Battle of the cringe, part ?: The belligerent blabbermouth vs the pretentious professor

 So, apparently Allen Lichtman and Cenk Uygur went on Piers Morgan and got into a nasty argument. Basically, Cenk dared insult Lichtman's keys and Lichtman started this stupid "DO YOU KNOW WHO I AM" type thing where he acts like he's the smartest person in the world and can never be wrong.

I'm going to be blunt, unlike my normal battles of the cringe, this one has a clear winner for me, it's Cenk. The only reason I called it a battle of cringe is because both people tend to have cringey dispositions sometimes, and fit certain archetypes that I find kind of offputting, but of the two....yeah I prefer Cenk here.

Like, I actually like Cenk, I just understand he goes off the rails sometimes and starts screaming people down and going on rants, and he does that here. often times when he does this on TYT you can see Ana Kasperian try to reason with him only to give up, roll her eyes, and let him go off, and it can be cringey, but here it was a treat to see.

Because let's face it. Allen Lichtman...is a hack. He's a cope peddler who's been wrong all election cycle, and he's been selling copium to resistance liberals who wanted to ignore polls. he leaned into ignoring the polls and the data, and as we know now, but not only were the polls MOSTLY correct (about as correct as they typically are, but if anything, despite liberals complaining of herding and pro trump bias, they actually UNDERESTIMATED republicans. As such, as Cenk put it, Mr. Lichtman needs a nice tall glass of "shut up juice" over here. He was wrong, he's a pretentious hack, and he needs to shut up and sit down and let the adults in the room talk. 

And I'll say this. Cenk, despite whatever disagreements I have with him at times, was right all along. He was warning the democrats that Biden was gonna lose and even ran a quite cringey and fruitless primary campaign against him that even I struggled to take seriously (mostly due to his lack of eligibility to be president due to the natural born citizen requirement). And you know what? He was right. He tapped into the vibes, and even I knew it. I dont like Biden myself and subjectively, my gut was on board with him.

I just understood the political reality of the situation and that replacing Biden was somewhat problematic, and we didnt even have polling data suggesting an alternative would do better, if anything Harris's initial odds were roughly half of Biden's at any given time, and Gavin Newsom was so far down in swing states that he probably WOULD have lost to Trump with Trump getting 400 electoral votes. 

Now, as far as Kyle Kulinski goes, since he talked about owning our failures, yeah. Okay, let's talk about that. 

Here's my final prediction as per my model:

I had Trump winning. I had him at a 54% chance, with Harris at a 46% chance, so my model got the right outcome, but was wrong on two states, both of which were tossups. That's....pretty good. About on par. I got the right outcome, but I got Michigan and Wisconsin wrong.

Now, I kind of knew my actual prediction would be about as...correct as it was. 48 states with me getting two tossups or a tossup and a lean wrong is pretty average. I've tested my model going back to 2004, and yeah, all of my predictions, minus 2016, were roughly this accurate vs the final results.

As such, I made a second personal prediction about polling error. I knew that we were gonna see one side overperform and another underperform. I figured Harris would overperform and we would get Pennsylvania and Nevada, and something akin to this:

I was wrong, and because the polling error went the OTHER way, we ended up with something like this instead:

Oops. And now that we have the full vote totals, that's actually pretty accurate as far as how the swing states went and how the "likely blue" wall went. Of course, the "likely red" states were catastrophically off, with up to 14 point margins out of states I expected like 6-9 out of, but yeah...I estimated a polling error about one point in Harris's favor since I had reason to believe that the polls were lowballing Harris...and it turns out they actually lowballed Trump.

Am I update by this? Does it hurt my pride? No. If anything it just tells me to trust my model and injecting my personal ideas of what would happen doubled my error rate. This is why I made two separate projections, one based on the model and one based on what I thought would happen. The model gave the median outcome, but that median outcome never actually comes up, we normally see it go in one direction or another, often by 1-4 points. It went Trump's direction. Oops. My bad. Oh well, I had a 50-50 shot and I called it wrong. Not a big deal. I take the L on that one. 

Not like the outcome was that out of the range of expectations. Remember me cherrypicking the simulations I thought would be most likely? Yeah, we got the exact Trump blowout scenario I expected. 226-312. It was actually a pretty likely outcome to pop up in the case of a Trump overperformance, as anything from R+1 to R+4 would get similar electoral college results.

So yeah. Was my model relatively accurate? It was on par with its historical performance. Was my personal prediction accurate? No, I bet the wrong way. It happens. I bet heads and got tails. Oh well. 

And that's the thing. Lichtman doesnt deal in probabilities. Lichtman deals in this set in stone model of "this is who is going to win and I'm almost never wrong". Even though his model has been called into question several times before. And thats why he was a snake oil salesman. Because his model was out in line with the mood of the country and expectations. It ignored various factors like dems suppressing a genuine primary challenge. His economic indicators dont account for inflation. His indicators dont account for voter sentiments at all. He just has a bunch of things generally correlated with economic success and he misread the situation. 

I'm going to be honest. I didnt know who would actually win. It was a coin flip. Nate Silver said it was a coin flip. Allen Lichtman ripped on nate silver for saying it was a coin flip because polling isnt accurate. Well, polling is basically simulating the election by gathering voter opinions. They're imperfect simulations, but when you gather enough data, it gives you an idea. And polling is rarely ever so far off that it's flat out wrong. We had a range of possible outcomes that could have happened, the actual outcome was not just well within the range of possible outcomes, but actually one of the most likely outcomes we could've gotten by sheer probability (seriously anything from the 16th percentile to 45th percentile would have amounted to 226-312, it had a 29% chance of happening overall give or take). And yeah, the polls werent really wrong. They were off by like 1-2 points and sometimes more in some states, but nothing that was shockingly unexpected.

So yeah, Allen, go take Cenk's glass of "shut up juice" go in the corner, and let the adults in the room talk. You werent just wrong, you were disastrously wrong. You were calling it for Biden even when my own model was like "yeah he has a 13% chance". Seriously, we had this map for Biden, and he still was like "Biden's gonna win". 

That isn't just wrong, that's literal malpractice. You screwed up SO BAD that you were ignoring the data giving us stuff like this because your keys said blah blah blah. Your model was wrong. It was laughably wrong, and I can be "oh okay" with someone who wrong and had since admitted it and moved on, but you're literally one of the most insufferable people of the 2024 election cycle, and you're STILL insufferable in defeat, so yeah, SHUT UP.