Replayed, not idealised
Strategies run on market data from Hyperliquid and Binance, with real fees and slippage, and hourly funding on Hyperliquid.
Brierz Lab · Closed beta
Test a strategy before anyone trusts it, including you.
Most backtests look good because nobody counts how many were tried. Lab counts every attempt, prices every fill with real fees and slippage, and keeps part of the data sealed until the end.
How a test ends
Every strategy test ends in a plain verdict and the number of attempts behind it. No equity curve without its attempt count.
Promising, not proven37 attempts
Example verdict, illustrative values. The four verdicts:
- Passes the luck check
- Promising, not proven
- Looks like luck
- Too few trades to judge
Checked for honesty
Tested for look-ahead and walk-forward on request, checked for execution realism, with an estimate of how likely it is to be overfit.
Frozen, then followed
A strategy is frozen before the result and tracked on data it has never seen: Pine Script, ready-made strategies and rules today.
Written your way
Start from plain words, rules, Pine Script, Python, or a wallet’s real trades replayed as a strategy.
The assistant
An AI assistant that works inside the rules
Describe a strategy in plain words, or bring Pine Script or Python. The assistant writes the code, reads every check, explains why a result failed and suggests a fix. Every version that runs counts as an attempt, and the sealed data stays closed until the final test.
“The edge thins out on ETH once costs double. Trading BTC only would keep it. Want me to try that? It will be attempt 38.”
Venues
| Venue | Ratings | Lab |
|---|---|---|
| Hyperliquid | Public | Live |
| Binance | Not yet | Live |
| Bybit | Not yet | Next |
| OKX | Not yet | Next |
The engine reads a record, not a venue. Ratings need every account’s full history to be public; Hyperliquid publishes it.
From Lab to a live record
A strategy that holds up can be frozen before it trades and followed on data it has never seen. If it runs as a Hyperliquid vault, it is rated by the same rules as every other vault. Testing in Lab never changes a rating.
Traders and vault leaders
Test before you launch, and launch with a record you can stand behind.
Prop firms
Check a strategy before you fund it, with every attempt behind it on the table.
Strategy teams
Count every variant your team tries, not just the one that looked best.
Apply for the beta
A small group uses Lab with us while we test the method itself. Tell us what you trade and we’ll come back to you.