2318

submitted 6 months ago by fossilesque@mander.xyz to c/science_memes@mander.xyz

184 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[-] TropicalDingdong@lemmy.world 1 points 6 months ago* (last edited 6 months ago)

You can project whatever narrative you want into the data but what is I'm saying is fundamentally the case.

Trump outperforms his polling. He did so in 2016 by a wide margin, and he did so again in 2020. You can just go look at the week prior polling. This isn't some grandiose fiction it's a statement of fact, that you seem to be ignorant to.

Your interest in a particular narrative doesn't change what is. What matters is that Biden needs around an 8% lead on Trump nationally to be secure, and has been trailing, basically the entire time.

If the election were tomorrow, and we believe the offsets observed in the two previous national elections, and we should because those were real events made from real data, then Biden would lose in a landslide today.

Because I can't stand all of your group think naivete:

I went and pulled the 2020 data. The above is the relative error in polling from polls during the months of October and November 2020, calculated against the real votes cast in 2020. Biden underperforms his polling by about 4% and Trump overperforms his polling by about 8%. You can argue with why this is the case, but you can-not pretend that this isn't the case. You should be adjusting how you see polls with this in mind. When you see Biden trailing Trump in national polling (and he has been for 400 days in a row), you should see that as a CLEAR Trump lead considering that Trump CONSISTENTLY overperforms on election day relative to his polling.

Sources: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/42MVDX

https://electionlab.mit.edu/data

https://projects.fivethirtyeight.com/polls/data/president_polls_historical.csv

[-] KevonLooney@lemm.ee 0 points 6 months ago

You didn't answer the question:

Do you know anyone who is voting for RFK Jr? He is polling at 10% right now, so if it's real then statistically you should know someone.

[-] LibertyLizard@slrpnk.net 4 points 6 months ago

It’s almost certainly true the RFK’s support will decline as we get closer to the election. This is a common trend with third party candidates.

However, it’s not totally clear which candidate those voters will choose. My sense is that RFK is not particularly popular but is a stand-in for the rejection of both candidates at the moment. However, most of these voters will switch over as the reality of RFK’s loss becomes more real. Often, last minute voting decisions will be based on the conditions and media narratives happening immediately around the time of the election. So the implications of what you’re saying on who will win are not certain currently.

[-] KevonLooney@lemm.ee 2 points 6 months ago

Yes, that's right. Basically you can't rely on polls this far from the election.

The question says "if the election were held today, who would you vote for". But the election isn't today, and the person answering the question knows that. So you see more people answer with third party candidates then would actually vote that way.

Not only that, but candidates' GOTV efforts do not happen until the election is approaching soon. That's what actually wins elections, not polls from 6 months before.

[-] Instigate@aussie.zone 1 points 6 months ago

That’s… not how statistics work there, friend. If there’s a 10% chance of something happening per person and I have ten people in a room, that doesn’t guarantee that one of them will have the thing happen. In fact, my sample could have 10/10 happenings or absolutely nothing happen and the statistic value would stay the same, because it’s an average of the entire population.

Trying to apply anecdotal evidence to statistics and then calling the statistic false when it doesn’t align with your anecdote is kinda doing things arse-backwards.