5 ms·

nice visuals, but these polls and simulations were wrong in 2016, and are likely wrong now.

edmundsauto 6y ago
Are you aware of the adjustments that were made by pollsters and meta-pollsters like 538 since 2018? Also, it's not like they were wrong in 2016 -- 538 had a 30% chance of a Trump victory. I expected better statistical understanding from this audience, but maybe that only underscores how difficult it is to get people thinking correctly about probabilities.
kilbuz 6y ago
What does wrong even mean in these contexts? You can't reliably evaluate a probability model with n = 1 test cases.
UncleOxidant 6y ago
The polls were right in 2016 in that Clinton did win ~3Million more votes than Trump. The problem was with polls in some states that ended up being crucial for an electoral college win. Hopefully 538 (and other similar organizations) have learned something from that and have made some adjustments.
- verylittlemeat 6y ago
  That seems outrageously obtuse. The electoral college is fundamental to predicting the US presidential election. The idea that 538 got it right but just didn't factor in the electoral college is a real stretch. If they screwed up polls in states that were crucial for an electoral college win then they screwed up full stop.
  - rsynnott 6y ago
    538 et al don't, as a general rule, run their own polls. In 2016, some swing states were poorly polled, so they were flying kind of blind on those. And, of course, a few tens of thousands of votes different on the day, below the margin of error of any poll, and no-one would be bothered about them. In a way, 538 can't win here; they said "there's a 30% chance of this thing happening", it happened, based on differences below the resolution that they could even see at, and everyone's annoyed with them.
    - dane-pgp 6y ago
      If differences below the resolution that you can see can cause the opposite outcome to happen, then perhaps an estimate of "30% chance of this thing happening" should be paired with a "1% chance of that estimate being correct".
  - rconti 6y ago
    Especially given the fact that "538" (electoral votes) is right there in their name.
  - colinmhayes 6y ago
    How can you possibly say they got it wrong? What evidence do you have that Trump didn't have a 30% chance of winning? Your sample size of 1?
iron0013 6y ago
How is it possible to miss the point this hard? The website was specifically designed to teach people not to make the mistake in understanding that you are making.
gdulli 6y ago
A poll is a fact. It's not wrong. A model that gave Clinton a 70% chance of winning is not proven wrong if she loses. We can't know if the model is right or wrong unless we run the election many times.
- bra-ket 6y ago
  Ok, here is a better word - "garbage", as in Garbage In, Garbage Out.

Are you aware of the adjustments that were made by pollsters and meta-pollsters like 538 since 2018? Also, it's not like they were wrong in 2016 -- 538 had a 30% chance of a Trump victory. I expected better statistical understanding from this audience, but maybe that only underscores how difficult it is to get people thinking correctly about probabilities.

kilbuz 6y ago

What does wrong even mean in these contexts? You can't reliably evaluate a probability model with n = 1 test cases.

UncleOxidant 6y ago

The polls were right in 2016 in that Clinton did win ~3Million more votes than Trump. The problem was with polls in some states that ended up being crucial for an electoral college win. Hopefully 538 (and other similar organizations) have learned something from that and have made some adjustments.

verylittlemeat 6y ago
That seems outrageously obtuse. The electoral college is fundamental to predicting the US presidential election. The idea that 538 got it right but just didn't factor in the electoral college is a real stretch. If they screwed up polls in states that were crucial for an electoral college win then they screwed up full stop.
- rsynnott 6y ago
  538 et al don't, as a general rule, run their own polls. In 2016, some swing states were poorly polled, so they were flying kind of blind on those. And, of course, a few tens of thousands of votes different on the day, below the margin of error of any poll, and no-one would be bothered about them. In a way, 538 can't win here; they said "there's a 30% chance of this thing happening", it happened, based on differences below the resolution that they could even see at, and everyone's annoyed with them.
  - dane-pgp 6y ago
    If differences below the resolution that you can see can cause the opposite outcome to happen, then perhaps an estimate of "30% chance of this thing happening" should be paired with a "1% chance of that estimate being correct".
- rconti 6y ago
  Especially given the fact that "538" (electoral votes) is right there in their name.
- colinmhayes 6y ago
  How can you possibly say they got it wrong? What evidence do you have that Trump didn't have a 30% chance of winning? Your sample size of 1?

iron0013 6y ago

How is it possible to miss the point this hard? The website was specifically designed to teach people not to make the mistake in understanding that you are making.

gdulli 6y ago

A poll is a fact. It's not wrong. A model that gave Clinton a 70% chance of winning is not proven wrong if she loses. We can't know if the model is right or wrong unless we run the election many times.

bra-ket 6y ago
Ok, here is a better word - "garbage", as in Garbage In, Garbage Out.