Full analysis is selected. Choose Quick brief to scan the key points first.
WFN QUICK BRIEF · UNDER 1 MIN
The short version
MODE 2 OF 2
What the source establishes
NIST describes Bonferroni’s method for controlling a family of comparisons by allocating error probability across the specified set. The key point is that the collection of tests matters, rather than only the one result eventually highlighted. NIST/SEMATECH — Bonferroni multiple-comparison method ↗
What the distinction changes
An advertiser can make a backtest look impressive by showing only the best combination of indicator, expiry and date range. The omitted trials are part of the research process. Without them, the reader cannot tell whether the result was specified in advance or selected because it looked strongest.
Questions for the next review
Record all candidate rules and selection criteria before evaluating a final holdout. Preserve unsuccessful tests in the research log. Describe a selected historical result as such, and avoid treating a nominal significance level as a probability that a trading strategy will make money.
Quick Brief is a reading aid. The full article and source trail stay on this page.
WFN STORY COMPASS
Choose your reading angle
Move between the reported record, market context and open questions. Status labels keep claims distinct from findings.
All context layers are visible. Choose a lens to jump to that part of the story.
EDITORIAL ILLUSTRATIONWFN EDITION · 15 OCTOBER 2026
Editorial noteSource-based explanation with original WFN analysis. Historical records retain their original periods. Numerical examples are labelled as illustrative, not observed market data. AI-generated conceptual illustration; it does not document an actual event or product interface.
WFN REALITY LEDGER
What is on the record?
Primary-source explainer
Primary-source explainer; examples are explicitly illustrative
Primary record / reference periodPublished statistical methodology; no performance dataset asserted
The cited material establishes the documented method or historical record. WFN analysis and illustrative calculations are labelled separately.
NIST’s multiple-comparison framework explains why selecting the best result after many tests can make chance look like a discovery.
01 · VERIFIED RECORD
What the source establishes
NIST describes Bonferroni’s method for controlling a family of comparisons by allocating error probability across the specified set. The key point is that the collection of tests matters, rather than only the one result eventually highlighted. NIST/SEMATECH — Bonferroni multiple-comparison method ↗
For an explicitly illustrative set of twenty independent null tests, each with a 5% false-positive probability, the chance of at least one false positive is 1 minus 0.95 to the twentieth power, approximately 64.2%. Independence is part of this example; correlated tests require their own analysis.
02 · WFN EDITORIAL ANALYSIS
What the distinction changes
An advertiser can make a backtest look impressive by showing only the best combination of indicator, expiry and date range. The omitted trials are part of the research process. Without them, the reader cannot tell whether the result was specified in advance or selected because it looked strongest.
Multiple-testing adjustment also cannot repair every weakness. A test with leaked future information remains defective even after an error-rate correction. The selection process and the information timeline require separate checks.
03 · WHAT TO CHECK NEXT
Questions for the next review
Record all candidate rules and selection criteria before evaluating a final holdout. Preserve unsuccessful tests in the research log. Describe a selected historical result as such, and avoid treating a nominal significance level as a probability that a trading strategy will make money.
