(And now, if you have no interest in voting systems, or in me taking Twitter arguments over to LJ so I can have more space to have them, then feel free to page down.)
So, the Electoral Reform Society electoralreform put up this:
After changing to STV 76% of Scots got their first preference elected:Which seemed to me to be both true, and a direct effect of STV.
The reason for this is that multi-member constituencies reduce the effect of winner-takes-all voting systems. If you have three FPTP constituencies, all of which are 40% SNP, 30% Labour, 20% Lib-Dem, 10% Other, then the effect is to elect three SNP representatives. In a three-member constituency you'd (probably) end up with either 2 SNP and 1 Labour, or one each of SNP, Labour, and Lib-Dem, depending on how the votes split. This would would mean either 70% or 90% of people saw their first-choice elected.
Now, you can argue that "Having your first choice elected" isn't actually what we're trying to optimise for here. And, personally, it's not the top thing I'm trying to optimise for. But a lot of people care about having a representative that they feel represents them. Telling them that changing the voting system to STV will give them a greater chance of having a candidate elected who was their #1 choice will make them happier about it.
And you can also argue that if your prime metric is "Having your first choice elected" then there are ones more guaranteed to do that - changing FPTP from "Elect the highest ranked candidate from a single-choice vote" to "Elect the highest-three ranked candidates from a single-choice vote" would probably be better at that, for instance. But this assumes that I only have one metric I care about, and I don't. I care about giving people a representative they want representing them _and_ I care about giving those representatives a voice which is in proportion to the number of electors who supported them. Plus, of course, nobody is actually suggesting Nth-Past-The-Post as a voting system.
Different people prioritise different metrics. I care more about proportionality, and giving small parties a voice, for instance, which means I like AMS, which gives the Scottish Greens two seats on 4% of the vote. While Andrew Hickey (for example) cares more about not having party machines controlling the lists of who can be elected (and what order candidates are elected from them), and so favours STV. And I think that picking a metric that I don't care about, but some people do, and pointing out that a particular voting system has that metric, and thus some people who care about that metric might be interested in supporting it, is a good thing.
Original post on Dreamwidth - there are comments there.