Browse categories
Explore
Fiverr Pro
English
$
USD
I will backtest your crypto trading strategy with an honest pass fail verdict
your go to for polymarket automation and trading bots
I build trading bots and data infrastructure that run unattended for months without babysitting.
I currently run a live automated trading system on Polymarket: 42 markets, 18 forecast models feeding ...
About this Gig
Most backtests look good because the rules were tuned on the same data that grades them. I run your rules through the same 10 gate test I use on my own strategies, then tell you whether they pass.
WHAT YOU GET
- A one page plain English verdict: PASS, NEAR, FAIL or INCONCLUSIVE
- G1 to G10 gate table with every measured number
- A quantstats tearsheet of the graded returns against BTC buy and hold
- Trade list and raw report, so you can recompute every number
THE TEST
Fees, slippage and funding included. Random timing null. Best 3 trades removed. Costs doubled. Deflated Sharpe for the versions you tried. Quarterly stability. Assets you did not tune on. A never seen 2017 to 2019 era. The bar was written down before my first run. More than 20 of my own ideas went through it. None passed.
WHAT I NEED
Your rules (Pine Script, Python or plain words with exact numbers), asset and timeframe, the dates you built it on, and how many versions you tried. Binance USD-M perpetuals, 1 hour bars and above, history to September 2026.
I grade your rules exactly as sent and never tune them. I validate: no predictions, no advice, no promised returns. Written messages only, no calls.
Platform:
TradingView
•
Binance
Development technology:
Python
•
PineScript
FAQ
Do you predict returns or tell me what to trade?
No. I validate a historical simulation of your rules against a fixed test. I give no predictions, no advice and no promises about future results.
Will you optimize or improve my strategy?
No. I grade your rules exactly as sent. Tuning on the test period is what makes backtests look better than they are. The gate table shows which test failed so you can decide what to change.
What if I built it on the last two years?
Then those years are not a fair test of it. I run the longest period your design did not touch and say so on page one. If none is left, the verdict is INCONCLUSIVE and I explain why.
Which markets and timeframes do you cover?
Binance USD-M perpetual futures, 1 hour bars and above, history to September 2026, plus BTC and ETH spot back to 2017 for the deep era check. Anything else: message me before ordering.
What do you do with my code?
I read it as text and re-implement the rules myself. I never run your code. Your files are used only for your order and deleted 14 days after delivery.
What does INCONCLUSIVE mean?
There are too few trades, or no untouched data left, so the numbers cannot support PASS or FAIL. The verdict says which. It is a true statement about what the data can show, not a soft fail.
Can we do a call?
No. Everything is written through Fiverr messages, so the record is clear for both of us.
