Sep 11, 2021, 11:10:40 PM9/11/21
I have taken quite a while to get my thoughts together about Mr. Hammond’s DECWER (Declarer weighted error rate), DDOLAR (Double dummy opening lead accuracy rate), and DEFWER (Defensive weighted error rate) home-grown, currently non-standard, bridge playing measurements.
1. They are too complicated to comprehend, and especially to verify.
2. They rely on a false premise: A binary scored result can be derived from 13 possible initial choices for DDOLAR, and one and two less initial choices for the other two. Furthermore, how those binary scores are derived is not explained in his terminology explanations.
3. They are expressed in percentages only, and therefore have insignificant, to no, statistical significance.