[EM] Experimental data on chicken dilemma (post 1 of N: Forsythe 1996)

[Jameson Quinn]

oops, there were a few mistakes in what I sent, because I hit "send"
prematurely:

They publish their data in a singularly useless format, but here's what I
> see in the tables for approval:
> without polls:
> minority wins: 5/48
> minority in 2-way tie: 7/48
> minority in 3-way tie: 2/48
> Total minority expected wins: 19% (was 26% for approval)
>

I meant, 26% for plurality.

>
> with polls:
> MiW: 2/48
> Mi2T: 10/48
> Mi3T: 0
> TMiEW: 30% (17% for approval)
>

Again, I meant 17% for plurality. Also, for approval, the total by my count
of the data table is only 14.6%.

I got the 30% number because there is an EVIL error in this paper. It says
that for approval with polls, "the minority candidate achieved only 14.6
victories (out of 48 chances, counting two-way ties as 1/2-victories and
three-way ties as 1/3-victories)." Obviously, I realized that you can't get
14.6 by adding wholes, halfs, and thirds, but I assumed that they'd just
improperly rounded 14.667. But no, the 14.6 was a correctly rounded
percentage for 7/48.

I wonder if I should write to Forsythe and let him know...

Jameson
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20121230/07ca3787/attachment-0004.htm>