Pricing
One price for the whole pool
Purchasers do not choose a position, so there is one price. It is the harmonic mean of every active backing, plus a surcharge.
Formula
HM = n ÷ Σ(1 ÷ bᵢ)
price = HM × 1.03
The harmonic mean is deliberate. It stops any single large backing from dragging the price up, which keeps entry cheap while heavy positions stay rare.
House edge
≈ 7.8%
1 − (0.95 ÷ 1.03), and it holds at any pool composition.
Where the edge comes from
A purchaser pays 1.03 × the harmonic mean. The position they draw can be sold back to its depositor at 0.95 × that position’s backing. The gap between those two multipliers is the edge, and no pool composition changes it.
The only way to beat it is to keep NFTs whose market value exceeds 95% of their backing. That judgement is the entire skill component of the game — everything else is arithmetic.
For context: the protocol Tuck is modelled on runs a 85% sell-back, which gives roughly a 17.5% edge, and offsets it with daily token emissions. Tuck raised the rate to 95% precisely because Phase 1 has no emissions to offset anything.
Selection weight
Weight is 1e36 ÷ backing, so it falls as backing rises. A 5 USDG position is drawn sixteen times as often as an 80 USDG one. Light positions turn over constantly for a small return; heavy positions sit, accrue fees for longer, and cost more when finally taken.
Fee distribution
Every acquisition fee, minus the protocol cut, is split equally across all active positions — not in proportion to backing. Per acquisition every position earns the same amount.
Equal splitting keeps the acquisition price low and rewards many small deposits, which is what keeps the pool liquid. Lifetime earnings still scale with backing, because heavier positions survive far more acquisitions before being drawn.
The dynamic surcharge
The surcharge flexes with time since the last acquisition. Below a 60-second gap it is distributed normally; at or beyond a 3600-second gap the whole of it is rebated to the purchaser, sliding linearly in between.
The rebate goes to the purchaser who ends the quiet, not to the one after them. That is the only reading that produces the effect the mechanism exists for: at full rebate they effectively pay 1.00 × the harmonic mean and the edge drops to about 4.9%. Paying it forward instead would leave the person who breaks the silence on the full 1.03, and nobody would break it.
It is paid on allocation only. A draw that expires already refunds the whole price, so a rebate on top would return the same money twice. The whole surcharge is rebated, the protocol’s third included — a pool that needs reviving pays for its own revival.