Why Everyone Starts Flat

Your challenge line does not start scoring the moment the challenge starts. It starts at the first moment your account is flat on the markets the challenge scores, at or after your entry. Until then, you are entered, you are on the page, and you are earning nothing.

This is the rule most likely to surprise a first-time entrant, and it is the rule that makes the rest of the scoring worth anything.

The problem it solves

Imagine the rule did not exist. The optimal strategy for a PnL challenge would be:

1. Open a large position several days before the challenge.
2. Wait for it to go in your favour.
3. Enter the challenge.
4. Close the position inside the window.

Every dollar of that profit would book as realized PnL during the contest. You would have banked a move that started before anyone else was competing, and the leaderboard would report it as trading skill inside a 24-hour window.

That is not a hypothetical exploit, it is the obvious one — and in a contest where entry is permissionless and the ticket is cheap, obvious exploits get used. The flat-start gate closes it by making the challenge measure only what happened inside it.

The exact boundary

The mechanics are more precise than "you must be flat at the start", and the precision matters.

Your line's window opens at the first instant your net position on the scoped markets reaches zero, at or after your entry time. Not at the start of the challenge, not at your first trade — at the moment you went flat.

A consequence that surprises people: the trade that closes your pre-existing position is not scored. Your window has not opened yet when that fill happens; it opens the instant the fill leaves you flat. So the profit or loss on unwinding a position you carried in belongs to the period before the contest, which is exactly where it was earned.

From that boundary onward, everything counts: realized PnL on every fill, volume toward your bracket, and — for whichever line is still live at the end — unrealized PnL on open positions.

Two more details worth knowing:

• "Flat" means flat on the scored markets, not on your whole account. A challenge scoped to BTC and ETH does not care about your SOL position. A challenge with no market scope cares about everything.
• Never being flat means never scoring. A trader who holds a scoped position continuously from before the start to after the end books nothing at all. If you are running a challenge, say this on the page — it is the single most common way an entrant ends up with a zero they did not expect.

For re-buys, it is not optional

For a first entry, the flat-start gate is a challenge setting. For a re-buy it is unconditional, and it cannot be turned off.

The reason is structural. A re-buy creates a new line that freezes the previous one, and a single venue account can only trade one book — so the two lines have to partition your timeline cleanly. The handover point is the moment you are flat. Without that, a fill would belong to two lines at once, or to neither.

This is also why buying a re-buy while holding a position is such a bad deal: the ticket is spent, the new line does not start, and the old line keeps running underneath. Which brings us to the check.

Asking the venue before taking the money

The scoring engine enforces flatness silently and after the fact. That is correct for scoring and terrible as a user experience — the player finds out by not being on the board.

So the entry path asks first. When you buy a re-buy, Voltrade reads your live positions from the venue and refuses the purchase if you are holding anything in scope, naming the markets and their rough notional so you know exactly what to close.

Deliberately live, not from Voltrade's own rollups. The rollup tables refresh on a five-minute cron, and five minutes is precisely the window in which someone closes a position and then clicks the button. Reading a cached position would refuse a purchase that should succeed, or allow one that should not.

Reads are venue-specific in ways worth knowing:

• Hyperliquid: positions come from the clearinghouse state for your wallet. Any non-zero size counts.
• Lighter: only the main account is read. Challenges score the main account, so sub-account trades do not count and a position held there does not block a re-buy either. Spot pairs are excluded; perps only.

Market matching is case-folded on both sides, which sounds like housekeeping and is not. Hyperliquid names its 1000x markets kPEPE, kBONK, kSHIB while the market picker uppercases everything it stores. An exact match found nothing — challenges scoped to a k-market scored zero for everyone and, once the flat gate read the same list, cheerfully reported a wallet flat while it held an open position. The k is not stripped, though: kPEPE is a distinct market from a hypothetical PEPE, and folding them together would scope in fills the challenge never meant to score.

Failing closed

The venue read can fail. Hyperliquid can be slow, Lighter's reads queue behind a shared throttle, a request can hang. What happens then is a design decision, and Voltrade makes it explicitly.

The read reports three states, not two: flat, not flat, and unknown. Unknown is what you get when the venue could not be reached inside an eight-second deadline. It is never quietly converted to "flat".

An unknown read blocks the purchase, with a message telling you to try again in a moment. That is the fail-closed choice, and it costs you a retry. The alternative — assuming flat when nobody knows — sells you a ticket into a state that was never verified, and the failure is silent and unrecoverable: you have paid, your line will not start until you close whatever you might be holding, and your old line is frozen or still running with no way to undo any of it.

A retry is an annoyance. A ticket sold on a guess is a refund conversation. The eight-second deadline exists for the same reason: a hung venue connection on a path where a person is waiting to spend money has to become "unknown" long before it becomes "the request died".

What this means as a trader

Three practical habits:

• Be flat before the start. If you know you are entering a challenge at 14:00, plan to be flat on the scoped markets at 14:00. Otherwise your window opens late and the field has a head start.
• Read the market scope first. It tells you what "flat" means for this challenge. A scoped challenge may not care about most of your book.
• Close, confirm, then re-buy. The panel has a re-check button and the check itself is live, so give the venue a beat after closing rather than clicking through a refusal.

And as a creator: put the rule on your challenge page in plain words. "Your score starts when you are flat" prevents more support questions than any other sentence you can write.

The short version

The flat-start gate makes a challenge measure the challenge. Your line opens at the first moment you are flat on the scored markets, the closing trade that got you there is not scored, and a wallet that is never flat never scores. For re-buys the rule is unconditional and checked live against the venue before your money moves — and when the venue cannot answer, the answer is "not yet", never "probably fine".

More on the scoring machinery in how to run a crypto PnL challenge, on per-line brackets in volume brackets explained, or see it running on the challenges page.