Verdict #4 in a series of 7. Previous: Verdict #3 — Definite Optimism.
The Claim
Thiel’s power law argument has two layers:
For VCs: Venture returns follow a power law — the best investment returns more than all others combined. Therefore, VCs should stop diversifying and concentrate on their top conviction bets.
For founders: Startups should put all resources into their single most promising direction. In a power-law world, diversification is the enemy.
幂次法则 — “The Power Law”
Thiel goes further, generalizing the power law into something close to a universal principle: it dominates investing, company-building, and even life choices.
Evidence Strength: 3/5
What Thiel gets right
- VC returns genuinely follow a power law. This is well-documented across decades of data. A tiny number of investments generate the vast majority of returns. Thiel’s Founders Fund data (Facebook + Palantir dominating the fund) matches industry-wide patterns.
- His critique of VC diversification has merit. Many funds overdiversify, spreading capital so thin that even a 10x return barely moves the needle.
- The “focus” advice is directionally correct for early-stage startups. Limited resources + scattered attention = death by a thousand distractions.
The major problems
1. The descriptive-to-prescriptive leap is unjustified.
幂次是事后描述，不是事前预测 — “The power law describes outcomes after the fact; it doesn’t predict which bet will win before you place it”
Knowing that 1 in 20 investments produces 95% of returns does not tell me which 1. Telling a founder “put all resources into one thing” is only good advice if you know it’s the right thing.
2. Thiel confuses VC logic with founder logic.
A VC can invest in 20 companies; one success covers the losses. A founder typically has one company. If your single product fails, there’s no second company to save you. The power law is a blessing for VCs and a potential trap for founders.
幂次法则对VC是福音，对创业者可能是诅咒 — “The power law is good news for VCs, potentially a curse for founders”
3. Self-selected sample.
Thiel uses Founders Fund data to “prove” the power law. But these are survivor datasets. Companies that failed due to insufficient focus aren’t in the sample. Neither are companies that failed because they overfocused on one product that didn’t work.
Counterexamples
Alphabet’s “Other Bets”
Waymo, Verily, DeepMind — all are diversification plays. If Google had followed Thiel’s advice and put everything into search, it wouldn’t have Android, YouTube, or Cloud, all of which are now major growth engines.
PayPal’s Pivot
PayPal started as a cryptography company for Palm Pilots. The pivot to web payments was a strategic shift, not singular focus. Many successful companies found their real product through exploration, not fanatical focus.
SoftBank’s Vision Fund
Masa Son explicitly bet against the power law — massive capital deployed across many companies, actively managing each. Whether it ultimately works is debatable, but it shows the power law isn’t inevitable.
Net Judgment: Trust 60%
The power law is real in VC. As a universal life philosophy, it’s overextended.
I’d split my verdict: ~80% trust for the VC-level observation, ~40% for the founder-level prescription.
理解幂次法则，但别让它吓到你不做合理配置 — “Understand the power law, but don’t let it scare you out of sensible allocation”
For founders: 70–80% of resources on your core bet, 20–30% on exploration. For VCs: the power law is truth — build a portfolio that accounts for it. For everyone: the power law is background knowledge, not a decision framework.
Next in this series: Verdict #5 — The Secrets Thesis
