A unique object
has no bid
This page was a market terminal. It held a wallet, priced a ticket and ran a settlement clock. The room it spoke to was designed, written, and never deployed, so what is left is the question it was built to answer, and an honest account of where we stopped.
The problem
Most days, the honest bid is nothing
A floor price is the cheapest ask in a collection. It says what someone wants for a piece, not what anyone will pay for yours. For a one-of-one there is no other side of the trade standing by, because there is nothing identical to quote against.
The art world answered this a century ago by paying a person to carry the risk. A dealer takes the work on consignment, holds it for months, and finds a buyer for twenty to fifty percent. That works. It also does not answer on a Sunday, and it does not scale past one address book.
Tokenised art has the same problem and a faster clock. Every attempt so far swaps it for a different one.
Lend against it
Borrow against the piece and keep it. You keep the price risk too, and every loan in the book is written against the same floor. In August 2022 BendDAO ran near-empty and rewrote its own liquidation rules mid-crisis.
Break it into shares
Fractionalising an object makes it tradeable by dropping the thing that made it worth owning. The vault trades as an index of the floor, the one-of-one premium goes, and in most jurisdictions you have just issued a security.
Bid on the collection
Pool AMMs and collection-wide offers put real, funded money on the bid side. The bid is for any piece, so it gets filled with the worst one every time. The market maker eats that adverse selection until they quit.
All three are real and useful. None of them leaves a named piece with a funded bid that survives a bad month.
What we designed
Backing pays someone to keep a bid open on one specific piece.
A depositor escrows the work with ETH behind it and names their own number. That number does three jobs: it funds a bid the winner can always take, it sets how often the piece is drawn (thin backing often, thick backing rarely), and it earns the depositor a share of every ticket the room sells. Holding a standing bid stops being a favour and becomes a paid position.
A purchaser buys a draw. Win, and a piece is yours to keep or to sell straight back into its standing bid. A ticket was priced at the expected backing of the piece you would draw, plus a surcharge, so a purchaser who always sold back would recover 85.71% of what they staked: the discount over one plus the surcharge, an identity anyone can check. The rest paid the depositors.
Whether ticket fees ever cover that position is the open question. We did not get to find out.
What we tried, and what it taught us
The room had a flaw anyone could see. The money you promised also set your odds, so anything genuinely good was almost never drawn, and anything drawn often was being given away. The top of the room was paid to be unreachable.
We tried to fix it by setting the odds from the gap between what a work is worth and what its owner had promised, rather than from the promise on its own. A grail and a cheap print would then come up equally often, and the gap would be the honest thing you were declaring.
It pays you to lie. Say your work is worth less than it is and you save about a ninth of its value every time it sells, while nothing else about your position gets worse: the same odds, the same income, less money tied up. Not sometimes, and not at certain prices. Always, by the same amount.
A room where understating your own work is the winning move is not a room a gallery should open. So we did not, and that is worth more than the room would have been.
What survived
The oldest arrangement in the trade, with the money actually there.
You leave the work with us with a price on it. Someone pays that price, or nobody does and you take it back. If a buyer wants out, they hand it back for the money you put up, which sits untouched the whole time and is never ours to spend. No draw, no odds, no luck. We keep 4.76% of what a buyer pays.
And the limit, so you can hold us to it: this beats the room we published a critique of only while your asking price stays within about 23% of your buy-back price. Price it far above what you would pay to own it again and a buyer is worse protected than in the thing we criticised, not better.
Which is why that crossing point is on this page and the flattering number is not. The flattering number is only reachable where you make nothing.
The questions, and where they stand
Three of the five are answered now, one of them by finding out we were wrong. The two left are the two that matter.
- Whether it could be offered at all.answered
- This is the one that stopped it. Paid entry plus chance plus a prize is a lottery in most legal frameworks, and a verifiable random draw does not change that. We did not find a way around it that we believed, so the room was never deployed.
- That the room could never hold a good piece.answered
- Promise enough money to be safe and your work is almost never drawn. Make it drawn often and you are giving it away. We tried to fix that by setting the odds from the gap between what a work is worth and what its owner had promised, and the fix pays you to understate. That is the answer, and it is why there is no draw.
- That promising more money did not pay better.answered
- However deep an owner went, the same money came back per day: the work came up rarer at exactly the rate each loss grew larger. Answered by removing the draw. There is no rate to trade against any more, only a price and a buyer.
- Whether money you put behind your own work counts as a bid.
- Unchanged, and the reason to stay careful. Escrowing it makes the number expensive, which is not nothing. It does not make it true, because nobody ever has to disagree with it for it to stand.
- Whether anyone comes to a room with no game in it.
- New, and the whole bet. The draw was the thing that brought people through the door, and a price tag is not. Better to write that down as a bet than dress it up as a finding.
So 生涯 ships on ordinary audited contracts and nothing more exotic, and the room stays a published design rather than a product. Every figure above is a division anyone can redo, which is the same arithmetic we applied to somebody else's pool at the pulse page. It still holds. It simply never got a chance to be wrong about ours, and the part of it we did test turned out to be wrong in a way worth publishing. The reasoning has been settled since 2026-07-26.