Unverified50% confidenceFactExact time
When a team tests 1,000 combinations of strategy parameters, even if all combinations have zero predictive power, roughly 50 combinations will appear to perform excellently in backtesting purely due to randomness (the multiple comparisons problem).
1
Sources
50%
Confidence
Long-term
Relevance
7/24/2026
First Seen
Sources
TradingSpy：本地优先开源AI交易助手，数据隐私不妥协
hackernewshackernews7/11/2026
Related Claims
Unverified在5%显著性水平下测试100个随机因子，期望会有5个看起来显著（多重假设检验问题）79% similarUnverified对于48支球队上百场比赛的赛制，5万次模拟通常可使冠军概率的标准误差控制在约0.2%以内75% similarUnverified一个随机猜测的模型在单次二元预测中有50%的概率押中72% similarUnverified置换检验通过大量随机置换（通常1000到10000次）构建零假设下的经验分布来估计p值，无需参数假设但计算成本相对较高72% similarUnverifiedThe project's 50 samples are at the proof-of-concept level, and accuracy may fluctuate in larger-scale real production environments.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/608875API
curl https://kongchang.com/api/v1/knowledge/claims/608875MCP
get_claim(id=608875)