01/ Create
Have an idea? Build it.
Describe how you want to trade. Pulse turns that into editable rules you can backtest and refine.
Build with Pulse
Crypto strategy design and automation
Crypto strategies on trial.
Most platforms sell the curve. Edge Theory shows the kill process first: a luck test, unseen data, a walk-forward and a stress test, before anyone treats that curve as proof.
Edge Theory is a crypto Bot-building platform. Describe an idea to Pulse in plain language, then review the editable rules. Test them in Strategy Lab on historical market data. Challenge a saved version in Evidence Lab. Simulated Bots are included on every plan. Supported Live automation on paid plans depends on your strategy, plan and exchange connection.
Dashboard Bot picks need a current B-or-better grade and a positive held-out simulated return.
Free · 40 AI credits each month · 10 new designs each month · 1 simulated Bot · no card required.
Live trading involves risk and can lose money.
Pulse · AI strategy designer
Bring your idea into the Lab.
Existing example idea
Long when a bullish engulfing candle forms and ADX is above 25, 4 hour chart
BTC · 4HBullish engulfingADX above 252× ATR stop2R target
Free account · 40 AI credits every month (10 designs) · no card needed
Automated trading involves risk. Backtests and simulated results don't predict future performance.
Built on real market data
Every Strategy Lab preview is labelled simulated, and Pulse makes no promises about returns.
- OKX
coming soon
Supported connections and market data sources vary by feature. Hyperliquid is coming soon and is not available for Live automation.
Choose your starting point
Your approach. Your starting point.
Already have a strategy, want to inspect an existing Bot, or need a harder look at the evidence? Each path still ends in a test, not a promise.
02/ Discover
Rather explore? Find a Bot.
Browse Edge Theory Originals and Marketplace strategies. Review the available rules, research and access before you choose what to inspect.
Explore Bots
03/ Research
Want a trial, not a headline?
See how Evidence Lab examines historical results: luck, unseen candles, walk-forward, trial history and stressed costs.
Explore Evidence LabEvidence gradeA-FGraded in the Evidence Lab
- Random entries
- Unseen candles
- Trial history
- Other exchanges
Interface example · each Original's stored scores appear on its member page.
From idea to Bot
01
Describe it
Describe your idea.
Pulse turns your description into a draft you can review, with assumptions shown in writing.
Long when a bullish engulfing candle forms and ADX is above 25, 4 hour chart
Example readoutGo long BTCUSDT on 4H when an Engulfing Bullish candle AND ADX(14) is above 25. Stop 2× ATR(14), take profit at 2R.
From the existing strategy examples · no performance result is shown
02
Test it
Test the draft.
Run the draft on historical candles. Interface charts on this page are examples, not strategy results.
Illustrative chart motion only · no market values or performance result is represented
03
Grade it
Review the evidence.
The Evidence Lab tests more than headline returns. A profitable backtest can still receive a low grade.
Evidence gradeA-FGraded in the Evidence Lab
- Random entries
- Unseen candles
- Trial history
- Other exchanges
Interface example · each Original's stored scores appear on its member page.
04
Run it
Choose how it runs.
Bot creation and execution follow your plan permissions. Supported Live automation depends on your strategy, plan and exchange connection. Backtesting and simulation do not guarantee future results. Live trading involves risk and can lose money.Learn about Bot automation →
Execution modeMember choice
PaperSimulated
LiveEligible Bots
Strategy example · review onlyBTCUSDT · 4HNo order is sent from this preview.
CHECK RULESUse a supported exchange connection for eligible Live automation. No member connection, trade or result is represented.
From idea to Bot
Scroll once to move to the next step
Long when a bullish engulfing candle forms and ADX is above 25, 4 hour chart
Example readoutGo long BTCUSDT on 4H when an Engulfing Bullish candle AND ADX(14) is above 25. Stop 2× ATR(14), take profit at 2R.
From the existing strategy examples · no performance result is shown
Illustrative chart motion only · no market values or performance result is represented
Evidence gradeA-FGraded in the Evidence Lab
- Random entries
- Unseen candles
- Trial history
- Other exchanges
Interface example · each Original's stored scores appear on its member page.
Execution modeMember choice
PaperSimulated
LiveEligible Bots
Strategy example · review onlyBTCUSDT · 4HNo order is sent from this preview.
CHECK RULESUse a supported exchange connection for eligible Live automation. No member connection, trade or result is represented.
The Evidence Lab
Is it overfitted? The Lab will say so.
A simulated backtest can look strong because it fit the history. Evidence Lab checks a saved version against random entries, unseen candles, a walk-forward where the data supports it, trial history and stressed costs. Then it gives a reasoned A–F robustness grade.
Explore the Evidence Lab →Chance
The Lab compares your strategy with random entries.
Unseen candles
The Lab tests on data held back until you reveal it.
Every trial counts
The Evidence Lab tracks how many versions you tested.
Repeatability
Choose the exchange and market for each run.
A reason for every point
Each grade comes with reasons, not a green light.
Free Quant Tools
Free tools for assumptions, failed tests and chance.
Four free tools in the app help you inspect trading costs, failed tests, overfitting and scenario paths. They do not promise a trading result.
Cost Drag Calculator
Vary fees, turnover, funding and slippage, then share the assumptions you used.
Indicator Graveyard
Review failed, inconclusive and not-enough-data outcomes separately.
Overfitting Explainer
Explore how selection bias can distort strategy testing.
Monte Carlo Sandbox
Run scenario checks while keeping the assumptions visible.
Open Quant Tools in the app Free tools · available from Quant Tools in the app
Bot Marketplace
Explore a Bot and its research.
Review the stored evidence, price and available copy modes before deciding what to inspect. A listing does not start a Bot for you.
Official groupFree
Pulse Signals Top 10
Free
Published signal streamJoin the groupReview each published setup and its recorded context. Not a prediction.
Open Pulse SignalsYour Home dashboard
More than a Bot builder. Your trading workspace.
Inspect Bots, review published setups and take part in the Edge Theory community from one dashboard.
01 / Bot picks
Featured Bots with their research.
Review the research behind the featured Bots. Bot picks need a current B-or-better grade and a positive held-out simulated return.
02 / Published setups
Published research in one view.
Review the latest published setup and its available context before deciding what, if anything, to do.
03 / Community
Join the conversation.
Share a trade, ask a question or bring an idea into the community feed.
Blog
Trading ideas, tested properly.
Read notes on market regimes, strategy tests, Bot ideas and the rules behind them.
We backtested Trading Rush's MACD + 200 EMA strategy on three years of Bitcoin
Trading Rush's Bitcoin test showed about a 62% win rate on 100 trades. We took his MACD + 200 EMA rules as published, added three ideas of our own and backtested all four versions on three years of Bitcoin, fees included. Here's where the money went.
What Is Backtest Overfitting? How to Spot It
Backtest overfitting means rules tuned so closely to past candles that they describe noise. Learn the warning signs and four checks to run on your own crypto strategy.
We backtested Data Trader's 4-hour range strategy on 4,634 trades
Data Trader showed seven trades and about a 72% win rate. We turned his 4-hour range strategy into rules and tested it on 4,634 trades over three years of Bitcoin and Ethereum, with fees, unseen data and random entries.
Your first step
Start with an idea. Put it on trial.
Explore Pulse, build a strategy and practise in simulation before deciding whether you need more.
Free membership
$0 / month
- 40 AI credits each month
- Up to 10 new AI designs
- Strategy Lab and historical backtesting
- 1 simulated Bot
- No card required
Questions before you start
Questions worth answering.
Can I use Edge Theory without coding?
Yes. Describe your trading idea to Pulse and review the editable strategy rules it creates.
Do I need real money to test a Bot?
No. Backtesting and simulated Bots do not place Live orders using real funds.
Can I start with a ready-made Bot?
Yes. Explore Edge Theory Originals and Marketplace strategies. Review each Bot’s research, price and access requirements before choosing.
Are Bots included in membership?
Membership sets platform features and capacity. Individual Bots may have separate charges or activation requirements, so check each Bot before choosing.
Can I try the platform for free?
Yes. Free membership includes 40 AI credits each month, Strategy Lab, historical backtesting and 1 simulated Bot.
Are the Quant Tools free?
Yes. Cost Drag Calculator, Indicator Graveyard, Overfitting Explainer and Monte Carlo Sandbox are free in Quant Tools. They help you inspect assumptions and failed tests; they do not promise a trading result.
Can a Bot lose money?
Yes. Historical and simulated results do not predict future performance. Live trading involves risk and can lose money.
Edge Theory
Ideas are the beginning. The trial is what matters.
Build a strategy. Explore a Bot. Examine the evidence. Find your place in Edge Theory.
Start free