Talk:Allan Lichtman

Can't have it both ways

He was wrong in predicting the President in 2000, as he choose Gore.

However, if his model simply chooses the winner of the popular vote, then he was wrong by choosing Trump in 2016. So, which year was he wrong: 2000 or 2016? — Preceding unsigned comment added by 2602:304:CE5F:2CE0:9C06:1CC:3EF8:1D84 (talk) 05:52, 23 November 2016 (UTC)

This is the place to discuss well-sourced article content. Instead of asking others to do the research, you will be a more effective contributor if you research the matter, find WP:RS references and propose any edits you feel reflect the sources within Wikipedia's editing policies and guidelines. SPECIFICO talk 15:46, 23 November 2016 (UTC)

I've since corrected the article to properly reflect this editor's concern. It would appear that Litchman, in his desire to be "right" has indeed contradicted himself. While Litchman excused his previous error of 2000 by pointing out the discrepancy between the popular vote and the electoral vote, he should not be permitted to then reverse this same logic in the 2016 election, and still claim that he was accurate in both elections. He cannot "have his cake and eat it too."

Still, I think that it is fair to give Litchman credit for the 2016 election and to disqualify his prediction on the 2000 election, because Litchman does not appear to be consistently referring to popular vote results, but to instead be predicting electoral outcomes. As requested above, I have now inserted an appropriate cite regarding Litchman's erroneous prediction for the 2000 election. Thanks, Scott P. (talk) 16:03, 3 December 2016 (UTC)

No, he's been pretty consistent in that his model is only for the the popular vote. His model retrospectively claimed Tilden and Cleveland should have won in 1876 and 1888, respectively, and I recall in his book on the 13 keys, he flat out said that the Electoral College was irrelevant. I've updated with the relevant source. Somebody Who Exists (talk) 01:27, 9 December 2016 (UTC)

I think it's inconsistent to fault Lichtman on both the 2000 and 2016 presidential races. If his system predicts the popular vote, then he was correct in 2000 and incorrect in 2016. But if his system predicts the electoral vote, then he was incorrect in 2000 and correct in 2016.

From there it becomes the task of determining whether his system is supposed to predict the popular or electoral vote. Since his books are called "The Keys to the White House," it seems to be indicating that his system is designed to predict who will end up in the White House, i.e., who will win the electoral college. After all, a model that predicts the popular vote but doesn't predict a winner seems futile and pointless. So, I believe the intro to this article should indicate that he has predicted the winner for every presidential election since 1984 except 2000 (with no mention of popular vs electoral vote).

That being said, Lichtman's possible inconsistency as to whether his model predicts the popular vs the electoral vote is a worthwhile inclusion in the article, but should be addressed in greater detail in the main body of the article. Walkingf00l (talk) 16:02, 16 August 2024 (UTC)

He’s publicly calling out this page for misquoting him and proporting false information-are we sure all linked sources are entirely accurate beyond a shadow of doubt? 180.200.201.84 (talk) 23:15, 30 August 2024 (UTC)

I don't see how his model can be considered valid for the electoral vote. He doesn't consider differences between states in his analysisJsnyder527 (talk) 00:02, 22 August 2020 (UTC).

He's still teaching and seems very approachable; why don't you write him at his university email address and ask? – Aboudaqn (talk) 19:58, 24 August 2020 (UTC)

We should settle this point, as there's been a lot of dispute about this recently on the 13 Keys page. Here are several sources regarding this discrepancy:

1. This source is the one most cited, but I cannot get access.
2. His 2016 book (preview on Amazon says popular vote on page 2). A citation to his 2016 book here would also be consistent with this inconsistency, there is a quote in it: “predict only the national popular vote and not the vote within individual states.”
3. Not that a Reddit thread is a source, but it explains this more coherently: here
4. Another article explaining the discrepancy: here.

His books from the 1990s do say popular vote only, so I am inclined to count 2000 but not count 2016, consistent with the sources. Anyone with access to that document or with insight, it would help Caraturane (talk) 02:05, 24 June 2024 (UTC)

Someone was able to get a link to the document in question over on Talk:The Keys to the White House, they strongly support the fact he was using the popular vote in 2016. Caraturane (talk) 21:41, 25 June 2024 (UTC)

I do not have a reliable source yet , but I remember him saying that after 2000 he revised his model to consider the electoral vote instead of the popular vote. I will continue to look for a source, before editing the article. 09:04, 25 July 2024 (UTC) RevDan (talk) 09:04, 25 July 2024 (UTC)

He revised it after 2016. At least, that's the first time he communicated it officially. Tomcleontis (talk) 13:08, 25 July 2024 (UTC)

Inaccurate Intro

The intro makes this false statement: It says he "accurately predicted the winner of every U.S. presidential election since 1984, with the exception of 2000 … and 2016, where he predicted Donald Trump would win, despite Trump's popular vote loss." Of course, Trump did win, despite his popular vote loss. After looking up the prediction, I'm changing it to this: "… and 2016, where he predicted Donald Trump would win the popular vote." (At least I think that's what Lichtman predicted. When he makes his prediction, he's not clear if he's predicting a popular vote victory or an electoral victory. They're usually the same thing, but his two failures happened when the winner lost the popular vote. MiguelMunoz (talk) 00:22, 9 July 2024 (UTC)

I updated this a bit to say he correctly predicted the "outcome" as I think that's a more useful word here since he changed what it was predicting (this lets him get due credit for 2000 but not for 2016, in line with the sources). Tomcleontis (talk) 13:09, 25 July 2024 (UTC)

Lichtman pushed for edit

Just wanted to highlight that Lichtman specifically called on his fans to edit this page in his Youtube livestream tonight, which looks like it corresponds with edits made regarding his 2000 and 2016 predictions. Not sure how this sort of thing is usually handled but seems like it should be reversed. — Preceding unsigned comment added by AmeriMike (talk • contribs) 02:17, 16 August 2024 (UTC)

This is a conflict-of-interest edit and raises other policy violations. I have reverted these edits and informed Wiki admins. Caraturane (talk) 03:20, 16 August 2024 (UTC)

Protection has been requested. Knitsey (talk) 12:20, 16 August 2024 (UTC)

It's not a conflict of interest. Lichtman was correct. He predicted Trump would win the 2016 election and that's exactly what happened. Jimv1983 (talk) 18:16, 31 August 2024 (UTC)

Regardless of whether Lichtman was correct (which was discussed at length months ago, and included the Wikipedia community paying to access his October 2016 paper mentioned by recent reporting and explanation on several platforms, which was a major part in resolving the question), it is a conflict-of-interest edit in violation of Wikipedia policy to ask supporters to remove material on a Wikipedia page related to you, that it was only material which serves to flatter the subject makes it additionally suspect. Caraturane (talk) 20:32, 31 August 2024 (UTC)

FYI he is doing it again: https://x.com/jaffejuice/status/1835639646774006021 Tomcleontis (talk) 16:35, 16 September 2024 (UTC)

Clarifying Lichtman's Keys

It seems that Lichtman revised the system to predict the winner of the election after the 2000 election, but before the 2016 election. So his 2016 prediction was correct as it was accounting for the overall winner. [1] Wikentromere (talk) 17:17, 16 August 2024 (UTC)

The Keys-model was wrong in 2016, just as the references say. Lichtman wrote a paper where he described the model:

As a national system, the Keys predict the popular vote, not the state-by-state tally of Electoral College votes.

He also believed the popular vote was very correlated with the Electoral College votes (of course), but he was very clear that "the Keys" predicted the popular vote. But in 2016 the Keys predicted that Trump would win the popular vote, who lost the popular vote. So the Keys were wrong in 2016. Paditor (talk) 17:37, 16 August 2024 (UTC)

So is he just lying when he says he revised the model after 2000 to predict the actual winner, not just the popular vote winner, in that NYT piece? (Forgot to login, also genuine question) 2601:194:8380:7D60:7D9D:1CC:DA74:EF1C (talk) 18:04, 16 August 2024 (UTC)

His recent statements about his 2016 prediction are untrue (we shouldn't be calling people liars), there's no doubt that every time he was pressed before 2016, and in his own works, he always said it was just the popular vote being predicted by the keys. Tomcleontis (talk) 18:14, 16 August 2024 (UTC)

So when Lichtman predicted in 2016 that Trump would be impeached, was he predicting that Trump would not be president? He has made it clear he was calling the winner in 2016. Wikipedia1010121 (talk) 23:28, 17 August 2024 (UTC)

He released a paper in October of 2016 (after he said he was predicting Trump would win and would be impeached) which said that he was only predicting the popular vote and it goes a step farther and says he was not predicting the electoral college. He cannot have claimed to have it both ways because he said the keys only predicted the popular vote in 2000 so he was right when he predicted Al Gore. Caraturane (talk) 04:10, 18 August 2024 (UTC)

submitted for your approval

Please append to the "2024 presidential election" subsection:

Vice president Kamala Harris was nominated as the Democratic presidential candidate the next month. Lichtman predicted on September 5, 2024 that Harris would win the 2024 presidential election.

Robert Tait (September 5, 2024). "Kamala Harris will win election, predicts leading historian Allan Lichtman". The Guardian. soibangla (talk) 05:12, 6 September 2024 (UTC)

 Not done as this page is not longer protected, an administrator is not required. You may edit the page as appropriate. — xaosflux ^Talk 14:47, 17 September 2024 (UTC)

I updated it, good add! Caraturane (talk) 16:48, 17 September 2024 (UTC)

Protected edit request on 15 September 2024

Allan Lichtman did not fail to predict the winner (Al Gore) of the 2000 Presidential election, both in the Electoral College and in the popular vote. Five members of the U.S. Supreme Court stopped the vote count in Florida and then voted to install Bush as President. Later, after Bush had been inaugurated, the votes were counted by NORC and it was shown that Gore had won, no matter which decision-rule (concerning hanging chads, dimpled chads, etc.) was used. Little attention was given to this in the media, as the attacks of 9/11 had occurred in the meantime. Norshanson (talk) 06:02, 15 September 2024 (UTC)

 Not done Please provide a specific edit you would like made (ie "please add 'this text' in such-and-such paragraph"), which must be supported by reliable sources. Happy‑melon 09:32, 16 September 2024 (UTC)

I oppose this chance due to the speculative and hypothetical nature. The page already notes that he did predict the popular vote outcome correctly (what his book, model, and himself said he was predicting), but that Al Gore did not win the election. This is asking for an additional step, to assert that Gore actually won the election. That is a dispute better resolved on pages on the 2000 election. Perhaps adding a word like "...with the exceptions of the disputed election of 2000, where he..." would do the job? Caraturane (talk) 13:14, 16 September 2024 (UTC)

Unacceptable source

A source used in this article and in the article about the Keys uses an unacceptable source. I've started a discussion on the other Talk page -- Talk:The_Keys_to_the_White_House#Unacceptable_source. The discussion there will be applicable to this article as well, so let's keep it all in one place. JamesMLane t c 03:30, 3 October 2024 (UTC)

To be clear, you're *arguing* it's unacceptable but others like myself disagree. But yes lets keep discussion just over there. Apprentice57 (talk) 20:08, 4 October 2024 (UTC)

The discussion on the Keys talk page got nowhere. The next step in resolving the dispute was the thread on the Noticeboard for discussing whether Postrider was a reliable source for use here. Apprentice57 expressed the concern that Lichtman might recruit his fans to astroturf the comments. That didn't happen. Three uninvolved editors chimed in. They were obviously legitimate Wikipedians, having edited the encyclopedia since 2021, 2013, and 2005, respectively.

All three uninvolved editors agreed that Postrider should not be cited. Accordingly, I'm removing those links from the Lichtman bio and from the Keys article.

As I've made clear, removing the Postrider criticism doesn't mean the removal of all criticism. The next step is to examine how the articles address any contentious issues concerning the Keys and to see whether the treatment might be improved. JamesMLane t c 16:34, 8 October 2024 (UTC)

I'm going to write this mostly to document what happened, as that whole discussion became a mess. I can't expect editors here in the future to track all that, but I also want to summarize what actually happened because I think we should keep an eye out on how this editor handles themself acrimoniously and in opposition to wikipedia's consensus based approach.

Three (maybe four now) of us who have been active in editing/in the13 keys talk page objected to the removal on the merits, to the opposition of only JamesMLane. They sought uninvolved editors to weigh in on The Postrider being an unacceptable source (fair enough), and as soon as they got the first agreement with them there (*two* editors not three) they made the change unilaterally. They did so while one of us reached out to The Postrider on their independent review standards. JamesMLane agreed it was possible that the email could change even their judgement on the merits, yet made the unilateral edit here and there after 2 days.

We also proposed a good faith compromise that would've clearly been in line with wiki standards and removed any factual reference to The Postrider while still citing their criticism on the 13 keys model, which they dismissed out of hand. I think the other editors prefer we not fight this further, but I want to reiterate that I remain disappointed by the way they used themself as a gatekeeper, and bludgeoned the process to effectuate that. Apprentice57 (talk) 01:26, 11 October 2024 (UTC)

Thanks for your summary of this discussion. I do think this article has the potential to become a niche space where a few editors bludgeon in a bunch of puffery. the article had quite a few weasel words in it, some of which I've removed, and I'm going to continue to work on improving this article. Jjazz76 (talk) 20:48, 11 October 2024 (UTC)

@Apprentice57 pointed out on the other thread that we got another source (in the Atlantic) to use that makes most of this moot. I added a citation on this page to it in the area where the Postrider source was once included. Tomcleontis (talk) 19:48, 16 October 2024 (UTC)

Removing policy violations

There has been extensive discussion on the Keys talk page, including an RfC. The most salient point is what to say about the system's past predictions. In particular, reliable sources offer differing views about whether the system's record is perfect or only almost perfect.

In the RfC, only one uninvolved editor has weighed in so far. She cited WP:NPOV and WP:BLP in favoring the version of the Keys article that did not take sides on that question, but merely reported the competing viewpoints.

Pending further development of the RfC, I've edited this bio article accordingly. This is obviously not final. If other uninvolved editors come to a consensus that the Keys article should adopt one viewpoint, then this bio article can be edited accordingly. In the meantime, the version that best reflects the current status of the RfC should remain in place. Editors should take note of the information at the top of the page, which states in part, "Contentious material about living persons that is unsourced or poorly sourced must be removed immediately from the article...."

Ultimately, once the Keys article is stable, this article should follow WP:SS, with a brief summary here and a wikilink to the relevant section(s) of the Keys article. JamesMLane t c 00:52, 21 October 2024 (UTC)

As per WP:BRD, I'll be specific about my reasons for reverting to the version described above. The edit that I have reverted:

1. Removed the undeniably truthful statement that the Keys have received a lot of media attention. Why, there was even an article about them in The Atlantic!

2. Went back to asserting only one side of the controversy about the system's record, by changing "all or nearly all" (a report of the controversy) to just "nearly all" (the Wikipedia editor's opinion).

3. Removed the undeniably truthful statement that Lichtman hosts live streams. It is not "advertising" to give a factual report of a bio subject's activities in the bio subject's bio article. JamesMLane t c 03:17, 21 October 2024 (UTC)

I've adjusted and explained:

1. It's not notable, useful knowledge, or newsworthy that the Keys get media attention, that's why they are on Wikipedia. Conciseness suggests this is a superfluous note.
2. Essentially every source gives him 9/10 or 8/10. Aside from being wrong about the outcome in 2016, Lichtman also predicted Al Gore in 2000. If you care not to specifically add an explanation about this (like the article originally had) then the only neutral opinion is the commonly held one: that he got most right. Not all.
3. Promoting and linking to a live stream that gets a few thousand views is promotional. Wikipedia:NOTADVERT notes that "Wikipedia articles about a person, company, or organization are not an extension of their website, press releases, or other social media marketing efforts. External links to commercial organizations are acceptable if they identify notable organizations which are the topic of the article. Wikipedia neither endorses organizations nor runs affiliate programs. See also Wikipedia:Notability (organizations and companies) for guidelines on corporate notability. Those promoting causes or events, or issuing public service announcements, even if noncommercial, should use a forum other than Wikipedia to do so." None of these exceptions apply.

Caraturane (talk) 13:13, 21 October 2024 (UTC)

Yeahhh there's pretty much no reasonable reason to say "all elections", come on... Tomcleontis (talk) 15:54, 21 October 2024 (UTC)

1. This is like saying that the bio of Greta Thunberg doesn't need the section on Public response and impact because the public response and impact is the only reason she has a Wikipedia bio. There are plenty of people out there who've never heard of Greta Thunberg or Allan Lichtman. Wikipedia, seeking to make all human knowledge available to everyone in the world, gives readers the salient information about a bio subject, including public response to that person.

2. Note, for example, the article on the Brandeis website, which says that his system "has enabled him to predict the winner of every U.S. presidential election going back to 1984." I think "all or nearly all" is the best description, but your "most" is terrible; that could be just six out of ten. Most say nine out of ten; you say eight out of ten but offhand I don't remember seeing any that were even that low.

As to the wording, "all or nearly all" is best. Nevertheless, since you still object to the application of NPOV (report disputes, don't adjudicate them), I'm changing it to say that "he is credited with a high degree of accuracy in predicting the outcome of the elections from 1984 through 2020." That's less informative to the reader but its accuracy is indisputable, so perhaps this point at least won't need to be further argued.

3. Reporting the fact of a bio subject's role as a YouTuber is not advertising. For example, our Jake Paul bio states "Paul launched his YouTube channel on May 15, 2014.[15] His channel became known for pranks, controversies, and his hip hop music.[16]" The first footnote is a link to the channel itself. Lichtman's channel is not independently notable, and we probably wouldn't mention it in any other article -- but no other article would mention that he has a sister named Ronnie. The bio article is an appropriate place for reporting things that are significant in the context of that person's life, even if they aren't otherwise notable. JamesMLane t c 01:30, 22 October 2024 (UTC)

1. Figured the introduction could use some tightening but I'm not going to nickle and dime you on it.

2. "High degree of accuracy" is fair, see we can get along :)

3. Jake Paul is known for being a YouTuber and his videos get tens of millions of views. Allan Lichtman is not, and his videos don't. I believe including the channel is promotional and probably a violation of Wikipedia:EXTPROMO but I'm open to some additional input; I just really don't see how a YouTube video is super relevant in terms of Lichtman's career or history if it's not providing some sort of reporting, academic review, or major life event. Again I don't really want to nickle and dime these much more, I am not going to lose a lot of sleep over it, but in the interest of getting these pages in tip top shape, it stood out.

Thanks for your work otherwise, I don't mean to be a pain. Caraturane (talk) 04:09, 22 October 2024 (UTC)

It does feel promo'y at this point. To add some specifics: the channel was registered in 2019 but the oldest video viewable on the platform is from July 2023 and is commentary on this election cycle (older videos could exist if privated/unlisted/removed, but the video in question has 2.3k views - very few, so we're probably not missing a wealth of videos). He has 130k subscribers and 9 videos over 100k views (only 1 over 500k views). This is an up and coming channel, but still a medium sized one even in the context of YouTube (so small in the context of wikipedia). Usually notable channels are at least at the 1M subscribers mark. Apprentice57 (talk) 18:52, 22 October 2024 (UTC)

Glad we're at a point we're just debating if a Youtuber channel is relevant. Personally I don't think it's the end of the world to include it, but I was also one for including the Postrider as a source, which seems just as relevant/popular as a Youtube channel getting a few thousand views, so I don't want to be a hypocrite myself. I just generally lean in favor of overincluding! Tomcleontis (talk) 18:57, 22 October 2024 (UTC)

You're comparing apples and oranges. Postrider is not a reliable source per WP:SPS, so we don't cite what it says as a source for a factual assertion. As for the YouTube channel, however, we're not citing anything it says as a source. Instead, we're merely reporting on the simple (and undisputed) fact of its existence, because it's worth mentioning as one aspect of the life of the person whose bio this is. If the Postrider authors ever have Wikipedia bios, then, yes, those bios will mention that they have a blog (or whatever people want to call it), with a link.

There appears to be agreement to restore the info about media coverage and the YouTube channel, so I'll do so. JamesMLane t c 23:09, 24 October 2024 (UTC)

No, there is not agreement about the YouTube channel. As I argued above it's probably promotion. Apprentice57 (talk) 03:37, 29 October 2024 (UTC)

In the interest of getting along, I'm acquiescing in your removal of the term "model" to describe a system that is widely described as a model. That term was even used in the articles in The Atlantic and Newsweek that were advanced by Lichtman detractors as being so helpful. I'm sure there are people who get into detail about what does or does not constitute a model; I'll leave it to them to argue the point. JamesMLane t c 23:15, 24 October 2024 (UTC)

Look at us go, getting along! I believe The Postrider (their editor in chief noted that was the correct mechanic) may qualify as a reliable source ultimately in consideration of how work is reviewed, who is running it, and reliable and independent verifiability, but I have asked their editor to publish an outline of their editorial standards publicly and I'll hold my comment until that is done (for what it is worth, they were good sports about it).

The YouTube thing we can come back to eventually if we need to but I am willing to let it lie at this juncture, honestly I am hoping to take a couple weeks away from these pages and Wikipedia with the upcoming election filling my head. Caraturane (talk) 23:22, 24 October 2024 (UTC)

Again, you need to stop making unilateral edits to these pages while you wait for comments on your proposed changes. I explicitly pointed out this was problematic last time and asked you to not repeat this (waiting for the first feedback that agrees with you and changing things after a trivial amount of time).

Also remember, RfC welcome *all* to weigh in on proposed changes and does not distinguish between involved and uninovlved. The responses on a whole did not support your reading that there are any policy violations. Apprentice57 (talk) 17:20, 21 October 2024 (UTC)

Neutrality concerning prediction record

The accuracy of The Keys to the White House in the elections of 2000 and 2016 has been the subject of extensive discussion on that Talk page and elsewhere. In the published sources, opinions differ. (See here for an explanation of the differing assessments, in the published sources, of those two years.) Wikipedia therefore cannot take a position and endorse one side or the other.

Editors can see 25,000 or so words worth of dispute over at the Keys talk page. While that dispute continues, we settled on saying, in the intro section to the bio article, that he had a "high degree of accuracy" because it's undisputed that he's usually been right. Any reader wanting more detail can follow the link to the daughter article.

I'm restoring that neutral wording, contra to the edit by User:108.15.35.13 that inserted clear POV. The simplest thing will be for editors wanting to address this subject to participate in the discussion at the Keys talk page. JamesMLane t c 19:09, 6 November 2024 (UTC)

I've reverted the edit by User:Kalbome22 and encouraged that user to engage in a discussion here. The subject has already been extensively discussed, this is the current compromise wording, and any changes should be discussed here. JamesMLane t c 20:18, 6 November 2024 (UTC)

You are clearly incorrect biased and making attempt to suppress information. Please report me Kalbome22 (talk) 20:25, 6 November 2024 (UTC)

I've explained to you on your Talk page that reverts of long-standing language, worked on by several editors, should be explained and discussed here.

And, no, "You are clearly incorrect" etc. does not constitute the kind of discussion that is supposed to characterize Wikipedia editing. JamesMLane t c 22:29, 6 November 2024 (UTC)

Kalbome2 seems to have two points. First, of course, one particular assessment of the Keys is right, everything else is wrong, and therefore the opinion favored by Kalbome2 must be stated as The Truth (nevermind WP:NPOV). Second, details about the system's record are so important that they must be included in the bio article, and prominently at that, because relying on WP:SS constitutes suppression of information.

In an attempt to avert an edit war, I'll edit to give a neutral description of the different points of view. This is an attempt to address Kalbome2's second gripe. My inclination is to think that this much detail in the lead section is excessive and can be left to the child article, so I don't feel strongly about the difference between this language and the version that was in place as of yesterday. JamesMLane t c 22:41, 6 November 2024 (UTC)

"high degree of accuracy is hardly NPOV" - he's 9 for 11 based on a model for 'checks notes' predicting earthquakes that did not 'checks notes' ever predict any earthquakes.

Has he usually been right? Sure. But most of those elections were total blowouts. Let's stick to what we can actually include with reliable sources. And also let's stop trying to reinclude the same material multiple times in one day @JamesMLane. We can also use the talk page to discuss inclusion, but I'm not seeing significant consensus on what you keep trying to include. Jjazz76 (talk) 23:15, 6 November 2024 (UTC)

I don't see what is disputable or not neutral about my revision. There are no rules about the length of the intro and you aren't any sort of authority on this matter. Kalbome22 (talk) 23:19, 6 November 2024 (UTC)

In light of the POV-warring snap reverts, it may be inconvenient for other editors to grasp the context. Here's the key language that neutrally presents the differing POVs about the prediction record:

Lichtman is credited with a high degree of accuracy in predicting the outcomes of the elections from 1984 through 2020 using the system. Assessments state variously that he got them all right, or that he was correct in all but 2000, or that he was correct in all but 2016. It is undisputed that his prediction of a win for Kamala Harris in 2024 was incorrect.

This language acknowledges the opinion stated here by Jjazz76. (Actually, I don't know whether Jjazz76 agrees with the people who think Lichtman missed in 2000 or with the people who think that he missed in 2016, but whichever it is, it's covered.) In a case like this, where there are multiple POVs that have significant support, the policy of WP:NPOV requires that we present each side fairly but without endorsing any of them. JamesMLane t c 23:58, 6 November 2024 (UTC)

Again, I think "high degree of accuracy" is not NPOV. If we are going to include that particular language it needs specific cites.

As an alternative I think we should just simple say. "Lichtman predicted" the outcomes of the elections from 1984 through 2020 using the system." And then go on to with the language you provide. Jjazz76 (talk) 00:09, 7 November 2024 (UTC)

I agree that it should be cited. The citations are the six footnotes in the next sentence. For each POV about 2000 and 2016, there are citations to examples of notable statements of that opinion. As for "high degree of accuracy", that language (consistent with all six citations) was a compromise. I originally wrote "all or nearly all", which I still think is more informative. The editors who want to push a POV hostile to Lichtman didn't like it, so we settled on this that everyone could live with. JamesMLane t c 00:17, 7 November 2024 (UTC)

Again, if there is no cite for "high degree of accuracy" it should be removed. I think you've done a fine job with the other material and do a good job of teasing out the mixed reporting in RS. But "high degree of accuracy" seems to border on OR, particularly when not backed by sources. Jjazz76 (talk) 01:09, 7 November 2024 (UTC)

It's pretty routine to say that 9/11 and 10/11 are high. Alternatively, how would you feel about going back to "all or almost all"? JamesMLane t c 02:33, 7 November 2024 (UTC)

9 out of 11 is 81 percent. That doesn't sound super high to me. All is obviously wrong based on the RS.

How about to maintain NPOV we just leave it out because we haven't identified any RS to support the claim and Wikipedia runs on RS? Jjazz76 (talk) 04:32, 7 November 2024 (UTC)

I agree with leaving it out. There's no issue with the article later covering both sources that credit him with a high level of accuracy (some do!) OR low level of accuracy (some do!), but the introduction shouldn't favour one POV of his track record.

Further, it's certainly not routine to call 9/11 "accurate" — accuracy is always subjective and contextual. A 90% target hit rate for a marksman firing a rifle from 5 metres would almost certainly be called inaccurate, for example. So even if the article can fairly state that other sources credit him with high accuracy, it's not NPOV for us to decide ourselves that this is a fair and routine assessment.

-----

"all or almost all" is probably untenable. At bare minimum, for example, Lichtman himself wrote repeatedly in 2016 (in his book and academic paper) that the keys predict the popular vote. *Even if* Lichtman ALSO made an Electoral College prediction that year, the popular vote prediction would still be definitively wrong. So you'd always have to add the qualifier of "all or almost all Electoral College results" (or something similar) to maintain 'all' in there — but THAT would be objectively wrong since even Lichtman himself states he didn't make EC claims until after 2000. There's just no coherent way to support Lichtman having gotten ALL his predictions correct and the level of semantic trickery required to make it non-falsifiable just isn't worth the squeeze.

And now that his 2024 prediction is also definitively wrong, what would even be the purpose in saying that he'd predicted "all or almost all elections from 1984 until 2024"? You're truncating the dataset a bit arbitrarily to point out something about a specific 'run' within that dataset which isn't even certain.

Something like "a majority of elections since 1984" is indisputably accurate, easy to maintain, and doesn't get into any semantic difficulty or odd phrasing. I feel it's safe to let the reader decide (based on Lichtman's words & reliable sources later in the article) exactly whether Lichtman's record is more like 8/11 or 9/11 or 10/11. Hangways1 (talk) 10:26, 7 November 2024 (UTC)

@Hangways1 - Thanks for doing a good job of articulating my points way better than I could. Yes 'majority' feels the most accurate and NPOV for the reasons you highlight above. Jjazz76 (talk) 14:38, 7 November 2024 (UTC)

Just saying "majority" is misleading -- it could be six of eleven. As for 2016, concluding that "There's just no coherent way to support Lichtman having gotten ALL his predictions correct" is unacceptable. You're assessing the evidence, looking at some of Lichtman's statements, reaching a conclusion that you think one side of the dispute is better, and wanting Wikipedia to state that that side is correct. That's precisely what would violate WP:NPOV.

Note that: (1) Lichtman has said that he was predicting popular vote before 2016 but then switched. (2) Even if it were just his assertion, that doesn't justify stating, in Wikipedia's voice, that he's lying. We'd still report both sides and let the readers decide for themselves. (3) But it's not just his assertion. Multiple reliable sources say that he got 2016 correct. Yes, some of those sources also say that he got 2000 wrong -- but there are sources crediting him with having gotten ALL his 1984-2016 predictions correct. (There's no dispute that 2020 was right and 2024 was wrong.)

The basic problem, here and on the Keys article, is the natural human tendency to consider pros and cons and reach a decision. That's an appropriate response in many circumstances, but not here. JamesMLane t c 17:17, 7 November 2024 (UTC)

"high degree of accuracy" is also misleading. not sure we have consensus for either phrase, so let's just leave it out all together. how does that sound? Jjazz76 (talk) 20:31, 7 November 2024 (UTC)

The problem is that it omits relevant information, namely the media accolades for his general success. Wikipedia editors are entitled to think he's overrated but the fact of his image is worth reporting.

If you won't go along with anything specific about the predictions, what about something like: "He has been widely hailed as the "Nostradamus" of political prediction (a reference to the sixteenth-century reputed seer)." You can do a search for "Lichtman Nostradamus" and get lots of hits. I personally think this is a bad way to tell readers about Lichtman, because Nostradamus was a humbug; but he has enough cultural prominence that plenty of readers recognize this term as conveying notability in forecasting. JamesMLane t c 20:45, 7 November 2024 (UTC)

Yeah i don't support widely hailed as a modern Nostradamus either. Jjazz76 (talk) 00:45, 8 November 2024 (UTC)

To be clear, I meant that I think it's a bad idea for outfits like Newsweek to use that description. Given that numerous sources do use it, however, it's accurate for Wikipedia to report the use. JamesMLane t c 01:29, 8 November 2024 (UTC)

Another option would be to tweak 'majority' - perhaps 'large majority' or similar? Hangways1 (talk) 01:20, 8 November 2024 (UTC)

Bear in mind that Lichtman claims that, for 1984-2020, "all" is correct. Bear in mind also that some reliable sources agree. To say "majority" or even "large majority" doesn't fairly report that POV.

If you want to tweak, we could consider "all or a large majority", which picks up all three of the major POVs. JamesMLane t c 01:36, 8 November 2024 (UTC)

What Lichtman claims isn't relevant, nor what his alma mater says. We need to look at the balance of reliable sources. All isn't correct, full stop. It doesn't matter what Lichtman says.

Majority seems the most NPOV based on our discussion. Jjazz76 (talk) 01:55, 8 November 2024 (UTC)

If someone is accused of running a red light and causing an accident, and five witnesses say he did it, he's still allowed to take the stand and testify that he had the green light. That's evidence. It's relevant. Alternatively, if he and the five witnesses all say he had the green, and only the injured pedestrian says the light was red, that's relevant evidence. In either case, the judge won't exclude self-serving statements on the basis of "the balance of reliable sources". The evidence comes in and the jury decides.

The analogy here is that the jury isn't us -- it's the reader. Our role, per WP:NPOV, is to present the evidence, not to weigh it. If we report what Lichtman says, or what The Washingtonian says, etc., along with reporting what the other side says, then each reader can decide how much weight to give to each source. A model can be seen at O. J. Simpson#Prosecution and defense cases, where the article reports on the trials by summarizing the arguments without saying "O.J. was guilty, full stop" or "O.J. was innocent, full stop."

We certainly wouldn't write "One prominent historian said..." without disclosing that it was Lichtman himself. We report the explanation, attributed to him, and readers who decide to dismiss it as self-serving may do so. Here's what the applicable Wikipedia policy says on the point:

A common argument in a dispute about reliable sources is that one source is biased, meaning another source should be given preference. Some editors argue that biased sources should not be used because they introduce improper POV to an article. However, biased sources are not inherently disallowed based on bias alone, although other aspects of the source may make it invalid. A neutral point of view should be achieved by balancing the bias in sources based on the weight of the opinion in reliable sources and not by excluding sources that do not conform to the editor's point of view.

We apply that here by reporting that there are sources (including but not limited to Lichtman himself) stating that "all" is correct, just as there are sources disagreeing with him and saying he got all except 2016 right, just as there are sources disagreeing with both and saying he got 2016 right but got 2000 wrong. JamesMLane t c 13:52, 8 November 2024 (UTC)

I don't see a compelling reason not to just summarise Lichtman's entire track record (1984-2024) and remark on individually notable years. Unless there are reputable sources claiming Lichtman got all his predictions correct including 2024, this debate now seems redundant. You might as well debate how to present what people thought of his track record from 1984-2016, or from 1988-2024, or 2000-2024, or 2000-2008, or so on. Even if we could agree on the most NPOV way to do so, we'd still be making a choice as editors to focus on only a slice of his track record, and this particular slice seems very arbitrary to me. (It's not like Lichtman only became famous in 2021, or changed his keys after the 2020 election, or was never said to have made an incorrect prediction prior.)

The way the main Keys article currently handles it is as follows:

Lichtman has successfully predicted a large majority of the eleven presidential elections held since 1984 by using the keys. However, he incorrectly predicted that Kamala Harris would win the 2024 election, and the nature and accuracy of his predictions for Al Gore in 2000 and Donald Trump in 2016 have been disputed.

This is also NPOV and properly encyclopedic — every single reliable source I know of would agree with every claim that in that sentence. The paragraph as a whole is balanced, with the first sentence providing the merits of Lichtman's track record and the second sentence its challenges. It is up-to-date, and does not require us to arbitrarily summarise only a selected time-slice of Lichtman's predictions. It is exceptionally easy to read. And of course, the article provides a full table of Lichtman's track record & facts relevant to those disputes later on, so no information is absent from the article.

I am just not seeing sufficient justification for losing several of those merits in order to separately summarise how Lichtman did until 2020, when it's now 2024 and a full election later. Hangways1 (talk) 14:43, 8 November 2024 (UTC)

@JamesMLane It isn't in dispute that the prediction system didn't predict the Electoral College in 2000 and 2024 and the popular vote in 2016 and 2024. Kalbome22 (talk) 15:00, 8 November 2024 (UTC)

It also isn't in dispute that the prediction system didn't predict the outbreak of COVID-19. Should the article include that undisputed fact? If not, why not?

The obvious answer is that there's no reason to even mention a supposed prediction that the system didn't say it was making -- and certainly no reason to say "failed", which implies that there was an attempt. We have to make a neutral presentation of the arguments about 2000 and 2016. Anyone reading a neutral version will know who won the popular vote and the electoral vote in those years. JamesMLane t c 17:10, 8 November 2024 (UTC)

I'm still just not seeing the need to provide a summary of an truncated track record (1984-2020) instead of his overall track record.

Obviously you can always infer one from the other, but everybody agrees that Lichtman got a majority (or large majority, if you like) of the predictions from 1984 to 2024 correct.

What value do we get out of excluding an incorrect prediction from that dataset and then summarising the rest? This seems almost as arbitrary as writing "Lichtman predicted a majority of elections from 1988-2024. He also got 1984 correct."

If your goal here is to make sure the introduction includes references to how the media & notable figures have perceived his track record, then I don't see a need to accomplish this by collecting perspectives on only a truncated version of his track record. Let's just collate some of the sources crediting him with high/mixed/low accuracy over the years (no need to pick out 1984-2000 specifically) and summarise his media perception in a separate sentence.

Or even better, perhaps there are already enough sources by now that summarise his entire track record INCLUDING 2024..? Not just from 1984-2020?

I Hangways1 (talk) 01:56, 8 November 2024 (UTC)

The current POV language

As I write, the version that's up contains this language: "although his prediction system failed to predict the outcome of the Electoral College in both 2000 and 2024 and failed to predict the popular vote in 2016 and 2024." His system also failed to predict the outbreak of COVID-19. To say "failed" implies that it was something the system was trying to do, an assertion that some sources would support but that others would dispute. NPOV is satisfied by noting the existence of the differing opinions. JamesMLane t c 03:31, 7 November 2024 (UTC)

What's your alternate suggestion? Jjazz76 (talk) 04:34, 7 November 2024 (UTC)

I agree with @JamesMLane here that this is a wonky & contentious sentence. I wonder if we need to "draw a conclusion" at all in the introduction about whether PV, EC, or both were being predicted; can we leave the nuance until later in the article where we have more space?

I've implemented this phrasing over in The Keys to the White House article: "Lichtman has successfully predicted the majority of the eleven presidential elections held since 1984 by using the keys. However, he incorrectly predicted that Kamala Harris would win the 2024 election, and the nature and accuracy of his predictions for Al Gore in 2000 and Donald Trump in 2016 have been disputed."

Is this safe? It doesn't seem contentious to note that Lichtman was incorrect in 2024, given that Lichtman has publicly concurred. And as long as the remainder of the article outlines why people dispute the accuracy of 2000 & 2016, there's no fact being left out. Hangways1 (talk) 10:13, 7 November 2024 (UTC)

My alternate suggestion is stated above. Here it is again:

Lichtman is credited with a high degree of accuracy in predicting the outcomes of the elections from 1984 through 2020 using the system. Assessments state variously that he got them all right, or that he was correct in all but 2000, or that he was correct in all but 2016. It is undisputed that his prediction of a win for Kamala Harris in 2024 was incorrect.

The best solution will be to get the Keys article into neutral shape. My offering of a neutral version there was snap-reverted (because some editors insist that their assessment of the weight of the evidence is correct and that hence there is "no dispute"). My version of the Keys article dealt with the dispute by alluding to it briefly early on, with a cross-reference to a dedicated section under "Criticisms" that explained the differing POVs in more detail. I still hope that an NPOV solution will eventually be implemented over there. Then this bio article can have a short statement along the lines of the suggestion above, followed by a wikilink to the more detailed subsection in the Keys article, per WP:SS. JamesMLane t c 14:18, 7 November 2024 (UTC)

References

JamesMLane t c 14:18, 7 November 2024 (UTC)

References

2024 Election

"Litchman[sic] apologized and admitted being wrong."

Lichtman's tweet cited admits being wrong but neither contains nor infers any apology.

"Unlike Nate Silver, who will try to squirm out of why he didn’t see the election coming, I admit that I was wrong. I will assess the election and the keys on my live show this Thursday at 9 PM Eastern."

Far from any apology (which wouldn't take an object, or make any sense in the context, in any event; it's simply that his system finally broke), he takes the opportunity to take a shot at his eternal nemesis. JohndanR (talk) 20:39, 6 November 2024 (UTC)

Fixed it. Yes I am a nerd -XCBRO172 (How could you tell?) 22:56, 6 November 2024 (UTC)

Restoring appropriate information

Some edits have removed properly encyclopedic information, without discussing the matter here, and with the common thread being that everything removed reflects favorably on the bio subject. I'm restoring this information. This is unrelated to the prediction record discussed above. JamesMLane t c 00:23, 7 November 2024 (UTC)

Just some of the neutrality problems with the current version

Skimming over this version, in place as I write:

1. "Lichtman is credited with predicting most of the elections outcomes from 1984 through 2020 using his interpretations of the system.[5][6][7][8]." Omits the POV, taken by independent sources as well as by Lichtman, that he got ALL those years right.

2. "He did not correctly predict the outcome of the Electoral College in 2000 and 2024, or the popular vote in 2016 and 2024.[9]" Clearly implies a false prediction. In 2000 he was not predicting the electoral vote outcome, and his critics who focus on the 2016 popular vote haven't said that he was, so it's false to imply that he made such a prediction.

3. The discussions of 2000 and 2016 later in the article are biased toward the POV critical of Lichtman, by not adequately presenting the other side.

Also, with regard to 2000, my neutral version of the Keys article corrected this foolishness: "George W. Bush was declared the winner of the Electoral College by the Supreme Court...." As part of reverting from the neutral version back to the POV version, this incorrect explanation has been restored to the Keys article, and now even imported here. JamesMLane t c 03:42, 11 November 2024 (UTC)

The introduction in this article has lagged behind the improvements made in The Keys to the White House (where similar discussions & improvements have been discussed). I've copied over that version to address issues #1 and #2.

Regarding #3, I agree that the SC declaring Bush the winner is inaccurate and have removed it from both articles. Perhaps someone will want to allude to Bush v. Gore more directly in this version; we could discuss that for sure.

What specific improvements do you want to the 2000 & 2016 sections? Hangways1 (talk) 12:49, 11 November 2024 (UTC)

Even if he wasn't trying to predict the outcome for those specific elections there is sufficient media coverage and public interest to include in the intro. Kalbome22 (talk) 17:04, 11 November 2024 (UTC)

I can see that you've reworked this passage to have a more neutral tone, Kalborne, which is good!

To be honest, I still prefer the current lede on The Keys to the White House — I'll put the two next to each other below just for ease of reference in this discussion:

Keys article: Using the keys, Lichtman has successfully predicted most of the last eleven presidential elections held since 1984, often making his prediction months or years in advance. The system has been frequently regarded as a reliable predictor of U.S. election outcomes. However, he incorrectly predicted that Kamala Harris would win the 2024 United States presidential election, and the nature and accuracy of his predictions for Al Gore in 2000 and Donald Trump in 2016 have been disputed.

Lichtman article: Lichtman is credited with correctly predicting the outcomes of most presidential elections from 1984 through 2020 using his interpretations of the system. He did not predict the outcome of the Electoral College in 2000 and 2024, or the popular vote in 2016 and 2024.

Obviously, the two differ in several ways, but let's stick to the coverage of the disputed years. I agree with @JamesMLane here that the second version is unclear and even potentially misleading here, especially because the nature of why Lichtman didn't predict these outcomes varies a lot:

• He [arguably] didn't predict the EC in 2000 because he never wrote that he was predicting it
• He definitely didn't predict the EC in 2024 because he got it wrong
• He [arguably] didn't predict the PV in 2016 because he wrote that he was predicting it but claims to have changed the methodology
• He definitely didn't predict the PV in 2024 because he never wrote that he was predicting it

Conflating all these into a singular "Lichtman didn't predict them" at best obfuscates that there are heavily varying levels of agreement over these matters, and at worst misleads the reader into thinking that Lichtman actively made unambiguously incorrect predictions on all four matters. (I'm not aware of a single RS claiming Lichtman made a PV prediction in 2024, so *that* ambiguity is worth clarifying at the very least.)

Yes, I saw your edit note: you've technically just said that Lichtman did not predict these matters. Ok, technically correct, sure. But lots of people *are* going to read that as "he made an incorrect prediction" as though your second sentence is being presented in juxtaposition to your first (about his correct predictions), and I too favour finding a phrasing that has a lower risk of being misinterpreted that way.

The Keys article version handles this by specifically calling out 2024 as a unique case (everybody agrees it was incorrect) and then noting that both the nature and accuracy of 2000 & 2016 are disputed. You seem to view this—and please correct me if I'm wrong!—as insufficiently detailed on the popular vote vs electoral college front. i.e. that it's not enough just to say these years' predictions are disputed; we must also note the exact primary manner (PV vs EC) in which some of them are disputed or notable.

Would you be open to alternative ways of communicating this (whether in the same sentence or another) that don't run into the confusion/misleading risk that James & I are experiencing? Are there specific improvements you think could be made that accomplish the best of both versions? Hangways1 (talk) 22:17, 11 November 2024 (UTC)

@Hangways1 How do you feel about this or something similar: "While Lichtman's system has accurately predicted most presidential outcomes since 1984, certain years are disputed. In 2000, he incorectly predicted that Al Gore would win, in 2016, his incorrect popular vote prediction involved a later methodological adjustment. His 2024 prediction of a Kamala Harris victory, however, was incorrect." Kalbome22 (talk) 22:53, 11 November 2024 (UTC)

"Lichtman's system has been accurate in most elections since 1984, though certain years are debated: In 2000 he predicted that Al Gore would win despite his electoral college loss, In 2016 he adjusted his popular vote after he predicted Donald Trump would capture it, and incorrectly predicted a Kamala Harris win in 2024. Kalbome22 (talk) 22:59, 11 November 2024 (UTC)

Lichtman's system has accurately predicted most elections since 1984, though some years are disputed: in 2000, his prediction didn’t account for the Electoral College outcome; in 2016, he incorrectly predicted the popular vote, later citing a method change; and in 2024, he incorrectly predicted a Harris win. Kalbome22 (talk) 23:06, 11 November 2024 (UTC)

All three of these versions take a position on the dispute, endorsing one of the POVs, instead of merely reporting on the dispute, as WP:NPOV requires. JamesMLane t c 23:21, 11 November 2024 (UTC)

@JamesMLane How about something like this: "Lichtman's system has accurately predicted most elections since 1984, though some years are disputed: in 2000, he predicted a Gore win, but Bush won; in 2016, he initially predicted a popular vote victory for Trump, then revised his methods; and in 2024, he incorrectly predicted a Harris win." Kalbome22 (talk) 23:34, 11 November 2024 (UTC)

I agree with @JamesMLane on this one. I don't think the positions you state on 2000 or 2016 are unambiguously true enough for the article to lead with them.

I mean, you'd struggle to even find critics of Lichtman who would tend to agree with you. From what I can tell, most of Lichtman's critics think he got 2000 mostly or entirely correct (since his writing explicitly stated he was making a popular vote prediction, but never explicitly stated he was making an EC one as well), on the same basis that they think he got 2016 wrong.

---

"It is my opinion that 2024 and 2000 must be clarified as incorrect presidential predictions"

I think everybody agrees with the Wiki presenting 2024 as incorrect. I think almost nobody else here would agree with presenting 2000 as definitively incorrect.

---

"2016 should also be included as a disputed popular vote prediction"

• I think it's essential to note that some years' accuracy is disputed in the lede.
• I don't think it's essential to specify exactly how in the lead; the remainder of the article addresses that nuance and there's no need to clog the intro with the technicalities of each disputed year.
• If we do specify how it's disputed in the lead, I still don't think the phrasing used captures the situation well — e.g. "He initially predicted a popular vote victory for Trump" would be disagreeable both to Lichtman (who claims he never did this) and his critics (who would object to 'initially' on the basis that they don't think he actually did change the prediction in any real sense)

Hangways1 (talk) 18:37, 12 November 2024 (UTC)

Can you provide some independent sources that say that he correctly predicted the 2000 election? Kalbome22 (talk) 18:42, 12 November 2024 (UTC)

Sure, there are several already in the articles.

Here's PBS stating he's predicted the popular vote successfully since 1984.

Here's WaPo stating he's predicted the popular vote successfully since 1984.

Here's US News stating he correctly called every election since 1984.

Here's AOL that says (at the end, ignore Lichtman himself saying it earlier in the article) that he predicted Al Gore's popular vote victory.

Hell, here's The Atlantic, in an article very critical of Lichtman, that states that his system worked as advertised in 2000 and points out he wrote that he was predicting the popular vote.

As far as I can tell, independent media coverage typically (though not always) regards him as getting 2016 right but 2000 wrong, whereas his critics tend to regard him as getting 2000 right but 2016 wrong.

There are certainly people who regard him as getting both wrong — or at least having *a* prediction wrong in both years, since he could have made a popular vote AND electoral college prediction, and gotten one right and one wrong!

But there's no "one size fits all" media consensus on exactly what his predictions involved in 2000 & 2016.

This is part of what's driven my edits over the last few weeks to include more direct quotes from Lichtman, so the reader can see for themselves what he wrote and make their own mind up. Hangways1 (talk) 19:06, 12 November 2024 (UTC)

@Hangways1 It is my opinion that 2024 and 2000 must be clarified as incorrect presidential predictions, disputed or otherwise, whatever that looks like. 2016 should also be included as a disputed popular vote prediction. This is my biggest concern with the article Kalbome22 (talk) 23:24, 11 November 2024 (UTC)