Independent crypto strategy validation

Is your crypto strategy really profitable?

Most backtests look great and fail live. I test your strategy on the full Binance history, with real fees, slippage and funding, and validate it scientifically. You get a clear diagnosis before you risk real money.

Nicolas Poliz.Nicolas Poliz.
Medical Doctor (MD) · MBA in Finance
Strategy diagnosticBD-2026-0003
Golden Cross · SMA 50/200 · daily · every Binance crypto future since 2021
EDGE SCORE1 / 5
No evidence of edge. Loses money even before costs, and fails out-of-sample with every parameter set tested.
Winning trades18%
Average trade, after costs−12.7%
Funding cost per trade−1.1%
Parameter sets profitable on new data0 / 18

From a real demonstration report on actual Binance data.

The problem

Most backtests are flawed. Here is how.

A strategy can look perfect on a chart and still lose money live. These are the errors I screen for in every diagnostic.

×

Overfitting

Parameters tuned to past noise. I map results across the whole parameter space and validate on data the optimisation never saw.

×

Lookahead & repainting

Signals that use information not yet available. Only confirmed candle closes are used, and your code is reviewed.

×

Data snooping

Testing many ideas and keeping the best. Every variant is counted and significance is adjusted for it.

×

Selection & survivorship bias

One lucky period, or only the coins that survived. I test full market cycles, delisted coins included.

×

Hidden costs

Zero fees, perfect fills, no funding. Fees, slippage and the actual historical funding rates are modelled.

×

Too little data

A handful of trades proves nothing. You get significance tests and Monte Carlo ranges, not just a headline return.

Real data

Official Binance data. The full history.

Your strategy (Pine Script, Python or written rules) is implemented on my custom Python backtesting engine, reconciled with your own results, and run on any Binance crypto market.

MarketsSpot, USD-M and COIN-M futures
SymbolsAny crypto pair, delisted coins included
Timeframes1 minute to daily
HistoryFull available history
CostsFees, slippage, actual funding rates
Order flowTick-by-tick trades: delta, CVD, footprint

Scientific analysis

Evidence first. Like a medical diagnosis.

Every result is tested so that luck is never mistaken for skill.

Out-of-sample testingWalk-forward validationMonte Carlo simulationParameter robustness mapDeflated Sharpe ratioPer-coin breakdown
  1. 01

    Send your strategy

    Pine Script, Python or plain words, plus your own results if you have them.

  2. 02

    Rebuild & reconcile

    Your rules on my engine, checked trade by trade against your results.

  3. 03

    Full protocol

    Real costs, out-of-sample, walk-forward, Monte Carlo, robustness.

  4. 04

    Your diagnosis

    A structured PDF report and an Edge Score from 1 to 5.

What you receive

A clear verdict, and what to do next.

Where it fails

Before or after costs, which coins, which periods, which parameters. The exact weak point, not a vague "it doesn't work".

How to improve it

Concrete suggestions, and tested variants of your strategy, validated on data they were never fitted to.

An honest Edge Score

A single number from 1 to 5, backed by the full statistical evidence in the report.

1No evidence of edge
2Weak or unstable
3Inconclusive
4Probable edge
5Robust, consistent edge

Real reports

Two famous strategies. Real Binance data.

Demonstration reports on publicly known strategies, so you can see exactly what you get.

EMA 9/21 crossover + MACD

5-minute candles · Binance crypto futures · 2024–2026

EDGE1/5
Trades analysed1.1 million
Average trade, before costs−0.03%
Average trade, after costs−0.15%
Parameter sets profitable on new data0 / 27

Golden Cross, SMA 50/200

Daily candles · Binance crypto futures · 2021–2026

EDGE1/5
Winning trades18%
Average trade, after costs−12.7%
Funding cost per trade−1.1%
Parameter sets profitable on new data0 / 18

Packages

Choose your diagnostic

Every package covers any Binance crypto market, symbol and timeframe, the full history and real costs. Orders are placed securely through Fiverr.

Standard Review

$20

Delivery in 1 day

  • Full-history test with real fees, slippage and funding
  • Monte Carlo simulation
  • Edge Score verdict
Order on Fiverr
Most chosen

Full Diagnostic

$70

Delivery in 2 days

  • Everything in Standard Review
  • Out-of-sample and walk-forward validation
  • Order flow analysis (delta, CVD)
  • 3 strategy variants tested
  • Full PDF report
Order on Fiverr

Comprehensive Diagnostic

$200

Delivery in 3 days

  • Everything in Full Diagnostic
  • Tick-level footprint analysis
  • Parameter map and robustness tests
  • 10 strategy variants tested
  • Detailed PDF report
Order on Fiverr
Nicolas Poliz.

About

Nicolas Poliz., MD, MBA

I am a medical doctor with an MBA in Finance. In medicine, a diagnosis must be evidence-based and reproducible. I founded Backtest Doctor to bring the same standard to crypto trading strategies.

Every report states its method, its data and its limits. If your strategy has an edge, you will know. If it does not, you will know before you lose money.

FAQ

Frequently asked questions

What do I need to send you?

Your Pine Script, Python code or a clear description of the rules (entry, exit, stop loss, take profit), the coins and timeframe, and if available your own backtest results or trade list from any platform, so they can be reconciled trade by trade.

Which markets do you cover?

Crypto only, on Binance data: spot, USD-M and COIN-M futures, any symbol and timeframe from 1 minute to daily, across the full available history. Stocks, forex and commodities are not covered.

Which biases do you check for?

Lookahead and repainting, overfitting, data snooping, selection bias, survivorship bias (delisted coins included), ignored fees, slippage and funding, unrealistic fills and insufficient sample size.

How do you improve a strategy without overfitting it?

Each variant is developed on one part of the data and validated on a separate, untouched part. The report states how many variants were tested and adjusts statistical significance accordingly.

Do you have order flow data?

Yes: tick-level trades for delta, CVD, footprint and imbalances. Full historical level-2 order book data is not available, so analysis is based on executed trades.

What if my strategy has no edge?

You get an honest Edge Score, a clear explanation of where and why it fails, and concrete suggestions for improvement. Knowing this before trading live is the point of the diagnostic.

Is my strategy kept confidential?

Yes. Your code, rules and results are used only for your diagnostic and are never shared or reused.

Is this financial advice?

No. It is a historical statistical analysis of the strategy you supply. Backtest Doctor does not recommend buying or selling any asset.

Know if your edge is real.

Before you risk real money. Delivery in 1 to 3 days.

Order on Fiverr