This paper investigates forecasting accuracy of fuzzy extended group decisions for adjustment of statistical benchmark results. Delphi is a widely-used method for implementing accurate group consensus decisions. However, the concept of consensus is subject to expert characteristics, and sometimes it is ensured by a facilitator judgment. Fuzzy set approach is adapted for Delphi, named fuzzy-Delphi (FD) by several studies, and the present paper extends the recent literature by an implementation of the FD for adjustments of statistical results. The empirical results indicate the improvement of statistical predictions after the FD based judgmental adjustment.