Joshe

2026 US Midterms Shenanigans

288 posts in this topic

4 hours ago, Leo Gura said:

This is a known problem in science. It's called retrospective curve-fitting. You can pretty easily invent models that fit data retrospectively, fooling yourself that you have a predictive model, but you don't.

Just because you can retrodict does not mean you can predict. You can invent dozens of different models that successfully retrodict the past but still fail to predict the future.

So it's just epistemic slop.

Elections contain so many nebulous factors and causes that no simplistic model will capture them.

Elections are a nonlinear, chaotic phenomenon, similar to weather. Predicting them with formulas is a fool's errand. That's why Lichtman will keep failing.

I disagree. Some elections are predictable, some aren't. The tighter the margins, the harder to predict, and the tighter the margins, the higher the variance. 

The hard part isn't just knowing which variables are relevant, it's knowing when they matter and how much. It's a lot like poker. 

1. Identifying relevant variables. What information actually matters in this situation?

2. Assigning importance. Of the relevant variables, which are most consequential right now?

3. Reasoning accurately. Given those variables and their importance, what conclusion follows?

It's not easy, but I think it's solvable to an extent.


-

Share this post


Link to post
Share on other sites

 

Hakeem Jeffries is officially house speaker. It's nice to see Hasan Piker, Kyle Kulinski & far leftists crash out because of their excessive demonization & racist remarks of Jeffries. Also props to Jeffries calling out the far lefts bluff on Tuesday before the vote

Share this post


Link to post
Share on other sites
27 minutes ago, kguirnela said:

Hakeem Jeffries is officially house speaker.

He is a Zionist shill.


You are God. You are Truth. You are Love. You are Infinity.

Share this post


Link to post
Share on other sites
11 hours ago, Leo Gura said:

He is a Zionist shill.

There weren't any anti-AIPAC Democrats in the House who would've had the votes to become the House Speaker.

Share this post


Link to post
Share on other sites
12 hours ago, Hardkill said:

There weren't any anti-AIPAC Democrats in the House who would've had the votes to become the House Speaker.

Or hear me out, nobody wanted the job. Ayanna Pressley made it clear that she didn't want the job no matter how many times the far left begged her. And AOC made it clear that nobody wanted the job & had took time out of her schedule to educate people how voting for house speaker works & they still gave her shit. 

Share this post


Link to post
Share on other sites
On 11/18/2026 at 1:16 PM, Leo Gura said:

This is a known problem in science. It's called retrospective curve-fitting. You can pretty easily invent models that fit data retrospectively, fooling yourself that you have a predictive model, but you don't.

Just because you can retrodict does not mean you can predict. You can invent dozens of different models that successfully retrodict the past but still fail to predict the future.

So it's just epistemic slop.

Elections contain so many nebulous factors and causes that no simplistic model will capture them.

Elections are a nonlinear, chaotic phenomenon, similar to weather. Predicting them with formulas is a fool's errand. That's why Lichtman will keep failing.

I actually agree with you now that presidential elections involve too many unpredictable, complicated, and difficult-to-quantify factors to make highly reliable predictions, especially more than a month before an election. That's why I've learned not to put much stock in long-range election forecasts anymore.

However, I don't think it's entirely fair to dismiss Allan Lichtman's 13 Keys as merely retrospective curve-fitting. He developed the model back in 1981 and then used it prospectively for more than 40 years, essentially correctly predicting 10 out of 11 presidential elections from 1984 through 2024, depending on how you interpret the disputed elections of 2000 and 2016.

That includes the controversial 2000 election, when he predicted Al Gore would win. Lichtman subsequently conducted statistical research for the U.S. Commission on Civil Rights documenting substantial racial disparities in rejected ballots in Florida, lending support to his argument that Gore was the rightful winner, although the research didn't conclusively establish who would have won otherwise.

As far as I'm aware, nobody has demonstrated a clearly superior track record of prospectively predicting presidential elections over the same 40-year period.

Furthermore, scientific models, theories, and methods are routinely modified and refined for legitimate reasons as new evidence emerges. Even the practices and standards of the scientific method itself have evolved over time. Revising a model doesn't automatically invalidate it, provided the revisions are scientifically justified, transparent, and subsequently tested against new evidence rather than simply being used to explain away past failures.

That said, I'll absolutely give you credit for your 2024 presidential election prediction. You correctly anticipated Trump's victory, whereas Lichtman incorrectly predicted a Harris victory. I also appreciate your approach to understanding and anticipating political developments by drawing on epistemology, philosophy, history, empirical evidence, and other relevant factors rather than relying exclusively on rigid formulas. I think that kind of multidisciplinary reasoning, especially when combined with intellectual humility and an awareness of uncertainty, can be valuable for making informed assessments about the near future.

Obviously, Lichtman's failed 2024 prediction exposed limitations in his model, and I'm no longer convinced that any simplistic forecasting system can reliably account for all the complexities of modern elections.

But I still think there's a significant difference between acknowledging those limitations and dismissing more than four decades of remarkably successful prospective predictions as nothing more than retrospective curve-fitting.

Edited by Hardkill

Share this post


Link to post
Share on other sites
3 minutes ago, Hardkill said:

He developed the model back in 1981 and then used it prospectively for the next 40 years, essentially correctly predicting 10 out of 11 presidential elections from 1984 through 2024

That is a good point.


You are God. You are Truth. You are Love. You are Infinity.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now