Benchmark library
What a realistic win rate looks like, by category
Each category carries a Beta(α, β) prior over the win rate, where α and β are pseudo-counts out of 30 notional observations. Your own inputs are compared against this baseline, and the report says plainly whether you are assuming you will do better than the category typically does. Ranking still uses score = p × b.
Side business
Beta(α=2, β=28) · Source: Stripe-verified indie project revenue distribution (US-IND-01..04) + fitted Patreon earnings distribution (US-CRE-01..04)
Both underlying datasets are community analyses of scraped platform data, not surveys. MicroConf's widely cited '70% under $1k MRR' is not MicroConf's figure at all (US-IND-05) and is excluded.
Mean win rate
6.7%
Payoff P50
2.50×
Payoff P90
13.00×
Education
Beta(α=17, β=13) · Source: FREOPP lifetime return by programme (US-EDU-12..17) + BLS CPS earnings by attainment (US-EDU-06..09)
FREOPP measures net lifetime return per programme, not per student. A 57% positive share is not a 57% chance for you at a specific programme, and the earnings premium behind it is a cross-sectional association rather than a causal return.
Mean win rate
56.7%
Payoff P50
1.50×
Payoff P90
4.70×
Career
Beta(α=16, β=14) · Source: Pew job-switcher real wage gains (US-CAR-05..06) + Atlanta Fed Wage Growth Tracker (US-CAR-01..02) + Correlation Ventures financing outcomes (US-CAR-16..17)
Pew's 60% was measured over the 2021-22 record-quits window, so it is an optimistic anchor for a normal labour market: the quits rate has since fallen to 1.9% (US-CAR-07).
Mean win rate
53.3%
Payoff P50
3.00×
Payoff P90
6.00×
Property
Beta(α=19, β=11) · Source: Case-Shiller national index, real 10-year windows (US-RE-10..15) + NAR buyer profile (US-RE-01..05)
Real, not nominal. On nominal appreciation the same ten-year windows read above 90% positive, which is most of why housing feels like a sure thing. Choice of window matters too: the trailing decade annualises at 6.4% nominal against 4.3% since 1987 (US-RE-10, US-RE-12).
Mean win rate
63.3%
Payoff P50
1.25×
Payoff P90
2.50×
Small business
Beta(α=8, β=22) · Source: BLS Business Employment Dynamics survival (US-SMB-01..08) × Fed Small Business Credit Survey profitability (US-SMB-10..11)
BLS survival counts continued employment, not profit — a business losing money every month is counted as surviving (US-SMB-09). The prior therefore multiplies two-year survival by the share of sector firms actually operating at a profit, rather than using the 85% headline. Franchising buys no measurable edge here: the Census study finds the advantage vanishes conditional on first-year survival (US-SMB-17) and franchisees default on loans more often than independents (US-SMB-18).
Mean win rate
26.7%
Payoff P50
1.40×
Payoff P90
3.00×
Investing
Beta(α=23, β=7) · Source: S&P 500 total-return history 1928-2025 via Damodaran (US-INV-01..11) + ICI and Vanguard expense ratios (US-INV-13..15)
Real, not nominal, and geometric, not arithmetic. The arithmetic mean sits 1.84 points higher (US-INV-03) and is not what an investor compounds. Over five years, 88.3% of windows were positive nominally but only 76.6% in real terms (US-INV-05, US-INV-06).
Mean win rate
76.7%
Payoff P50
2.83×
Payoff P90
5.80×
Agriculture
Beta(α=13, β=17) · Source: USDA ERS Farm Household Income Estimates (US-AGR-02..07)
At the median, a US farm household is forecast to lose $1,161 a year on farming itself and stays solvent on off-farm income (US-AGR-02, US-AGR-03). The prior uses the 42% of intermediate farms with positive farm income; commercial farms above $350k of gross cash income reach 84% (US-AGR-06), so scale changes this answer more than anything else.
Mean win rate
43.3%
Payoff P50
1.20×
Payoff P90
2.50×
Other
Beta(α=14, β=16) · Source: Median across the seven calibrated categories above
No published base rate applies, so this is deliberately weak. It exists to be replaced by your own logged outcomes rather than to be believed.
Mean win rate
46.7%
Payoff P50
1.50×
Payoff P90
4.70×