Reply to “Reproducibility in Computational Science: A Case Study: Randomness of the Digits of Pi” [Bailey et al. 17]

ABSTRACT

[Bailey et al. 17] pursue the question whether recently published results on the non-randomness of the decimal digits of pi [Ganz 14] can be replicated and whether the employed test statistic is sufficiently robust. While the authors are able to reproduce the original results by Ganz, they claim that their additional tests using different parameters than those employed by Ganz fail to replicate Ganz's phenomenon. The authors however fail to assess the mathematical implications of their additional parameter choices and, finally, to evaluate the distribution of their obtained results. By completing the statistical analysis of their results, one finds that the phenomenon claimed by Ganz can very well be replicated.

KEYWORDS:

• Pi
• statistical randomness

2010 AMS SUBJECT CLASSIFICATION:

• 11Y60
• 11K45
• 65–05