How the floor works

Open Outcry is an hourly market-call tournament on Robinhood Chain. Six times a session you seal a call on a tokenized equity — direction plus how sure you are — and sixty minutes later the tape decides whether you were right. Right pays. Wrong burns. The record is permanent and it sells with your seat.

Two things make it different from the collections that came before it. The prize pool is funded by the players who called it wrong rather than by a fee on our own token, so it does not go quiet when speculation does. And seats cannot be bought — the top five hundred of each season win one.

The four pillars

• ›Zero-sum by construction. Nothing is minted to pay anybody. Every unit a winner receives was forfeited by somebody who was wrong in the same round.
• ›Seats are ranked into existence, never sold. Money cannot skip the proof.
• ›The floor under each seat is held in WETH, so it cannot fall alongside our own token.
• ›Art and metadata are returned by the contract. No gateway, no image folder, nothing to go stale.

Bells fire on the half hour across the US cash session, 09:30 to 14:30 ET. Each publishes a rotating slate of eight tokenized equities drawn from the deepest pools on the chain, and the slate for the next bell is committed one bell early so it can never be edited after the crowd is visible.

One round, four moments

• ›The bell — the round opens and a fifteen-minute submission window starts.
• ›Seal it — instrument, direction, confidence and stake. Only a hash lands on-chain.
• ›The lock — every commitment is revealed in the same block, so the crowd becomes visible at the exact moment it stops being changeable.
• ›Settlement — sixty minutes later a five-minute average decides it, and the call is written to the record forever.

A public chain leaks everything written to it. If calls were stored in the clear, the last person to submit could read the whole floor and take the contrarian side purely to farm the originality multiplier. That single attack would destroy the scoring rule, so the design closes it rather than patching around it.

• ›Commit — a hash of your call lands on-chain during the window, and your stake is escrowed in the same action.
• ›Reveal — after the lock, every preimage is submitted in one batch so nobody positions against a crowd they can already see.
• ›You sign once. The reveal costs you nothing; its network fee is paid out of the rake.
• ›If the batch never arrives, a public ten-minute window opens and anyone can push their own preimage. The relayer is a convenience, never a dependency.

Calls are not simply right or wrong. They are graded on how sure you claimed to be, using a logarithmic rule with one useful property: your expected score is highest when you report exactly what you actually believe. Overstating certainty is punished harder than understating it is rewarded, so bluffing is not a bold strategy — it is arithmetically the worst one available.

score if right

1 + log2(p)

score if wrong

1 + log2(1 − p)

confidence range

50% to 90%

best possible

+0.85

worst possible

−2.32

Confidence is capped at ninety percent deliberately, and no clamp is applied to the loss side. A clamp would make overclaiming free above the cap and break the honesty property the rule exists for.

Originality

The other half of the grade is how few people took your side. Your slice of the pot is multiplied by (1 − c) raised to the power 1.5, where c is the share of staked capital on your side. Being right in a crowd pays badly; being right alone pays extraordinarily. That is a real property of markets, and it means the profitable seat is never obvious.

Losers forfeit a slice of stake proportional to how confident they were: wrong at sixty percent costs about fourteen percent of stake, wrong at ninety percent costs all of it. The forfeited pool is the only prize money that exists.

Where the rake goes

The rake is five percent of the forfeited pool and it is the only thing the protocol takes. Winners split the remaining ninety-five percent by weight — stake times score times originality — and get their own stake back on top.

Anyone can play free with a paper desk — no wallet, no signup, scored on the same leaderboard as everybody else. A season runs twelve weeks and at the end the top five hundred unseated players mint a seat for the network fee only. It is a ranking, not a pass mark, which is what makes farming it self-defeating.

minimum to qualify

30 settled calls

spread requirement

10+ distinct bells, 4+ instruments

ranked on

lower bound of skill, not the average

seats per season

500

lifetime cap

5,000

mint cost

network fee only

Rank uses the lower bound of your mean skill rather than the mean itself, so a short run of perfect calls has a wide interval and stays low. Sample size is a structural requirement, not a formality, and a hundred lucky throwaway accounts rank below one player with a real body of work.

The ladder

• ›Junior Analyst — three instruments, small stake ceiling.
• ›Analyst, at 30+ calls — the full slate and a bigger ceiling.
• ›Senior Analyst, at 100+ calls — tune your own parameters.
• ›Vice President, at 250+ calls — run a second desk.
• ›Partner, at 500+ calls — blend strategies into your own ensemble.
• ›Chairman, at 1,000+ calls — publish original strategies and take a royalty forever.

Rank falls as well as rises, and the portrait follows it in both directions. The record is never edited: a poor stretch is a discount on the secondary market, not something that can be deleted.

Every seat is backed by WETH held in escrow and funded by a fifth of the rake. Hand a seat in and escrow pays the published floor; anyone can buy a dormant seat back out at the floor plus five percent, which refills escrow and tops up the pot.

The number starts at zero and ratchets upward only. It is published live and can never be revised down. A floor priced in the project's own token falls with that token — this one cannot.

The mechanism does not need a token. Stakes are WETH, prizes are WETH, the floor is WETH. $OUTCRY exists for the capability sinks below, and it launches only after Season 1 has settled in public — so anyone considering it can read a full season of settled rounds first.

total supply

1,000,000,000

team allocation

0 — fair launch

mint function

does not exist

pause / blocklist

none

admin powers

empty by design

needed to play

no

governance rights

none

What it may never do

• ›Buy a seat — seats are ranked into existence.
• ›Buy rank, skill, or a better score multiplier.
• ›Improve the odds on any call, at any size.
• ›Grant an early look at the crowd before the lock.

Stake size scales a win and a loss by the same number, so a larger stake is exposure, not advantage. The moment a token can buy an edge, the leaderboard stops meaning anything — and the leaderboard is the entire product.

Settlement price

A forecasting game is only as honest as its settlement price, and on a young chain with thin pools a single tick is trivially movable. Resolution uses a five-minute average sampled on-chain, checked against an independent reference over the same window. If the two disagree beyond a published tolerance the round voids for that instrument: every stake returns in full, nothing scores, and the void is written to the record.

Voiding is cheap. Settling one manipulated round would be the end of the leaderboard's credibility, so the bias is deliberately toward voiding.

Anti-gaming

• ›Reading the crowd then fading it — impossible, the crowd does not exist until it is frozen.
• ›Farming seats with throwaway accounts — seats are ranked, so sybils compete against each other and dilute their own odds.
• ›Getting lucky in a small sample — rank uses the lower bound, and small samples rank low.
• ›Hedging two seats on opposite sides — the loss forfeits more than the matching win collects.
• ›Coin-flipping — the expected score of a random call is zero before rake and negative after.

Do I need a wallet to play?

No. The free paper desk needs nothing at all — no wallet, no signup — and it is scored on the same leaderboard as seated players. A wallet is only required once you hold a seat and want to stake.

Can I just buy a seat?

Not from us. Seats are only ever minted to the season's top five hundred. You can buy one on the secondary market from somebody who won theirs — and you inherit their record, good or bad.

Is there an edge in the published strategies?

No, and we will not pretend otherwise. A moving-average crossover is not alpha; if free published strategies beat the market they would not be free. The edge is never the strategy, it is being on it while everyone else is not.

What happens if nobody trades?

Nothing breaks. The pool is funded by players who were wrong, not by trading volume, so it scales with participation rather than with speculation. Fifty players makes a small pot; fifty thousand makes a large one.

When does the token launch?

After Season 1 settles. Not before. The tournament runs with real money first so the curve opens against a public track record rather than a promise.