Of the 12 checks we run on an idea, exactly one changes when the founder is funded
We had to put a boundary on something founders argue about endlessly and nobody scores: which barriers on a celebrated idea list actually move when you have money.
Saying funded teams can do more is easy and unimplementable. So we picked one class of barrier and made it the only thing in the system that responds to the founder's situation: a regulatory or compliance cost of entry.
The part that took longest was the predicate, and I would not have guessed that going in. Capital, full time, and more than one person — all three at once. We tried looser versions and each one cleared something we did not want cleared. A well-capitalised solo founder on nights came out approved for compliance work that is mostly coordination, which is not a thing you can buy your way through alone.
The rule we settled on: keep the clearable list short and the prohibited list broad. Being too strict is recoverable — the user argues with us and we look harsh. Being too permissive passes a safety-critical idea and nobody arrives to complain. Choose the error that comes back to you.
The one that surprised me most is the exemption. Capital intensity does NOT move with funding. Money does not turn a capital-heavy dud into a business, it funds it for longer, which is worse. Leaving that check indifferent to money is the single line that stops the whole feature collapsing into rich founders get more yeses.
Net result: of the 12 checks, one moves. If you are building anything that judges ideas per-person, my honest warning is that you will want to make many more of them situational than is actually true, and each one you add teaches your system to agree with whoever has the most resources.
The full reasoning, with both lists: https://whittleos.com/guides/y-combinator-startup-ideas
I like the decision to keep the situational exceptions narrow, but I’d question whether capital + full-time + team is the right clearing condition.
Those are proxies for capacity. They do not show whether the team has the domain expertise, licensed person, approval timeline, internal controls or distribution trust needed for a specific regulated market. Money can purchase legal work and certification support; it cannot necessarily compress a regulator’s clock or replace accountable expertise.
I reckon the output should say which part of the barrier funding can clear and what remains unresolved, rather than flipping the whole check from fail to pass. Otherwise two nearly identical founders can land on opposite sides of a hard threshold even though neither has demonstrated the capability that matters.
Capacity versus capability is the right cut, and the check is a capacity test wearing a capability label.
Two things that soften it in practice: the relaxation is scoped to a named list rather than a blanket pass, and when too much of the picture comes back unknown the composite goes to hold rather than to a yes — so missing information can't quietly resolve in the founder's favour. Neither answers your actual objection. A verdict is what comes out, not a decomposition, so nothing in the output distinguishes the part of the barrier money genuinely clears from the part that needs an accountable licensed person or an approval clock that does not care how well funded you are.
The regulator's clock is the sharpest example because it's also the most checkable: approval timelines are usually public in a way that internal controls and distribution trust are not. A founder can find out that the path is fourteen months before building anything, and no amount of capital moves it. Reporting that as an unresolved component beside a cleared one is more useful than either verdict, and the two-near-identical-founders-on-opposite-sides problem is the standing cost of every hard threshold — worth paying only where the binary is doing real work.