Pickled Lizard · Learn
Learn
How we test AI trading ideas before we trust them. Plain words, no hype.
October 27, 2026 · Michael Shane
Why We Refuse to Count Trades We Can't Verify
We found trade records we couldn't trust. We kept them, and we don't count them. Here's why honest results matter more than good-looking ones.
October 26, 2026 · Michael Shane
Position Sizing for Beginners: How Much Should You Risk Per Trade?
Position sizing decides how much you put into each trade. Here's a plain way to size trades by risk, with simple math, a losing-streak example, and the mistakes that hurt beginners most.
October 25, 2026 · Michael Shane
The Deflated Sharpe Ratio: Why 1,000 Backtests Need a Higher Bar
Test enough trading ideas and one will look great by pure luck. The deflated Sharpe ratio raises the bar based on how many things you tried. Here's how it works, in plain English, with a simple simulation.
October 24, 2026 · Michael Shane
Monte Carlo Simulation for Trading Strategies, in Plain English
A Monte Carlo simulation reshuffles and resamples a strategy's trades thousands of times to show the range of outcomes a backtest hides. Here's how it works, how to read it, and what it can't tell you.
October 23, 2026 · Michael Shane
Point-in-Time Truth: What Did the AI Know, and When?
Look-ahead bias makes bad trading ideas look great. Here's how we record what a strategy knew at each moment, so results can be checked later.
October 23, 2026 · Michael Shane
In-Sample vs Out-of-Sample Testing: What a Fair Split Looks Like
Out-of-sample testing checks a trading strategy on data it never saw while you built it. Here's how to split your data, how big each part should be, and the mistakes that quietly ruin it.
October 22, 2026 · Michael Shane
Walk-Forward Analysis, Explained With a Simple Example
Walk-forward analysis tests a trading strategy the way you'd really use it: tune on the past, then trade the next stretch blind. Here's how it works, step by step.
October 20, 2026 · Michael Shane
Backtest vs Forward Observation: Why a Nice Chart Is Not Proof
A backtest shows what would have worked. Forward observation shows if an idea survives. Here's how we test AI trading ideas before we trust them.
October 19, 2026 · Michael Shane
Can You Trust an AI Trading Bot's Backtest? A 10-Point Checklist
An AI trading bot shows you a great backtest. Before you believe it, run through these 10 plain questions about data, costs, lookahead, trade count and more.
October 18, 2026 · Michael Shane
Paper Trading vs Live Trading: Why Your Results Drift
Why paper trading results differ from live trading: fees, slippage, missed and partial fills, latency, funding and behavior. Plus how to measure the gap first.
October 17, 2026 · Michael Shane
Backtest Overfitting: 7 Signs Your Strategy Only Works on the Past
How to tell if a trading strategy is overfit (curve fit) to old data: 7 plain warning signs, quick tests for each, and what to do next.
October 16, 2026 · Michael Shane
How to Backtest a Trading Strategy Without Fooling Yourself
A plain guide to backtesting a trading strategy honestly: clean data, real fees, no peeking, a held-back test period, and counting every try.
October 15, 2026 · Michael Shane
How to Test an AI Trading Strategy Before You Use Real Money
Got a trading idea from ChatGPT, Claude, or Grok? Here's a plain, step-by-step way to test it on paper before you trust it with real money.