Time-Value Mechanics

1. Early conviction earns double. Late certainty earns nothing.

Every stake carries a time weight set at the moment you place it. The entire first hour of a poll counts at the full 2× toward your share of the losing side's pool, so you have time to rotate funds between polls without racing the clock. After that the weight falls steadily and reaches zero in the final 10 minutes.

A final-minutes stake that wins gets its money back minus the fee, and no share of the pool. Late money can't free-ride on early conviction, and it can't dilute the people who committed first. Riding the bandwagon isn't banned. It's just pointless.

2. Your downside is capped at about 12%

Being early doesn't mean being trapped. You can exit any position before the poll closes, at face value, never at a crashed market price. If you exit the trailing side, you leave 10% of your stake in the pool for the eventual winners, plus an exit fee of up to 2%.

Compare that to prediction markets, where a wrong-side exit can cost 50–80%. On b⁴, early money gets the most upside and a known, capped downside. That asymmetry is the product.

3. You see the odds, not the race

While a poll is live, b⁴ shows you a price, not a scoreboard:

• Live payout multiples per side, personalized to your time weight: “Your $25 on NO → est. $85 if NO wins.”
• The price chart, showing each side's share of the staked money in cents. 30¢ means 30% of the money is on that side, and it pays roughly 1/0.30 ≈ 3.3× if it wins.
• Total activity: combined dollars staked and number of stakers.

What the app doesn't display until the poll resolves: exact vote counts and each side's exact pool. Votes decide the winner, and they aren't shown while the poll runs, so there is no leader on screen to pile onto and no scoreboard to game. (Like everything on b⁴, raw poll data lives on the public Solana blockchain; what changes is that the game no longer rewards watching it.) One thing to remember: odds reflect money, not votes. 85¢ doesn't mean an 85% chance of winning. It means the payout on the other side is large. A lopsided market is an opportunity, not a verdict.

4. Exits and switches

• Exits are always at face value, minus the fee (1–2%, rising over the poll's life).
• Leaving the trailing side leaves 10% of your stake in the pool. Leaving the leading side costs only the fee.
• You get one switch per poll (YES↔NO). Switching sells your old position (10% stays behind if it was trailing) and your new position gets the time weight of the moment you switch. Switch late and the new position earns ~0.
• Exiting forfeits your resolve claim. Holding to resolve keeps your full entry-time weight.

5. The final 10 minutes: trusted members only

The last 10 minutes of every poll have one flat rule, for entries and switches alike:

Gold reputation or above · $25 minimum · no time bonus

• Gold requires real history: at least $1,000 of lifetime staking volume held to resolution, and an account at least 2 weeks old.
• The $25 minimum and zero weight apply to everyone in the window. No exemptions, same rules for both sides.
• Below Gold? Your window closes 10 minutes before the poll does. Stake earlier. It pays better anyway.

Why this exists: the final minutes are where coordinated vote manipulation used to be cheapest. Now every wallet acting late has months of reputation behind it, and something real to lose.

6. The reputation ladder pays for itself

Reputation levels are earned with real history: staking volume held to resolution (exited positions don't count), account age, wins, and quality comments. The higher tiers come with perks that give back a real share of the fees you've paid.

Platinum perks

• 10% fee rakeback: a tenth of the base fees you generate, rebated to your wallet monthly in USDC.
• 1.25× time-pool boost on your early-holder claims.
• Higher Loser Recovery ceiling when you hold, lose, and engage.
• Platinum name color, profile flair, and pinned comments.
• Priority support and higher send limits.

Diamond perks

• 25% fee rakeback, paid monthly in USDC.
• Creator access: create markets and earn creator fees on every poll you launch (0.25% of the winning pool plus 0.5% of the losing pool).
• Diamond club access: vote on upcoming market topics, shape the roadmap, and get a direct line to the team.
• High-roller markets: Diamond-only polls with deeper pools.
• IRL events and merch.

7. Fair play policy

Coordinated vote manipulation (flooding a side with sybil wallets, mass-switching at the buzzer, faking momentum) is detected after resolution by reviewing clusters of coordinated wallets. Penalties are published; detection thresholds are not.

• Strike one: demotion by one tier, and your volume clock restarts from the strike date.
• Strike two: permanent exclusion from final-window staking.
• Coordinated clusters lose platform access entirely (staking, Loser Recovery, leaderboard, referrals).
• Poll outcomes are never reversed. Penalties apply to accounts, not results.

One honest stake per person: wallets are bound to a physical device, one wallet per device, for the life of the market.

8. The numbers at a glance

9. Why we built it this way

The old pattern was bad for everyone honest: nothing happened for 99% of the poll, a bandwagon formed in the last 10 minutes on the visible leader, and losers switched off before close, leaving winners a near-empty pool. Early conviction was punished; late certainty was free.

v2 inverts it. Early money earns double weight with a capped downside. Late money keeps only itself. Exits are gentle but never free, so winning always pays. And the endgame belongs to members with real history, playing blind, with reputation on the line.

Stake early. Stake what you believe.