# [EM] General PR question (from Andy Jennings in 2011)

Toby Pereira tdp201b at yahoo.co.uk
Sun Oct 5 12:58:16 PDT 2014

```On Sun, Oct 5, 2014 at 2:02 AM, Toby Pereira <tdp201b at yahoo.co.uk> wrote:
> From: Kathy Dopp <kathy.dopp at gmail.com>
>
>
>> Well, it changes the overall proportion they should have, but it does not
>> change the 5:3 correct ratio of A to B seats.
>Yes. However, it changes the NUMBER of seats allocated in the 5:3
>ratio, allocating those seats in EXACTLY proportional fashion.  I.e.>For three seats:>5 voters 1.875 seats  gives 2 seats
>3 voters 1.125 seats  gives 1 seat>For four seats:>5 voters 2.5 seats - a tie for the forth seat gives 2 or 3 seats
>3 voters 1.5 seats  - a tie for the forth seat gives 2 or 1 seats>It's just SIMPLE arithmetic.>If you don't like it, then you are arguing AGAINST PROPORTIONAL
The example wasn't about three or four seats. It was about four seats both times and who was more deserving of the fourth seat out of A and B with or without C's inclusion. (It doesn't matter that you might argue that C is more deserving than either when present because we can still order non-winning allocations in terms of proportionality.)
>> I would argue that if [SNIPPED]

>Yes. I AGREE and so does my formula minimization algorithm.
OK - I should have added that it is specifically with factions 1 and 2 having the same voting patterns. I'll formulate it differently and give a specific example. So I would argue that if
Faction 1: x seatsFaction 2: y seatsFaction 3: z seats
is more proportional than
Faction 1: x-1seatsFaction 2: y+1 seatsFaction 3: z seats
then if the only change in votes concerns faction 3, this would mean that
Faction 1: x seatsFaction 2: y seatsFaction 3: z seats
is still more proportional than
Faction 1: x-1seatsFaction 2: y+1 seatsFaction 3: z seats
For example:
2 to elect
Faction 1: 302Faction 2: 100Faction 3: 1
The proportional allocations are:
Faction 1: 1.499Faction 2: 0.496Faction 3: 0.005
Faction 1 would win both seats. However, if we have:
Faction 1: 302Faction 2: 100Faction 3: 3
The proportional allocations are:
Faction 1: 1.491Faction 2: 0.494Faction 3: 0.015
This would make it one seat all between factions 1 and 2. This is why I would argue that there are better definitions of proportionality than your method.
Toby
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20141005/01a84633/attachment.htm>
```