A few myths:

1 Preference deals don’t apply in Hare Clark, they only apply to its sister system the Senate voting system (another variant of the STV family).
2 Those who claim they “pick up” Hare Clark right away are usually wrong, whether they realise or not. eg: They would not know in which cases you redistribute the last bundle assigned to a candidate and in which cases you redistribute from the first bundle.
3 Not all votes are fractionalised – this is only the case where a vote starts with a candidate who topped quota.
4 The graphs are easy to follow. Don’t claim you can follow the system if you cannot follow the graphs.

The up side of the Hare-Clark is that we don’t have to worry about preference deals between the parties. Nothing shits me more than voting for my prefered candidate and finding out later that they had a preference deal with someone I particularly dispise!
In the ACT, I can direct my preferences where I want them, even if they are worth 1/3 of a vote (thanks Sepi).

jakez said :

Sepi has it. I picked up the system pretty much straight away so it’s hard for me to help people with their problems, however I’ve noticed a lot of people don’t know how it works.

It has its structural flaws but every system does. I would like to know what the knockers would prefer.

I’ve lied most of my life in Victoria, and though I know all systems have flaws this Hare-Clark system has my mind boggling. The time it takes to work out, the way candidates with a tiny proportion of votes can win, and the instability that seems to be a consequence of it are all things that make me wince. But this is my first election here in the ACT, so maybe with time I’ll get used to it…?

Jakez, I would be surprised if you picked anything up quickly.

Thanks Jimbocool, my only concern is that the average punter is not interested politics and I’m pretty sure the average voter has no idea which political party is pushing which policy, sure they may have a general idea what the parties stand for, but if you went down the street now and asked the first 10 people what they thought about the Greens transport policy? You would get a blank look and then maybe something along the lines of they want more buses; that would be the extent of the detail. The more complicated the voting system the more likely people will disengage further from the political system. Further example of this is the number of people not enrolled to vote, it is amazing!

The various discussions on this topic have certainly helped me grasp the system better, though I’m with Sepi – I don’t follow the tables on the Elections ACT website at all.

Sepi has it. I picked up the system pretty much straight away so it’s hard for me to help people with their problems, however I’ve noticed a lot of people don’t know how it works.

It has its structural flaws but every system does. I would like to know what the knockers would prefer.

Redistribution of preferences is hard to understand, because they are distributed as tiny percentages of a vote, not as whole votes.

So – if a quota is 100 votes, and Shane Rattenbury gets 150 votes, there are 50 votes worth of votes to redistribute. (Not just 50 votes, cos they go back to everyone’s vote to see their second preference. TKs Jimbo for that!)

SO – of the 150 votes, 1/3 is excess, not needed to get Shane in, so each second preference is now worth 1/3 of a vote.

So if a 1/3 of Shane’s second preferences went to Green Caroline L, she gets 1/3 of 50 votes added to her existing total.

All the other second preferences in Shane’s pile of votes also get valued at 1/3 of a vote and passed on to the name second on that ballot paper.

Clear as mud hey!

I am starting to vaguely understand the theory of the system. I can’t follow the numbers in the graphs to save myself though.

So it’s magic then? I just can’t seeing in being that big a difference considering the small transfer values of anyone who isn’t Zed or Katy.

Areaman, the short answer is that the Green-destined preferences travel more efficiently the longer the Greens stay in the count. The effect is worth about 1,000 votes by my calculation so significant. The long answer involves half a page of handwritten calculations which would be tedious to reproduce here.

Granny & Arty – the confusion isn’t helped by the media picking up the daily fluctuations in the count and breathlessly reporting that ‘so and so is now in trouble’- Simon Corbell is not in any trouble. Nothing in the last couple of days changes the most likely outcome from being 7,7,3. The race for the final seat in Molonglo is very tight, though so there is a chance of 7,6,4.

caf: the basics are easy enough to understand, it’s the redistribution of preferences that doesn’t seem to be clearly explained anywhere… I’ve discussed this with mates many times this week, trying to grasp how these votes can be used a second time.

Anyone care to explain step 1.5?

Yeah … no … I’m with Arty.

It’s not really. Basically, to get elected you need to get a quota of votes in your pile. In Molonglo, a quota is 12.5% of valid votes cast.

You start by putting votes into candidates piles based on the first preference. You then:

1. Declare anyone with more than a quota in their pile elected, and distribute their excess votes to the next-preferenced candidates (repeat as necessary);

2. Knock out the candidate with the lowest number of votes in their pile, and distribute those to the piles of the next-preferenced candidates, then go back to step 1.

The major complication is that in step 1, you have to redistribute *all* the votes, but their value is diminished because they’ve already used up some of their “electing power” getting that candidate elected. But that’s the basics.

For a voting system that is supposed to be the most democratic it is bloody hard to understand!

Repost from the last thread:

Jimbo, I’m still not really understanding why you’re concerned about when the third green gets knocked out. Is it just controlling leakage within the greens? Surely any Pangello preference that was going to one of the greens will go to to any other still in.