EDGE
Make the edge defend itself.
Why do you believe this strategy has an edge? Why do you believe it does not? Is the edge broad or concentrated? What would change your conclusion?
How to interrogate your strategy
Do not just read the report. Make EdgeQuery defend it.
Your report is a starting point. Challenge the reasoning, ask what evidence supports a conclusion, test another explanation, and keep investigating until you understand what is actually driving the results.
You do not need to know the perfect question. The Research Agent should help identify the most useful unresolved questions from the evidence already collected.
Challenge the conclusion
If something does not make sense, push back.
A finding should be explainable. EdgeQuery should be able to show what evidence led to the conclusion, what remains uncertain, and what evidence would change its mind.
Example: “I don't understand why you believe my strategy does not have an edge.”
A useful answer should not simply repeat “the edge is weak.” It should identify the evidence behind that conclusion: whether expectancy is unstable, profits depend on a small group of trades, performance disappears outside one period or direction, drawdown overwhelms the return, or another specific weakness is present.
Then ask the next question: What evidence would make you change that conclusion?
Questions worth asking
Use these as starting points, not a script.
Beginners can use these questions to learn how to investigate a strategy. Experienced developers and quants can skip them and go directly to their own hypotheses.
WINNERS VS LOSERS
Find what separates them.
What conditions were present in my best trades that were missing in the failures? Are the same conditions producing both large winners and large losses?
ENTRY VS EXIT
Question where the problem starts.
Could the problem be my exit rather than my entry? Are losing trades bad immediately, or do good trades become bad because the strategy gives back too much?
MARKET STATE
Ask what changed around the trade.
Does the strategy behave differently under trend, chop, volatility, GEX, market structure, order flow, or another condition supported by the available data?
DRAWDOWN
Decide whether the damage is normal.
Is this drawdown abnormal, or has the strategy behaved this way before? Did the market change, did the strategy's trade behavior change, or both?
CHALLENGE
Try to break the explanation.
What is the strongest alternative explanation? What survives after we test it? Which finding are you least confident about, and why?
Keep digging intelligently
The Agent should help you ask better questions.
After a report or answer, BackTest Forensics may surface a small number of evidence-specific questions worth investigating next. These should come from unresolved findings in your strategy—not from a generic list designed to keep you clicking.
Questions worth investigating next
Why do you believe this edge is concentrated rather than broad?
What would change your conclusion?
Are my winning trades occurring under the same condition you are questioning?
Could the problem be my exit rather than my entry?
What happens if we separate the strategy by the market condition the evidence points to?
Is this drawdown abnormal, or has the strategy behaved this way before?
Pick the unresolved question that matters most and keep digging. The research history remains tied to the strategy so the next investigation can build on what has already been established.