[EM] Rank Codes
Forest Simmons
fsimmons at pcc.edu
Fri Feb 12 18:29:29 PST 2021
It pains me to see all of the ranked ballot proposals that unnecessarily
limit preferences to three or four alternatives because of ignorance of
simple higher resolution ballots that can be easily marked and read (by
hand or by machine) without ambiguity or confusion from poorly formed
characters, stray marks, etc.
A method that allows only three or four candidates to be ranked cannot
satisfy clone independence ... the only indispensable justification for
scrapping First Past the Post Plurality. And (beyond that) it exacerbates
the biggest IRV/STV/RCV defect, the high likelihood that one's choices will
be completely exhausted before the final rounds unless you rank lesser
evils at the expense of alternatives you like better, because of ranking
limitations that highlight the effect of premature eliminations.
It is alleged that because of ambiguous handwriting and lack of room for
more than a few "bubbles," only a handful of distinct ranks can be allowed.
But what if each bubble has a different value?:
[8] [4] [2] [1]
The rank of a candidate is the sum of its darkened bubble values ... a
number between zero and fifteen.
Suppose that there are to be 26 candidates, then instead of indicating
their relative ranks with mere numbers, you can order them with standard
alpha numeric code words ... Alpha1, Bravo2, Charlie3, Delta4, Echo5,
Foxtrot6, ... Victor22, Whiskey23, Xray24, Yankee25, Zulu26. So the
military already solved the ambiguity/ "noisy channel" commuunication
problem in the early days of Morse code.
These 26 code words cannot be confused with each other no matter how
illegible the hand writing.
If 625 alphabetically ordered code words are needed, there are that many
easily distinguished five-letter words that satisfy the following rules:
The 1st letter of each word must be a member of the set {A, F, L, Q}.
The 2nd letter must be from {B, G, M, T}.
The 3rd from {C, H, N, W}, the 4th from {D, J, P, Y}, and the last from {E,
K, S, Z}.
The important thing is that each of the five sets consists of four letters
that cannot be confused among themselves. Futhermore letters in the same
position come from different quarters of the alphabet, making alphabetical
order easier to discern.
The ballots are to be accompanied by an easily accessible table of code
words numbered in alphabetical order. However, the voters can skip numbers
that they don't need when there are more code words than candidates in the
race. Similarly, it goes without saying that the same code word can be
applied to more than one alternative when equal rankings are allowed. And
of course, these code words can be adapted for high resolution ratings if
needed in the Range/Score/Cardinal Ratings context.
These suggestions are intended for absentee and other mail-in ballots ...
electronic voting machines should allow in person voters to drag the names
into a list in any order, and then print out paper copies for voter and
precinct receipts.
I am sure there are better ways of doing this, but then why do we still
keep seeing proposals with unnecessarily crippling limitations on the
number of distinct ranks?
Comments? Suggestions?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20210212/caee19eb/attachment.html>
More information about the Election-Methods
mailing list