Discussion about this post

User's avatar
Ed Y.'s avatar

Prior to cancelling my paid sub, I’d like to share my thoughts despite the potential “just leave” comments that will inevitably come.

I’ve consistently been vocal on here that Trump was comfortably in the lead, that the mainstream polls were once again wrong, that Seltzer’s poll was going to miss horribly, and that Nate’s model is flawed. Because of that, I was ridiculed and called names, which is unbecoming in this type of a newsletter.

The reason I’m sharing my thoughts is the earnest hope that it will lead to future improvements in the Silver Bulletin.

For, despite being a critic of Nate’s, I’ve been following him for quite a while, enjoying the past 538 articles on politics, sports and society.

I’ve been a paid sub and even purchased his book.

All this to lay the groundwork and show I’m not a troll.

Other groundwork to lay is my background. I share this to blunt whatever stereotypes critics will make of anyone who is a Trump supporter, which I am one of.

Raised in Canada, lived in Southern California for 25 years. Bachelor’s in Poli Sci and Psych. MBA. Worked in the corporate world. For the past 20 years, founded and run a humble small business. Taiwanese-American.

I had high hopes for the Silver Bulletin, but as time went on and we approached Election Day, it was clear to me that what I was reading wasn’t what I was hoping for.

All the signs were there for a massive Trump win. Yes, Nate acknowledged the possibility as much, but his algorithm really didn’t as shown by his 50/50 result after tens of thousands of simulations.

The fatal flaw, from my humble opinion, is his reliance on polls. If the polls are wrong, his model will be wrong. I realize he incorporates other factors, but the primary driver are the polls.

And this cycle, again, the “right wing” polls were all extremely accurate.

The news media and university polls missed…some by quite a bit. Again.

But honestly, you didn’t need these polls to know Trump was a heavy favorite.

All the key indicators pointed to Trump having the massive advantage:

-Wrong track was at 28%. As Harry Enten pointed out, no president wins with those numbers.

-Party ID by Pew and Gallup showed a net swing from D+4 to R+3. That’s 7 points. Party ID has a strong correlation with who wins.

-Economy / cost of living, and immigration were consistently the top issues, which Harris ran behind on.

-Harris simply would never do the same or better than Biden in the Rust Belt. Scranton Joe vs far left California Kamala. That alone should have disqualified her from consideration to run.

-Harris could never detach herself from the horrible numbers of Biden. She was trying to convince voters she will bring change, but could never answer the fundamental question of why she didn’t enact change as VP.

Then as the polls rolled out, we consistently saw Trump overperforming his 2020 numbers by 4 to 10 points.

We saw these in swing states but also in blue states.

We saw him tie the national average, which signaled a massive EC victory.

We saw him up with almost every demo: blacks, hispanics, Catholics, Jews, Arabs/Muslims, union workers, young men.

Add up all the above, and it was clear Trump should have been a 2-1 favorite at least, which coincidentally was about what the betting markets trended towards in the final days.

Yet despite all of the above, those on the left were oblivious and kept trotting out “advantages” Harris had which we now know were simply non-existent. Yet many on the left here, on blue X, commenters on Daily Kos or Slate, were convinced Harris was trending to a massive EC victory with states like FL and TX going blue.

Ann Seltzer’s historically disastrous poll that missed by an astonishing 16 points only added to their false bravado.

Because of the above, I saw an arbitrage opportunity to profit, and placed multiple bets on Kalshi.

I went 5 for 5 betting on: Trump to win, GOP to sweep Presidency / House / Senate, Trump to win the popular vote, Trump to win about 312, McCormick to win the Senate.

Those profitable bets were helped by data from Rich Baris at Big Data Poll, host of Inside the Numbers, and Robert Barnes, professional political and sports better. The info they provided were far more accurate and dispassionate than Nate’s. Eg. They didn’t give hopium that Trump would win states like VA, NH, etc.

Contrast that to Nate’s insistence that FL could NEVER go +8 for Trump, in his infamous scuffle with Keith Rabois. Nate is lucky Keith didn’t take him up on his $100,000 wager. Because not only did Trump win by 8, he won by a staggering 13.2%, flipping Miami-Dade solid red, exceeding the 10-12 that Keith predicted.

Yet there’s no acknowledgement of being wrong from Nate, on this or other aspects.

To end my much-too-long diatribe, I’m leaving some suggestions that could help improve the Bulletin:

-Take a long, hard look at how you weight polls. I kept saying, and no one had a good retort, that it was disastrous to give a poll like Quinnipiac an overweight of 1.36. Quinnipiac missed 2020 by an average of 7, and 100% in favor of Dems. They should have been disqualified from that alone, but instead Nate actually values them higher. This cycle, Quinnipiac was right only 40% of the time, worse than a coin flip.

Same goes for polls like Washington Post, which was right only 33% of the time.

Conversely, the right wing polls were right:

Atlas Intel 100%

Big Data Poll 90%

PollFair 86%

Rasmussen 86%

Quantus Insights 86%

Trafalgar 71%

Insider Advantage 71%

Left wing pundits like Litchmann, Sabato, etc were also disastrously wrong.

-I hate the 13 keys as well, but there’s something to the meta-themes that polls can’t capture. Again, I know Nate tries to incorporate some of this, but perhaps look to reduce the importance of polls and increase the importance of other factors.

-Dissect where his left wing lens may have blinded him to what was right in front of his eyes. None of us are non-partisan, but being too partisan can make us see anything we want in the data.

For those commenters who ridiculed me, called me nasty names, etc., I don’t really hold any animus. But I do hope this gave you a dose of humility, to learn that sometimes when you’re in a bubble, you can’t see that you’re in a bubble.

Despite cancelling my paid sub, I’ll still follow Nate, and I do wish him well. There’s absolutely a place for a newsletter like his that can add value to the public discourse.

Expand full comment
Thoughts About Stuff's avatar

Question for Nate: If Trump does not in fact declare himself God-Emperor for life, will you do a public mea culpa for all your alarmism about “our sacred democracy!” over the years?

Expand full comment
773 more comments...

No posts