$PBAY is pre-launch — no contract, no price, no announced date. Here is what a prediction-market platform token can plausibly do, what it structurally cannot, and how to take one apart before you decide it matters.
Predict Bay ·
$PBAY is the token concept for Predict Bay. It has not launched. There is no contract address, no listing, no distribution event and no announced date — and this page will not invent any of those, because a token page that guesses at numbers is worth less than one that states plainly what it does not know. What follows is the version that is actually useful before a launch: what a token attached to a prediction-market platform can plausibly do, what it structurally cannot do no matter how the docs are worded, and how to take one apart.
ROUND CLOSES IN
--:--
LAST PRICE
—
STRIKE
—
NO POSITIONS YET THIS ROUND — BE THE FIRST
Practice mode · no wallet, no deposit, free to start
One distinction is worth making before anything else. A platform token belongs to the interface you trade through. A protocol token would belong to the on-chain system underneath it — in our case DeepBook Predict, the binary-options protocol on Sui that our Sui trading mode is built against, currently on that protocol's testnet deployment. Those are different objects with different claims on value, and conflating them is the most common error in this category; does DeepBook Predict have a token? handles that question on its own terms. Everything below about $PBAY is conditional. Where you read *could* or *would*, that is deliberate.
The accurate description of $PBAY right now is a named intention with a described utility surface. The token page at /token sets out that intent — the utility categories the token is meant to carry, and a planned distribution design. Read both as a plan rather than a state of the world: it is not a contract you can query, nothing described there exists on-chain, and treating a described utility as a shipped one is how people end up holding something that never had a sink.
The first row is the one worth holding onto. If a contract claiming to be $PBAY exists somewhere today, it did not come from us, and matching branding changes nothing about that. The only address that will ever be worth trusting is one published on our own surfaces and resolvable on a block explorer to a publisher we have named. Until such an address exists, the correct number of tokens to hold is zero.
Across prediction markets, exchanges and perp venues, platform tokens cluster into four utility categories. They are worth understanding as categories, because the marketing language varies far more than the mechanics do. Each is described here as a general design pattern, and as something $PBAY *could* be structured around — not as a commitment.
The most common design: holding or staking the token moves you into a lower fee tier, or earns a rebate on fees already paid. It is popular because it is the one mechanism where the token touches a cash flow that genuinely exists. The details decide whether it means anything. A discount tier denominated in *token units* silently becomes cheaper to reach as the price falls and harder as it rises, which makes the benefit pro-cyclical in the wrong direction. A tier denominated in the settlement asset's terms tracks intent better, but it needs a price feed for the token itself, which is a new dependency to maintain and to attack.
Tokens are also used to pay for behaviour a platform wants: early usage, referrals, quoting, holding a position through resolution. This is distribution dressed as utility, and that is not automatically bad — a market needs holders who use it, and issuance is one way to find them. The question is what happens when the emissions stop, because the behaviour that was purchased usually stops with them unless the product was worth using at zero subsidy.
Governance has the widest gap between claim and substance. "Vote on protocol parameters" means something concrete if the parameters live on-chain and the vote binds an upgrade authority. It means considerably less if the vote is advisory and the deploy key sits with the team either way. The useful question for any token, including this one, is not whether governance exists but which specific values it can change and what enforces the result.
Staking is a lock with a reward attached. It reduces float and can be paired with something economically real — backing market-maker inventory, bonding a market creator against bad-faith listings, insuring a settlement edge case. Where staking is *only* a lock paying issuance, it is a slower form of emissions and should be read that way.
Any honest token design starts from the revenue surface, because that is what a token can attach to. In a prediction market that surface is narrower than in a spot exchange, and it is worth laying out explicitly.
| Surface | Who pays it | What a token can plausibly do |
|---|---|---|
| Trade fee on entry or exit | Whoever crosses the spread, per position | Tiered discounts by holding or stake; rebate paid back in token |
| Settlement or redemption fee | The winning side, at expiry | Same idea, but harder: it lands at settlement, so tier eligibility has to be evaluated at resolution rather than at trade time |
| Spread and maker edge | Anyone taking liquidity | Nothing directly — this is a depth problem, not a token problem |
| Gas and oracle costs | The chain, or whoever sponsors the transaction | Can be subsidised from a treasury, but the cost is denominated in the gas asset, not in the token |
| Market creation | The creator of a market | Bonded creation in token, slashable if a market is malformed or resolved in bad faith |
Read down the third column and the pattern is clear: a token attaches cleanly to fees and to bonds, weakly to subsidies, and not at all to spread. Designs that claim otherwise are usually describing a treasury spending money — a decision reversible at any time — rather than a mechanism.
This is the section most token pages skip, and it is the one that tells you whether the author understands their own system.
There is a failure mode specific to prediction markets worth spelling out, because it is not obvious and it recurs constantly. A prediction market's actual output is a price — a number meant to carry information about the probability of an event. That is the product. Everything else is plumbing.
Now attach a token reward to trading volume. In a binary market the two sides are complementary: buying both sides of the same contract at a combined cost near the payout leaves almost no directional exposure and a large volume number. Volume-weighted emissions therefore pay people to generate volume that carries no information, and the price they push around while doing it is the thing everyone else is reading. The incentive did not merely waste tokens; it degraded the output.
Designs that survive this reward things that are expensive to fake. Quoting rewards paid on *resting depth at risk* — size that sits in the book long enough to be picked off — cost the earner real adverse selection. Accuracy-style scoring, where a forecaster is graded against realised outcomes across many resolutions, is harder to farm than volume, though it needs weighting by capital at risk or it fragments into thousands of tiny accounts. Neither is a solved problem. The point is that in a prediction market a rewards schedule is not a marketing line item; it is a change to the price signal itself.
Apply the following to $PBAY when it exists, and to every competitor's token now. If a page cannot answer these, the usual reason is that the mechanism has not been built.
Running one checklist across venues is more informative than reading any single token page — Predict Bay vs Polymarket compares the underlying market models these tokens would sit on top of, which is the more load-bearing comparison.
If and when a launch happens, the verification procedure is mechanical:
Start from the product, not a message
Take the address from the platform's own domain and app surfaces. Every launch attracts impersonation, and social channels are where it lands first.
Resolve the address on-chain
Put the published address into a block explorer. Confirm the publisher, the upgrade or mint authority, and that the module is what the page claims it is.
Find the sink inside the app
Open the interface and look for the mechanism in place — a fee tier that changes, a bond you can post, a gate that opens. If the utility exists only in prose, it does not exist.
Compare emissions against fees collected
If scheduled rewards outrun realistic fee income, the gap is funded by dilution. That may be an acceptable growth cost, but it is a cost, and it should be visible.
Watch one full settlement cycle
Let markets open, resolve and pay out once with the mechanism live before treating any of it as proven.
What is real on Predict Bay right now is the trading stack, not the token — and the honest version of "real" is narrower than most sites would admit. EVM real-money trading on Polygon is paused, and the DeepBook Predict path targets that protocol's Sui testnet deployment and is switched off for real funds. With EVM withdrawn, the mode switch on a market page renders two options rather than three — Practice and Sui — and selecting Sui shows a coming-soon panel in place of the trading controls. Practice mode is what you can actually trade today.
The markets themselves are short-duration price markets on BTC, ETH, SOL, SUI, DOGE and XRP, in hourly and daily rounds: a strike is set at the open, you take above or below, and the round resolves against an on-chain price feed. Practice markets mirror the real ones exactly — same assets, same strikes, same rounds, same on-chain resolution — with no capital at risk. How it works walks the full path, and the BTC market is the shortest way to watch a round run end to end.
Sign-in is Google via zkLogin, so there is no seed phrase to store, and transactions can be sponsored so you do not need to hold SUI for gas; connecting a Sui wallet directly is supported too. Prediction markets without a seed phrase covers that path in detail. None of it requires a token, which is rather the point. That is the state of things, and it is deliberately easier to find here than a token page would normally make it. When there is something concrete to say about $PBAY, it will come with an address and a mechanism attached rather than with adjectives.