Predicting elections

Rafael A. Irizarry
October 24, 2014

538 uses: State fundamentals

  • The generic congressional ballot.

  • Congressional approval ratings

  • Fundraising totals.

  • Highest elected office held.

  • Margin of victory in most recent Senate election.

  • Candidate ideology and state partisanship

Details

This works by treating the state fundamentals estimate as equivalent to a “poll” with a weight of 0.35. What does that mean? Our poll weights are designed such that a 600-voter poll from a firm with an average pollster rating gets a weight of 1.00 (on the day of its release55; this weight will decline as the poll ages). Only the lowest-rated pollsters will have a weight as low as 0.35. So the state fundamentals estimate is treated as tantamount to a single bad (though recent) poll. This differs from the presidential model, where the state fundamentals estimate is more reliable and gets a considerably heavier weight.

538: Margin of error

  • Uncertainty is larger the when there are more days to go until the election. This means the model will become more confident as Election Day approaches.

  • Uncertainty is larger when there are fewer polls.

  • Uncertainty is larger when the polls disagree more with one another.

  • Uncertainty is larger when the polling average disagrees more with the state fundamentals.

  • Uncertainty is larger when there are more undecideds or third-party voters in the polls.

  • Uncertainty is larger when the race is more lopsided.

Heavy tails

FiveThirtyEight model does not quite use a normal distribution; instead it uses a transformation of the normal distribution with slightly fatter tails. The transformation gives extreme long-shot candidates slightly shorter odds; it might mean, for example, that we would have a candidate with a 0.5 percent chance to win his race instead of a 0.05 percent chance. But this process is not complicated and makes little difference. Instead, we run simulations to deal with a couple of more important problems.

538 removes polls

They take these polls out: Strategic Vision, Research 2000, Zogby,

Read in old polls

Source: local data frame [5 x 6]

  state year real.dem real.rep vote.dem vote.rep
1    AK 2008    47.77    46.52       47       41
2    AK 2008    47.77    46.52       45       46
3    AK 2008    47.77    46.52       51       44
4    AK 2008    47.77    46.52       47       45
5    AK 2008    47.77    46.52       48       43

Note: deal with independents

Data wrangling

Note this one appears flipped. So take out

  vote.rep2 vote.dem2 real.rep2 real.dem2
1        32        65     63.36     36.52

When to start counting

plot of chunk unnamed-chunk-5

So we keep 2 months

Evaluating pollsters

plot of chunk unnamed-chunk-7

Evaluating polls

plot of chunk unnamed-chunk-8

Evaluating polls

plot of chunk unnamed-chunk-9

Evaluating polls

plot of chunk unnamed-chunk-10

Aggregate

plot of chunk unnamed-chunk-11

Aggregate

plot of chunk unnamed-chunk-12

Aggregate

plot of chunk unnamed-chunk-13

Results

plot of chunk unnamed-chunk-15

Results by year

plot of chunk unnamed-chunk-16

Bayesian

plot of chunk unnamed-chunk-17