Choosing a Payout Structure for Your Campaign

The payout rule is the campaign. It decides who bothers entering, who keeps trading on day nine, how much a farmer can extract, and whether a thin turnout produces an embarrassing result or a defensible one. Everything else — pool size, window, markets — is downstream of it.

Here are the structures, what each one does to a field, and how they break.

Fixed rank tiers

Fixed percentages by final position. Presets: top 3 at 50/30/20, top 5 at 40/25/15/12/8, top 10 at 40/20/12/8/6/4/4/3/2/1. Percentages must total exactly 100 — the creation form will not let you past the step otherwise — and a competition can define up to 50 tiers.

What it does. Concentrates the pool at the top and creates a visible podium. A large trader has a specific, nameable prize to size up for. It is the shape that produces a story.

Where it breaks. The cheques are fixed in percentage, so they are fixed in dollars regardless of turnout. Four wallets entering means three of them split the entire pool for whatever volume they happened to do. That is the single worst-looking outcome a competition can produce, and it is structural rather than bad luck: fixed tiers pay the same top three whether four people entered or four hundred.

Field-size fit. Small, known fields, or launch-day campaigns where you have reason to expect a crowd. Bad for anything speculative.

Abuse resistance. Weak on its own. The prize for rank 1 does not shrink as one wallet's volume grows, so an account able to out-trade the field captures a fixed, known amount. Fixed tiers need a per-day counted-volume cap to be safe.

Tie handling worth knowing: tied ranks pool the tier percentages they span and split them equally, so a tie for first across two wallets pays each of them half of the first and second tiers combined rather than paying two firsts.

Pro rata

Your share of the pool equals your share of qualifying points.

What it does. Pays everyone who qualifies, in proportion to what they did. It is self-scaling in the way tiers are not: a bigger field does not cost more, it just means each participant's slice is smaller, so cost per dollar of induced volume improves as the field grows. With a thin field it concentrates automatically into whoever showed up, which is a much better failure than three wallets splitting a full pool.

Where it breaks. Without a floor it shreds the pool into dust — hundreds of sub-dollar payments that cost more in attention than they deliver in goodwill. And because the split is continuous, there is no podium and therefore less drama; a pro-rata leaderboard is informative rather than exciting.

Field-size fit. Broad and unknown fields. Revivals, multi-token baskets, venue campaigns aimed at breadth.

Abuse resistance. Structurally the strongest of the three, for a reason that is easy to miss: a farmer's marginal dollar dilutes every other farmer. The pool is fixed, so manufactured volume does not create new money, it only redistributes. Combined with a per-day counted-volume cap, extraction is bounded regardless of how much anyone trades.

Top-percent buckets

Instead of naming positions, name proportions of the field: "the top 10% share 60%, the rest of the top half share 40%". The winner count resolves at settlement as max(1, floor(N × topPercent / 100)) over the final eligible field, so it scales with turnout.

This shape is a challenge feature — the entry-fee competition format — not something regular sponsored competitions accept; sending it to a regular competition is rejected outright. That restriction exists because the read side of a competition page understands fixed positions only, and a payout shape that silently fails to render is a pool that pays nobody with no error anywhere.

Within a bucket, the pool follows a poker-style curve: position p gets weight p^-0.8, normalised inside the bucket. That produces a top-heavy podium with a long paying tail — first is roughly 1.74× second — instead of a flat split that makes winning meaningless.

Two behaviours to understand before using it:

• Payouts never increase down the ranks. Where a wider later bucket would out-pay the seat above it, adjacent ranks flatten into an equal-payout band. That smoothing can shift a bucket's realised cut away from its stated share, and on extreme custom splits can flatten the podium ratio too. The standard presets keep the ratio at every field size, and the page always shows the resolved numbers.
• Empty buckets fold upward. In a field too small to give a bucket any members of its own, that bucket's share is redistributed to the winners above it in proportion to what they already hold. With two entrants, a 10%/50% split pays the leader everything.

Field-size fit. Exactly the case the other two handle badly: a field whose size you cannot predict. The structure adapts instead of being wrong in one direction.

Qualifying gates

Gates are not a payout mode; they are filters applied before any mode runs. They matter more than most designers expect, because they define the denominator.

A minimum is the standard partner to pro rata: it stops dust payments and it tells a small trader honestly, up front, that this competition is not aimed at them. Set it against counted volume, which is what remains after the per-day cap — gating on raw volume would let a trader clear the bar with volume that earns nothing.

Be aware that some requirements — a minimum trade size, deposit or hold duration — cannot be projected live, so a competition using them cannot show traders what their rank is currently worth. That is a real cost: a leaderboard without projected rewards is much less useful for deciding whether the next dollar of volume is worth trading.

The lottery reserve

A raffle draws a fixed prize among traders who cleared a volume threshold in a period. Every eligible trader gets exactly one entry and the draw is uniform — a small trader who clears the bar has the same odds as the leader. That flatness is the entire point.

The arithmetic that campaign designers get wrong: the reserve comes out of the same pool. Total reserve is draws × prize per draw, with draws counted over the campaign window — one per day for a daily raffle, one per week rounded up for a weekly one. A 3,000-unit pool with a 25-unit daily draw over fourteen days reserves 350, leaving a 2,650-unit main pool. Projected rewards are computed against the main pool, not the advertised figure, and the reserve is required to be strictly smaller than the pool.

So the real decision is: how much of your headline number are you moving from "rewards performance" to "rewards showing up"? A large reserve on a small pool guts the leaderboard; no reserve at all on a long campaign leaves the tail of the field with nothing to play for after the podium settles.

The rule of thumb that follows from the mechanics: the longer the window, the more the raffle earns its keep. A three-day competition does not need one. A month-long one, where most participants can work out by week two that their rank will not move, mostly does.

Choosing, quickly

And one rule that outranks all of them: publish the structure, then freeze it. Changing a payout rule mid-campaign is the most damaging thing you can do to a competition's credibility, and no improvement to the rule is worth it. If something must change, the campaign is over and a new one starts.

For how the numbers are computed live, read how projected rewards are calculated. For the entry-fee format where top-percent buckets live, see challenges and how to run a crypto PnL challenge. For the abuse side of the same decisions, wash-trading resistance in volume campaigns.