Hacker News Reader

Comment by lastofthemojito | Hacker News Reader

I don't pretend to know whether Trump will win Wisconsin or not. I did recently read the NYT piece entitled The Relentless Shrinking of Trump’s Base [0], which at least seems somewhat data driven. Obviously, the characterization of "White voters without college degrees" as the Trump base is certainly a very rough approximation, but using their demographic groupings, they estimate that the Trump coalition shrunk by 48,000 since 2016, and the "Biden coalition" grew by 140,000. This in a state that Trump won by less than 23,000 votes.

[0] https://www.nytimes.com/interactive/2020/10/22/us/politics/t...

View on HN

I live in this state. I know these people better than the schmuck reporters. Just because the official numbers shrank doesn't mean the actual constituency did.

Do you also only think that crimes happen only if it gets reported? Meaning if 50 burglaries happen to 100 people but only 2 people reported it, does that mean the burglary rate is 2% or 50%? This is why I don't bother with these worthless metrics as I'll explain in the next paragraph regarding trump.

Did you ever wonder why a lot of people don't publicly come out as trump supporters? They have very liberal friends that they don't want to lose because they said they'd vote for trump. That sentiment still lingers here. Most leftist people who argue against trump vehemently do so in a way that these regular folk that end up voting for trump keep it under wraps because they're too scared to publicly say anything.

Just because the "numbers" shrank, doesn't me the actual voters did. There is a lot of societal pressure involving politics and only on the left is it worthy of breaking a friendship over. This is why many white liberals who voted trump keep their mouths shut.

I think I visited Wisconsin once on a day trip when I had some family in neighboring Illinois back in the 90's. I definitely don't know a ton about the state. Again, I certainly don't know which way Wisconsin will end up voting.

But I guess my point here is that it seems like the NYT feature was based on an honest analysis of data from something called IPUMS, which Wikipedia tells me is "the world's largest individual-level population database", not a just "schmuck reporter" with a baseless hunch or something.

It'll be interesting to revisit pieces like this on November 3 (or whenever all the votes are tallied) to see if any of them turned out to be good predictors or not.

> Do you also only think that crimes happen only if it gets reported? Meaning if 50 burglaries happen to 100 people but only 2 people reported it, does that mean the burglary rate is 2% or 50%? This is why I don't bother with these worthless metrics as I'll explain in the next paragraph regarding trump.

Based on this middle school level of understanding of polling and crime statistics, I'll skip the rest of the comment, but thanks.

Reply on news.ycombinator.com