Calendar Time Differences

00:58:14 - Protocol Time vs Calendar Time Impact

01:00:44 - Decay Functions and Bitcoin Supply Modeling

01:04:19 - Building a Mathematical Model for Price Distribution

01:09:08 - Decay Functions vs Linear Quantile Regression

Hosts: Matthew Mezinskis,

Music: New Friend Music newfriendmusic.com/

Donations to Porkopolis Economics via BTCPay are appreciated: https://donations.porkopolis.io/

Show support appreciated: donations.cryptovoices.com

Podcast & videos

Bitcoin, privacy, crypto, economics & liberty

Thanks for watching!