[FRIAM] high turnout and tight races?

[uǝlƃ ↙↙↙]

My incompetence is overwhelming. But I have found a couple of peer-reviewed articles that reflect Roger's point about reduction from high-dimension to low-dimension [⛧], which doesn't imply Nick's misrepresentation of what Roger said. But it does imply that high turnout elections will be less *predictable*. And by being less predictable, you're left with estimating the distribution of closeness of the outcome.

So, if there are more points in the outcome space that are "not close", then "not close" is the higher probability.

But I *think* the causal links between turnout and closeness are swamped by other things like anti-incumbent bias, lesser of 2 evils, and the partisan effect Gary raises.


[⛧] This is the primary thing I think about when considering instant runoff, the algorithm for reduction. But if/when we made such a change, this whole area of statistics about elections might explode into ruin. My Conservative homunculus objects! How can we mine old research if you suddenly change the reduction algorithm? Screw the inhuman consequences of our current system. Better the devil you know.

On 10/28/20 7:43 PM, Gary Schiltz wrote:
> Arghhh, I meant to write that I’ve warmed to the idea of mandatory voting, not warned. The perils of typing on a phone. 
> 
> On Wed, Oct 28, 2020 at 9:40 PM Gary Schiltz <gary at naturesvisualarts.com <mailto:gary at naturesvisualarts.com>> wrote:
> 
>     I think that Republicans make up way less than half of the population, but the people who are traditionally less inclined to vote, would vote Democrat if they voted.  This is responsible for Republicans lately getting very slightly over half the votes cast.  I further believe that most of the newly motivated voters come from this heretofore underrepresented left leaning group, hence the likelihood of the Democratic margin of victory being larger this year. Whether that implies a correlation between turnout and margin of victory, I will leave it to the statisticians to argue over. 
> 
>     Incidentally, I have warned to the idea of mandatory voting, as we have here in Ecuador, and I believe in Australia as well. Without a “papel de votación”, i.e. a certificate that proves that you voted in the previous election, you have to pay a fine in order to get government services such ad drivers licenses. 
> 
>     On Wed, Oct 28, 2020 at 8:16 PM Roger Critchlow <rec at elf.org <mailto:rec at elf.org>> wrote:
> 
>         I would think that the more people who vote the less likely a tie or close outcome becomes, simply by the larger number of ways you can miss with more votes in play.

-- 
↙↙↙ uǝlƃ