Thu Oct 13, 2016, 05:48 AM

RBInMaine (13,570 posts)

How soon before Allan Lichtman changes his prediction?

The problem from the very start of Lichtman's pro Trump prediction, as he actually said, is that this year's Republican candidate is so "unusual" so it could throw off the historical trends upon which his predictive formula is based.

Remember that. Because I think that is just what we are seeing before our very eyes right now. This is why he gave himself this caveat this year. Make no mistake. Lichtman knew, and had already been clearly on record, saying that his model may be thrown off in this cycle because of the "unusual nature" of Trump. Well, Trump is certainly unusual. He is a confirmed sexual deviant. It doesn't get that much more unusual.

Also, I don't think Gary Johnson in the end will pull 5% but if he does, I think it will be more from registered Republicans than Democrats. It will be moderate Republicans.

But even if Johnson does get at least 5%, the thing that is really sinking Lichtman's model this year is that his model, the 13 keys, is based on a TRADITIONAL opposition candidacy, and Trump doesn't qualify as a traditional candidacy. So this will be interesting.

Meantime, get out there and vote early if you can, GOTV, write letters to the editor, donate to Hillary, and do anything you can to sink Trump and elect Hillary. We have to continue to charge hard to the finish line.

11 replies, 1842 views

Thread info Bookmark this thread Trash this thread

Reply to this thread

Always highlight: 10 newest replies | Replies posted after I mark a forum

Replies to this discussion thread

11 replies	Author	Time	Post
How soon before Allan Lichtman changes his prediction? (Original post)	RBInMaine	Oct 2016	OP
I think so too	Ardoewaan	Oct 2016	#1
Mostly agree, but they do need to also keep attacking Trump as an "unfit" candidate. Every campaign	RBInMaine	Oct 2016	#2
You're right	Ardoewaan	Oct 2016	#3
The candidate herself needs to do both the attack message and the positive message.	RBInMaine	Oct 2016	#8
i think johnson won't get 5%, and the model will have correctly predicted hillary.	unblock	Oct 2016	#4
He will not change his prediction	Cicada	Oct 2016	#5
Several of his categories are subjective and not empirical	bluestateguy	Oct 2016	#6
I'm very surprise that he says the Iran Nuclear Agreement doesn't count as a success	bigdarryl	Oct 2016	#10
But he left himself TWO outs: If Johnson gets below 5% or Trump being so unusual. The model might	RBInMaine	Oct 2016	#7
And will it discredit him if he is wrong?	RBInMaine	Oct 2016	#9
Not with me. A good model doesn't have to be perfect.	Cicada	Oct 2016	#11

Response to RBInMaine (Original post)

Thu Oct 13, 2016, 05:56 AM

Ardoewaan (144 posts)

1. I think so too

The model is based on rational agents. Rational candidates try to widen their audience during the general election, while Trump just throws out more red meat for his own voter base.

The polls do indicate that this is a losing strategy. I think Hillary just has to keep up the pressure, let Trump be Trump and go high.

Reply to this post

Response to Ardoewaan (Reply #1)

Thu Oct 13, 2016, 06:06 AM

RBInMaine (13,570 posts)

2. Mostly agree, but they do need to also keep attacking Trump as an "unfit" candidate. Every campaign

needs both an affirmative message and a negative message. The narrative is that Trump, on every measure, is "unfit" for the job. Americans will not elect someone that they are not comfortable with in the Oval Office. They need to keep reminding everyone that Trump is an unqualified and unfit person and could not be imagined in the office of the presidency.

Reply to this post

Response to RBInMaine (Reply #2)

Thu Oct 13, 2016, 06:55 AM

Ardoewaan (144 posts)

3. You're right

But at this point I think that Hillary can concentrate on getting her own message across, and let others (campaign staff) attack Trump himself. That being said, I'm not a paid political strategist

Reply to this post

Response to Ardoewaan (Reply #3)

Thu Oct 13, 2016, 07:17 PM

RBInMaine (13,570 posts)

8. The candidate herself needs to do both the attack message and the positive message.

Reply to this post

Response to RBInMaine (Original post)

Thu Oct 13, 2016, 08:31 AM

unblock (51,287 posts)

4. i think johnson won't get 5%, and the model will have correctly predicted hillary.

basically, lichtman just went into a panic because he felt pressured to make the call, and instead of saying it all depended on johnson breaking 5% or not, he called it for trump and hedged in a way that cast doubt on his entire model instead of sticking to his guns.

he handled it very poorly. now even if the model turns out to have been correct all along, he's done damage to it by saying that any new election might break it. that's always true, in a sense, but that's not how you promote a model! just say the model's prediction, and if it breaks, then you talk about and maybe fix it afterwards.

Reply to this post

Response to RBInMaine (Original post)

Thu Oct 13, 2016, 08:54 AM

Cicada (4,533 posts)

5. He will not change his prediction

His model makes the prediction, not Allan. I'm sure Allan knows Trump will lose but changing the prediction of the model would discredit Lichtman big time.

Reply to this post

Response to Cicada (Reply #5)

Thu Oct 13, 2016, 09:00 AM

bluestateguy (44,173 posts)

6. Several of his categories are subjective and not empirical

For example, the category about the candidate being charismatic, or the categories about the president's party having a "major" domestic or foreign policy achievement. Those criteria have never been sharply defined.

Reply to this post

Response to bluestateguy (Reply #6)

Thu Oct 13, 2016, 07:26 PM

bigdarryl (13,190 posts)

10. I'm very surprise that he says the Iran Nuclear Agreement doesn't count as a success

That's questionable

Reply to this post

Response to Cicada (Reply #5)

Thu Oct 13, 2016, 07:16 PM

RBInMaine (13,570 posts)

7. But he left himself TWO outs: If Johnson gets below 5% or Trump being so unusual. The model might

make the prediction, but he created and interprets the model. There is subjectivity with it.

Reply to this post

Response to Cicada (Reply #5)

Thu Oct 13, 2016, 07:18 PM

RBInMaine (13,570 posts)

9. And will it discredit him if he is wrong?

Reply to this post

Response to RBInMaine (Reply #9)

Sat Oct 15, 2016, 06:18 PM

Cicada (4,533 posts)

11. Not with me. A good model doesn't have to be perfect.

But if you ignore the predictions of the model to get the correct result? That would be game over. So I predict Allan will stick with predicting Trump will win according to his model.

Reply to this post

2016 Postmortem (Forum)