Trumponder: Rigging

I already wrote a lot about voter fraud before the election. To summarise:

  1. In person voting fraud in the US is rare and ineffectual, carries stiff penalties and would be a waste of time for an organised campaign.
  2. Fraud around absentee ballots is more common than in-person fraud but still not huge and affects mainly smaller elections. Notably supposed anti-fraud measures proposed by Republicans address in-person fraud rather than absentee ballot fraud.
  3. Voter suppression techniques are a much bigger issue.

How big an impact did voter suppression measures have on the election? I doubt there will ever be a clear answer because it is hard to track why somebody didn’t vote. However, Clinton clearly suffered from a reduced turnout and voter suppression measures aim to do that AND the election was close. Americans have reason to feel cheated but the sad fact is that disenchantment with electoral processes can also drive lower turnout, so even Democrats feeling cheated helps the Republicans. Well there is a depressing thought.

Now I don’t have a decent link for the next bit because it was described to me verbally and non-committaly but I’ll sketch something out as a strawman for me to knock down.

Maybe, what-if, perhaps the GOP cheated in some other way than openly via voter suppression. Short answer is all the reasons why it is implausible that the Democrats cheat en-mass:

  1. It is hard for many reasons.
  2. The risks are high.
  3. The rewards are uncertain.
  4. There are better things to do with your time if you are a campaign.

Ok, ok, but maybe, maybe things were different with the Trump campaign because

  1. It was desperate.
  2. It didn’t give a shit about the reputation of the GOP or the long term damage getting caught cheating would cause.
  3. It didn’t have a ground game and maybe cheating is a better use of resources when you are badly organised.
  4. And look – it would explain why the polls were off and why the stories about high turnout didn’t pan out!
  5. Also Trump always projects – he said the Dems would cheat ergo he was planning to cheat!

Yeah but, none of that is actual empirical evidence of cheating.

  1. The polls being simply wrong is a SIMPLER explanation than what would have to be a complex conspiracy. The high probabilities assigned to a Clinton win by most forecasters didn’t account for a possibility of systematic polling failure – mainly because they couldn’t put a number on that. Notably Nate Silver’s figures (wrong but lower) did attempt to account for it. Importantly, even a poll that captures accurately the split in preferences, will have a very hard time capturing who will actually vote. In an election where turnout was key, predicting WHO would vote is a bigger issue than who they would vote for. p(polling error)>p(complex conspiracy)
  2. Yes, Trump projects but we already know how the GOP intended to rig the election because they did it quite openly and at least quasi-legally: voter suppression. It isn’t like they made much of a secret about it, how it would work or what the intended effect would be.
  3. The kind of fraud that would have needed to have taken place would involve electoral officials hiding/not-counting/destroying some votes from particular precincts. OK, that is a theoretical way somebody could fraudulently create the kind of voting impact we saw but it would require the GOP to have suborned multiple officials in multiple states that use different voting systems and voting technology and NOBODY on the Democrat side watching the voting NOTICING and NOBODY giving the game away and NOBODY slipping up in anyway.

In short, it just isn’t plausible and amounts to a kind of Deus Ex-Machina hope of a terrible plot twist twisting back towards a happier resolution. I also doubt it would be a happy resolution – discovering whole chunks of uncounted votes from Democrat leaning precincts would (not unreasonably) be seen by Republican supporters as evidence of Democrat cheating.

The other argument is perhaps more metaphysical and hence beyond empirical considerations:

  1. It’s 2016 man and it ain’t done messing with us yet.

OK spooky-2016 granted, I’ll change my mind in the event of something that looks like something more substantial than wishful thinking.




So What is Going On in Vox Land Just Prior to Election Day?

Um. Unfortunately, having resorted to describing the position on Vox Day’s websites re the US election as gibbering nonsense, I’m not bereft of words to describe the level of nonsense.

In the final days, the conspiracy theories have run off the rails and are making weird bubbling noises in a swamp. Essentially, the current claims are that the hacked emails do reveal terrible crimes but ONLY if you read them using special code. Read the code correctly and a world of demonology is revealed. Defeatist talk has been banned in a manner that sounds like Vox is trying to script his own Downfall parody of himself

BUT do not get cocky. Yes, this all indicates a group of people desperately trying to rationalize what looks like a looming defeat. It is manifestly IS desperate clutching at straws. However, remember that Vox and the alt-right have no special insight into the facts on the ground – indeed their level of knowledge of what the actual state of play might be is negative because of how they engage with facts. Just because they are acting like they are losing does not mean that they are losing.

In short – if you are eligible to vote in the US Presidential election and haven’t already done so, then please do so. Remember this actually the nice timeline – the other one has weasel flu. Optimism is a weapon but only if it motivates action.

A Sail Around the Puppy Seas to Exemplify the Polls Using Appallingly Sized Samples

No, but seriously looking in depth at what individuals are saying is a totally legit research technique.

To recap, in previous posts I’ve looked at the Sad Puppy/Rabid Puppy political split in terms of the right’s anti-Trump vote and the alt-right’s pro-Trump vote. Simply, the difference between notable Sad Puppies (as defined by the 2015 Hugo Campaigns) and the Rabid Puppies was a simple as the Rabid Puppies being strongly pro-Trump.

How do things stand now? The Rabid Puppy camps remain pro-Trump obviously: racism and misogyny are seen as a feature rather than a bug of the Trump candidacy.

On the Sad Puppy side, Peter Grant the Tor-boycott guy has been leaning Trump for awhile.

Brad Torgersen and Larry Correia, remain in the plague-on-both-your-houses camp.Brad has been making digs about people supporting Hillary though.

The big change is Sarah Hoyt, who has announced that she intends voting for Trump:

I think her reasoning is interesting – not terribly coherent and based on a looming dread of leftists doing lefty things but still interesting. I say that because it helps show what may be happening in various ways with the polls (at least in part). People with various kinds of right (or anti-left) views finally caving under the weight of a freakish electoral cycle and picking the only side they can pick.

The Right’s War on Statistics

‘Zero Hedge’ is in a flap about poll ‘oversampling’ here


It even includes a hack email from John Podesta which discusses ways of ensuring that the Democrats own polling over samples minority groups. Again, gasp!

Except. Well over sampling a smaller demographic group is the right thing to do. When I say ‘right’, I don’t mean for opinion polls but for collecting statistics on a population in general.

Say you have a representative sample of a population consisting of a thousand people. Now, of that thousand people you are particularly interested in a sub group that represents 1% of the population. If your sample is exactly proportionate, then it should have 10 people belonging to that sub group. Unfortunately 10 is a shitty sample, if you are unlucky to get 2 odd people with unusual views they then form 20% of your sub-sample.

Sample size is a dark art but the easiest issue to understand is it that magnitude matters. A good sample size is less about the proportion of the whole population in your sample and more about the raw number of people. More is better, but ‘more’ is subject to diminishing returns.

Over sampling means you can get a better picture of the sub group. However, because you end up with more of group X than you should have, their response are then weighted proportionally when looked at the statistics overall.

Are polls manipulated! Well, if by ‘manipulated’ you mean ‘use statistics’ then yes.

Weird Internet Ideas: The Voting Dead

Are dead people voting in the US elections? No. At least not in any significant numbers.

Former mayor of New York, Rudy Guilliani is reported as saying “I’m sorry, dead people generally vote for Democrats rather than Republicans,”

Oh, if only I had some way of checking this using sources that Republicans would regard as reliable! Oh, wait, I do! Back to the Heritage Foundation ‘database’ and let’s see how many dead people have been voting.

Oh, wait, I do! Back to the Heritage Foundation ‘database’ and let’s see how many dead people have been voting…

Not many, is the answer, about 30 at most in a database that mainly covers this century and a wide variety of election. The biggest example was from the New York Police, as a demonstration that it was possible as part of an operation – I haven’t counted that as an actual case of dead people voting. I have included the voter registration examples that came up but there is no actual indication that these dead people voted. Most of the clear cases of the dead voting are cases of somebody voting on behalf of a recently deceased near relatives (some quite touchingly). The funniest one was a case of identity theft in which the guy also registered as a voter as part and parcel of getting a drivers license.

Here are all the cases I could find. The number at the start represents the number of actual dead people that may have voted.