Graveyard Archeologist

Overview

This pillar counteracts survivorship bias by actively seeking out data on failed projects, companies, and events. It provides a more realistic baseline for success by incorporating the full historical picture, not just the winners.

What It Does

The Graveyard Archeologist systematically identifies and catalogs historical failures that are typically excluded from standard datasets. It analyzes the common traits, timing, and causes of these failures. This 'graveyard' data is then used to adjust the base rate of success for a given category, correcting for overly optimistic projections.

Why It Matters

Markets often overvalue new ventures by looking only at past successes, creating a significant blind spot. By quantifying the hidden failure rate, this pillar provides a crucial, often contrarian, reality check that leads to more accurate probability assessments.

How It Works

First, the pillar defines a historical cohort for a market, such as all social media startups from the last decade. It then scours archived websites, news databases, and financial records for entities that ceased operations. Finally, it calculates a corrected success rate and analyzes failure patterns to inform predictions about the current market.

Methodology

The core calculation is the Survivorship Bias Index (SBI), which is the ratio of the failure rate in the augmented dataset versus a standard, survivor-focused dataset. Data is collected by scraping archive.org and querying news APIs with keywords like 'shut down', 'delisted', 'acquired for parts', or 'bankrupt'. Time windows are typically based on industry cycles or the average lifespan of similar past entities.

Edge & Advantage

This pillar provides an edge by systematically uncovering and pricing in negative information that the rest of the market ignores, leading to better-calibrated risk assessments.

Key Indicators

• Failed Entity Ratio

  high

  The proportion of failed projects or companies within a complete historical cohort.

• Survivorship Bias Impact

  high

  A score quantifying how much the perceived success rate is inflated by ignoring failures.

• Mean Time to Failure

  medium

  The average lifespan of failed entities in the dataset, useful for timing predictions.

Data Sources

• Archive.org (Wayback Machine)

  Provides snapshots of websites for defunct companies and projects.

• News APIs (e.g., GDELT Project)

  Used to search global news archives for announcements of shutdowns or bankruptcies.

• Crunchbase

  Offers data on tech startups, including those that have gone out of business.

• SEC EDGAR Database

  Contains delisting notices and bankruptcy filings for public companies.

Example Questions This Pillar Answers

• → Will this new AI startup be acquired for over $1 billion within 5 years?
• → Will this new cryptocurrency maintain a top 100 market cap for more than one year?
• → Will this independent film studio produce a profitable movie in the next three years?

Tags