Talk:Allan Lichtman

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

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)

Direct quotes from Lichtman are useful in this regard, I concur. I think writing it so it's parallel with the 13 Keys page: saying the accuracy/nature of his 2000 and 2016 predictions is disputed, but his 2024 prediction was clearly false, is the best approach in that it is concise and to the point. There's plenty of sourcing out there available, and the record and criticism section of the 13 Keys page does a good job getting into nuance that we don't need on his personal page. Tomcleontis (talk) 19:07, 15 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)

The problem with that is you then get into the legitimacy of Bush's win. It's been Lichtman's position since 2000 that he correctly predicted the EC but irregularities in ballot counting and the politicisation of the Supreme Court won Bush the election.

He wrote to the electoral commission to this end and as such I don't think we can present his 2000 prediction record as undisputed. 2A00:23C5:11E:F901:E91A:AC65:47F4:CE9F (talk) 01:42, 17 November 2024 (UTC)

I should add that he's been inconsistent in expressing this position to the media but he most recently expressed it on his YT channel before the election. 2A00:23C5:11E:F901:E91A:AC65:47F4:CE9F (talk) 01:44, 17 November 2024 (UTC)

References