PR-STV: query about use of my second pref

Geuze

Let's say I give a first pref for a strong candidate, likely to be eventually elected, or if not, then likely to stay in the count until the end.

So they are not eliminated in the first few counts, and if they are eliminated, it's at the last count.

They don't make the quota during first few counts, so no surplus to be distributed.

Am I correct in saying that my second and subsequent prefs may never be used?

If this candidate is eventually elected, say on the 3rd-6th count, it will be due to transfers from eliminated candidates.

So my vote will sit in their pile, and won't transfer.

Is that correct?

Godge

Geuze wrote: »

Let's say I give a first pref for a strong candidate, likely to be eventually elected, or if not, then likely to stay in the count until the end.

So they are not eliminated in the first few counts, and if they are eliminated, it's at the last count.

They don't make the quota during first few counts, so no surplus to be distributed.

Am I correct in saying that my second and subsequent prefs may never be used?

If this candidate is eventually elected, say on the 3rd-6th count, it will be due to transfers from eliminated candidates.

So my vote will sit in their pile, and won't transfer.

Is that correct?

Correct.

On two occasions I was able to criticise the government and opposition and say I had no hand, act or part in electing any of them because my vote was not used to elect anyone having been left in the pile of the last person eliminated.

Zubeneschamali

Geuze wrote: »

Am I correct in saying that my second and subsequent prefs may never be used?

To cover you question fully, there are 4 possibilities:

1) Your #1 candidate is elected on the last count: your 2nd preference does not matter, it is not counted. No worries, your #1 counted.

2) Your #1 candidate is the last one eliminated: your 2nd preference does not matter, it is not counted. You might as well have stayed at home.

3) Your #1 candidate is eliminated before the last count, your vote will definitely be distributed. Hooray.

4) Your #1 candidate is elected before the last count, then your vote may be distributed as part of their surplus, but may not. It's random.

Godge

Zubeneschamali wrote: »

4) Your #1 candidate is elected before the last count, then your vote may be distributed as part of their surplus, but may not. It's random.

It is not random.

Example:

Quota is 20,000.

Candidate A gets 19,000 votes on Count 1
Candidate B gets 3,000 votes and is eliminated.

Two-thirds of B's votes transfer to Candidate A in Count 2 and Candidate A is elected with a surplus of 1,000 votes. In the next count, distributing the surplus 1,000 votes of Candidate A, only the bundle transferred from Candidate B is examined for determining where the surplus goes.

Zubeneschamali

Godge wrote: »

It is not random.

Suppose A is elected with a surplus of 1000 on the first count. Which 1000 votes of his 21000 are distributed?

Suppose A is elected as in your example with 2000 transfers and 1000 surplus. Which 1000 of the 2000 transfers are distributed?

It's random.

In the North, they count the next preference on all 2000, and divide by 2 to transfer, but down here they just pick a random 1000.

Godge

Zubeneschamali wrote: »

Suppose A is elected with a surplus of 1000 on the first count. Which 1000 votes of his 21000 are distributed?

Suppose A is elected as in your example with 2000 transfers and 1000 surplus. Which 1000 of the 2000 transfers are distributed?

It's random.

In the North, they count the next preference on all 2000, and divide by 2 to transfer, but down here they just pick a random 1000.

But in the example I gave, it is a random 1,000 from the last bundle received which isn't a random 1,000 of all the votes.

Take the example of a FF candidate elected with a SF elimination and the two remaining candidates are FG and AAA. In distributing the FF surplus, FG would be more likely to get second preferences from a random 1,000 from all of the FF votes than from a random 1,000 from the SF first preference bundle.

In large rural constituencies the local factor would also come into play.

All that being said, I have no evidence that any of the close contests in recent years would have been changed by a different situation but given the way it is done, sooner or later we will have a situation like that.

Zubeneschamali

Godge wrote: »

it is a random 1,000 from the last bundle received which isn't a random 1,000 of all the votes.

Well of course it is - nobodies vote counts twice. Each vote lands with at most one candidate (or in the non-transferable bin). Your vote can't elect the FF guy and then transfer to FG as well.

plodder

Godge wrote: »

But in the example I gave, it is a random 1,000 from the last bundle received which isn't a random 1,000 of all the votes.

Take the example of a FF candidate elected with a SF elimination and the two remaining candidates are FG and AAA. In distributing the FF surplus, FG would be more likely to get second preferences from a random 1,000 from all of the FF votes than from a random 1,000 from the SF first preference bundle.

In large rural constituencies the local factor would also come into play.

All that being said, I have no evidence that any of the close contests in recent years would have been changed by a different situation but given the way it is done, sooner or later we will have a situation like that.

The way it actually works is that the 2,000 votes that were transferred to A are all divided according to the next preference on each. So, there is a separate pile of votes for each continuing candidate. Then they work out how many (of the 1,000 surplus) each of these is due to receive (from the relative size of each pile). They take that number of votes off the top of the pile for each candidate. So, technically, it's not random.

Though, it is unpredictable, because the votes are all mixed up at the start of the whole count. In that sense, you could say it is random.

Riskymove

Zubeneschamali wrote: »

Well of course it is - nobodies vote counts twice. Each vote lands with at most one candidate (or in the non-transferable bin). Your vote can't elect the FF guy and then transfer to FG as well.

exactly

Single
Transferable
Vote

jwwb

Searched on the forum for an answer. It may be there but I couldn't find it.

Scenario:
My 1st preference goes to candidate A who gets eliminated at the first count - vote transfers
My 2nd goes to candidate B who was elected on the 1st count - vote transfers
My 3rd goes to candidate C just short of quota, and as a result has a huge surplus, and gets elected, and their surplus is distributed.

Question: as my vote was "the last vote in" is my vote more or less likely to transfer than someone who voted for candidate C with their 1st preference. To phrase is another was - is the random surplus selection from all of candidate C's votes or just the surplus?

Thanks

seamus

jwwb wrote: »

Searched on the forum for an answer. It may be there but I couldn't find it.

Scenario:
My 1st preference goes to candidate A who gets eliminated at the first count - vote transfers
My 2nd goes to candidate B who was elected on the 1st count - vote transfers
My 3rd goes to candidate C just short of quota, and as a result has a huge surplus, and gets elected, and their surplus is distributed.

Question: as my vote was "the last vote in" is my vote more or less likely to transfer than someone who voted for candidate C with their 1st preference. To phrase is another was - is the random surplus selection from all of candidate C's votes or just the surplus?

Thanks

In the scenario you describe, the surplus is taken from the votes that were transferred after the last count, so your vote is more likely to be transferred.

So, for example, imagine the quota is 10,000

Count 1, the candidate gets 8,000 votes.
Count 2, the candidate receives 1,500 transfers. Total count 9,500.
Count 3, the candidate receives 1,000 transfers. Elected.

The surplus of 500 is only selected from the 1,000 votes transferred at the end of count 2. So if you vote was transferred then, it has a 50:50 chance of being redistributed in count 4.

plodder

jwwb wrote: »

Searched on the forum for an answer. It may be there but I couldn't find it.

Scenario:
My 1st preference goes to candidate A who gets eliminated at the first count - vote transfers
My 2nd goes to candidate B who was elected on the 1st count - vote transfers
My 3rd goes to candidate C just short of quota, and as a result has a huge surplus, and gets elected, and their surplus is distributed.

Question: as my vote was "the last vote in" is my vote more or less likely to transfer than someone who voted for candidate C with their 1st preference. To phrase is another was - is the random surplus selection from all of candidate C's votes or just the surplus?

Thanks

There is a good chance your vote will transfer. It depends where in the pile your ballot is, and that would be very hard to predict.

There is zero chance that the other vote for candidate C on 1st preference will transfer, because it wasn't in the last parcel of votes that C received.

Faugheen

Before people get shirty, I'm not saying this is 100 percent right, but I'll do my best.

I'll use a theoretical example of Carlow-Kilkenny. Quota of 8000. John McGuinness gets 10,000 first preference votes. His votes are then sorted by second preferences. Each pile is then counted, then percentages of the those preferences are represented in the surplus

So, let's say 40 percent of the second preferences of McGuinness are going to Bobby Aylward, so 40 percent of the surplus will go to Aylward, so Bobby Aylward will receive 800 transfers from McGuinness' surplus

I'm still reading into it, but I'm not sure how it's factored if a candidate is elected through transfers in terms of distribution of surplus.

Hope that clears one part of the query up at least.