Research education · Quantic Eagle operates only with its own capital. No client funds, investment advice or trading signals.
The test passes.
The evidence is missing.
Three experiments you can take apart with a click. See how a reassuring result can survive an incomplete check.
Synthetic counterexamples inspired by research failures. No trading performance is shown.
Same scores. Different decisions.
score = score / selection ≠ selection
The probabilities match. Selection chooses B instead of A. A score comparison passes; a decision comparison exposes the defect.
Every batch returns a response. The same objects appear twice.
8 results / 4 unique objects
Both requests receive the global A B C D context. The responses contain eight results for just four distinct objects.
The feature is a number. Its window is incomplete.
2 observations / 3 required
The current value is finite. The second session is missing from a three-observation window. A NaN check on the final value misses the gap.
The tests that must fail
A figure explains the defect. The planned fieldbook will include exercises to reproduce it, correct it and check that the test detects it.
The fieldbook is in preparation. The resources on this website are free to use. You can request one email when the fieldbook is ready; no purchase or reservation is available.
Notify me when it is readyWritten by Quantic Eagle LTD · General research education. About.