[maemo-community] STV voting strategy

From: Dave Neary dneary at maemo.org
Date: Fri Mar 20 13:26:21 EET 2009

Frantisek Dufka wrote:
> OK so how exactly surplus gets transferred? I thought it is random like 
> described in
> http://stv.sourceforge.net/votingmethods/random so it is not clear which 
>   vote is considered as surplus and thus which #2 position vote counts. 
> As I understood it maybe my vote for #1 was not considered as surplus of 
> winning person so my #2 position does not count at all. Correct or not?

Yes and no.

Ballots in random transfer are considered to be pieces of paper - they
cannot be split.

What happens in random transfer STV is this:

Jaffa has 48 first round votes, the quota is 27.

We count all of Jaffa's 2nd preferences:
Tim: 18
Kees: 7
sjgadsby: 6
qole: 6
Jamie: 5
Andrea: 3
anidel: 3

We multiply the number of 2nd preferences by 21/48, and round the result
to give the number of votes which will be transferred:

Tim: 8
Kees: 3
sjgadsby: 3
qole: 3
Jamie: 2
Andrea: 1
anidel: 1

The question is *which* 8 of the 18 votes that were Jaffa 1, Tim 2 (or
Jaffa 1, GA 2, Tim 3, since GA is no longer being considered) get
transferred to Tim's pile.

This has an incidence if Tim ends up with a surplus, or gets eliminated
later, since in that case, we move down to the next preference on the
ballot, and if there's a ballot that has 1. Jaffa, 2. Tim, 3. sjgadsby,
and another that has 1. Jaffa, 2. Tim, 3. Kees, then depending on which
of the two ballots gets added to Tim's pile, Kees or Stephen end up with
the vote later.

There is another way to count, which only makes sense with electronic
counting (and is thus harder to verify), and that's fractional transfers.

When calculating how many ballots to transfer, we look at the 2nd
preferences as before, but now distribute exactly the number of ballots
that you get from the multiplication by 21/48. If, later, someone is
eliminated and they have had some ballots transferred to them from
Jaffa's surplus, all of the future transfers are weighted according to
the probability that the full ballot would have transferred.

An example mighht help to explain:

Three ballots are 1. Jaffa, 2. Jamie, 3. Kees | 1. Jaffa, 2. Jamie, 3.
sjgadsby | 1. Jamie, 2. Kees

Now, the first two ballots will transfer to Jamie 21/48 ballots each,
for a total transfer of 42/48 ballots.

Later, Jamie is eliminated. The third ballot transfers in its entirity
to Kees, Kees will get +1 ballot from it in the following round. The
other two, thoughh, split evenly among Kees and sjgadsby, resulting in
21/48 ballots each for them in the following round.

We could use fractional transfer if you want - and it could even be
retroactive to this election.

To ensure we weren't running the risk of controversy, I ran the election
under all variants of STV available, with different methods of
calculating thresholds and transfers. I can confirm that under all
scenarios I tried, the results remained the same.

> I didn't mean you should vote for other than your #1. I just think that 
> with ranking method the next positions have bigger effect so it really 
> doesn't matter much if you put someone on first or second place (or rank 
> them as 100 or 90 on 0-100 scale) when council has 5 members. With STV 
> it is very significant since your one vote is counted really only for 
> one person (and most likely the first one on the list if the candidate 
> is popular).

Yes, this is true. In STV, you are voting primarily to elect your
preferred candidate and ordering is very important. We do not run the
election with a pairwise run-off system like Condorcet (which isn't
suitable to multi-seat constituencies, and which is very hard to


maemo.org docsmaster
Email: dneary at maemo.org
Jabber: bolsh at jabber.org

More information about the maemo-community mailing list