Low, but not zero (hooray statistics). The larger the sample size, though, the more it will tend towards average-- so if there were lots of people with 7-7-7- and a few with 1-1-1 then it would indeed seem to indicate a higher (20%) chance, but when it's skewed the other way that tends to suggest that the 7-7-7 is the outlier.