Shannon entropy of "QXZ" Hello everyone. I am trying to determine the Shannon entropy of the string of letters QXZ, taking into consideration those letters' frequency in English. I am using the formula: H(P) = –Ʃ pilog2pi What's puzzling me is that I am expecting to calculate a high entropy, since QXZ represents an unexpected string in the context of English letter frequencies -- but the first pi term in the formula, which takes very small values (e.g., .0008606 for Q), is greatly diminishing my calculations. I am obviously making a wrong assumption here or applying something incorrectly, because as I understand it, letters with high surprisals should increase the entropy of the string, not reduce it. Thank you in advance for your generous help.