# 3 Candidate Condorcet Simulations

donald at mich.com donald at mich.com
Tue Nov 26 03:05:45 PST 1996

```Dear Methods List,

I would like to thank DEMOREP1 for doing this "simulation" below:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
DEMOREP1 wrote: >I did 37 random simulations. Results-
>4 Circular ties, 33 Condorcet winners (beat each of the other two)
>------
>Average winning percentage of all winners- 55.8 (range was 50.1 to 69.1)
>------
>Voters voted for 1- 20.7 Pct
>Voters voted for 2 (2 subtypes)- 38.1 Pct
>Voters voted for 3 (2 subtypes)- 41.3 Pct
>With a random function, it would appear that each subtype would appear an
>equal percentage (i.e. 20 percent) of the time which is highly suspect for a
>real election.
- - - - - - - - - - - - - - - end of simulation - - - - - - - - - - - - - - -

Dear DEMOREP1,

The ratio of circular ties was something I was looking for.

Your random results of voters selecting one or two or three is very
acceptable. With three candidates there are three possible single
selections - six possible double selections - and six possible triple
selections. Which would yield 20-40-40 percentages.

But I agree that in a real election other forces would cause people to
vote differently.

What does your use of the word "subtype" mean? You have (2 subtypes)
in two lines - Voters voted for 2 (2 subtypes) - and Voters voted for 3 (2
subtypes).
What does that mean? I do not understand.

Thank you again,

Donald

```