John Van Pelt wrote: > What is the basis of your computation? Just the predefined Quackle > superleaves? or some other algorithm? > > I find it hard to believe a lot of these, ABERRS to cite one. surely > BARES, or even SEAR, is better than doubled R's. > > -John VP > I recently wrote a program that calculates bingo probability for any given stem.
ABERRS bingos with 14 different non-blank tiles representing 64.1% of the remainder of the bag. ABERS bingos with 114 non-blank 2-letter combinations with an aggregate draw probability of 44.5%. From that standpoint, ABERRS is indeed better. I am using this program to guide my study. By studying 5-letter leaves in probability order you first get a sense of what leaves are good, and you also become able to convert them. Remember that Quackle's heuristics are based on knowing the dictionary cold. AERST bingos 68.4% of the time... provided that you know what the underlying words are.
