Thanks -- does the fact that the top 1% and bottom 99% show different distributions not imply they are gambling on different scales though? Is it as simple as the 1% gamble on a log scale and the bottom 99% gamble on a linear scale?