Method
How we score a start
Two headline numbers, each from 0 to 10, called by hand rather than by formula and kept deliberately apart.
The two scores
Everyone gets two headline numbers, each from 0 to 10, kept apart so the same disadvantage never gets counted twice.
About money. Family wealth, a home to move back into, people depending on them, whether a failed attempt meant ruin or just a reset. You can check it against the D1, D2, D7, and D8 sections on each page.
Resistance that wasn't about money: bias and closed doors tied to race, gender, sexual orientation and gender identity, immigration, disability, and neurodivergence — plus the personal load some carried, like mental illness or trauma they lived through. Class is left out on purpose, since it already sits in cost of failure. Check it against the D0 baseline and the structural context, which ties the pattern to published statistics.
Both are judgment calls made by hand, not formulas. The sourced facts on each page are what they're based on, and the confidence tiers below say how solid each of those facts is. These two numbers are the softest, most subjective figures on the site, and that's on purpose.
Sourcing standard
Every claim is stored as { value, source, confidence }. The source has to be a working link, and the site won't build if a claim is missing one. Nothing shows up on a page without a source behind it.
Confidence tiers
Every field carries a stated confidence. Where sources disagree, the field is marked down rather than resolved silently.
Documented in a primary source.
Reasonably inferred from the public record.
A single unverified claim, kept out of any totals.
Sensitive fields
Some things, like childhood abuse, a mental-health crisis, family allegations, or early parenthood, are never shown, compared, or added into any total. They stay in the underlying file for completeness, and the code itself blocks them from appearing.
A blank is still an answer
Fields are never filled in just to look complete. When something isn't known, the page says “not established” instead of quietly dropping the row. A gap is information too.
Talent
Talent is scored because leaving it out would be dishonest. For the musicians, comedians, and athletes especially, raw ability was the thing that broke them out, not a side note. It's the fuzziest number here, a judgment call rather than something you can pull from a filing. It sits with the other advantages you can't just borrow: you can no more take on someone's talent than their inheritance. The one difference, and the reason it stands a little apart, is that talent is the only thing on this list a person can actually build.
What this can't tell you
These are all people who succeeded. How often the same path ends in failure, which is the number that would actually help you decide, isn't in any dataset, including this one.
A research instrument, not inspiration. Every claim carries a source.