[EM] Consistent ratio of round-off point differences

[MIKE OSSIPOFF]

The factor by which the round-off points of Hill differ from those of 
Bias-Free tends to be very nearly twice the factor by which the round-off 
points of Webster differ from those of Bias-Free. That ratio seems to keep 
getting closer to exactly 2, from above, as the size of the two consecutive 
integers increases. If I remember correctly, it's about 2.01 between 1 and 
2, and it's about 2.00001 between 10 and 11.

If it's necessary to propose a simple, familiar, precedented apportionment 
method, then people should be asked if they prefer Hlamilton's completely 
unbiased randomness or Webster's very slight tendency to favor large states.

The latter would probably sound worse to residents of small states, but the 
former would probably cause more anger in its operation.

Mike Ossipoff

_________________________________________________________________
Get the latest Windows Live Messenger 8.1 Beta version. Join now. 
http://ideas.live.com