Methodology

How a Hyperliquid trader is scored

Every position, fill and liquidation on Hyperliquid is public. That makes trader evaluation an engineering problem rather than a trust problem — but only if you measure the things that repeat instead of the things that impress.

In short

Candidates are scored on realized PnL consistency, win rate, profit factor, position discipline and account survivability. Each dimension carries a hard floor; failing any floor disqualifies the trader outright. Survivors receive a composite score, and capital is allocated in proportion to it.

Step one: reconstruct the record

Scoring starts from the account's own fill history rather than a displayed PnL figure. Fills are grouped into positions, positions into closed trades, and closed trades into a realized PnL series with timestamps. Unrealised PnL is deliberately excluded: an open position is a claim, not a result.

This reconstruction is also where records get rejected. Too few closed trades, too short a history, or a series dominated by a handful of outsized outcomes cannot support a score, no matter how large the headline number is.

Step two: the five dimensions

Each dimension answers a different question about repeatability, and each has a floor rather than a sliding scale, so a catastrophic weakness cannot be averaged away by a strength elsewhere.

  • Realized PnL consistency — is the curve broadly distributed, or one lucky trade with noise around it?
  • Win rate — read jointly with average win/loss, never alone.
  • Profit factor — gross profit over gross loss, the cleanest single ratio for edge quality.
  • Position discipline — sizing stability, adding to losers, liquidation proximity.
  • Account survivability — has the account ever been effectively wiped and refunded?

Step three: composite score and weight

Surviving candidates are combined into a single composite score, which becomes the allocation weight. Weighting by score rather than splitting evenly means exposure tracks the strength of the evidence, while per-leader notional ceilings prevent any single high score from dominating the book.

What the score is not

A score is a compressed description of a past record under a fixed set of tests. It is not a forecast, it does not detect intent, and it cannot see off-exchange hedges or coordinated accounts. Regime change can invalidate a high score without any visible failure on the leader's part, which is why scores are recomputed continuously and probation exists.

Past performance is not indicative of future results. Perpetual futures are leveraged instruments and carry a substantial risk of loss, including the loss of your entire position.

Questions

Frequently asked

Is the score based on displayed leaderboard PnL?

No. It is computed from the account's reconstructed fill and closed-trade history, which is why it often disagrees with short-window PnL rankings.

How long a history is required?

Long enough for the consistency and profit-factor tests to be meaningful. Records that are too short or too thin are rejected rather than scored optimistically.

Can a trader game the score?

The floors are deliberately hostile to the usual tactics — leverage spikes, adding to losers and wipe-and-refund patterns all reduce or disqualify a score rather than raising it.

Diversified copy trading. On autopilot.

Score-weighted allocation across up to 10 elite Hyperliquid traders, each isolated in its own sub-account. Your funds never leave your account.

Non-custodial · Agent cannot withdraw · Cancel delegation anytime