7000-Year-Old Trees Uncovered High Up In The Rockies

That’s a routine procedural standard following a dismissal.

That is incorrect. Nothing in the ruling indicates that the case was dismissed with prejudice. The “prejudice” referenced in the ruling relates to harm to the defendant caused by the delay. I think you are conflating two different legal concepts.

Apart from prejudice being presumed in the case of delay, some of Ball’s key witnesses died of old age (literally) during the period in question, so it’s hard to argue otherwise.

AI can be influenced heavily into saying what you want it to say by biasing the words.

Sage wisdom. Guess we should refer to primary sources, like the official court ruling.

We’d be done and off to lunch by now if you just looked at the primary source document, directly from the judge’s mouth.

But its Mann’s case so he cant be held responsible for being slow in his own case.

Of course he can, and that is exactly what Ball did. The defendant has some expectation of a right to an expedient trial. You are correct that nowhere in the official ruling does the judge say, suggest, or imply that Mann failed to comply with any court orders to turn over documents or information. Thus there is no basis, none whatsoever, for the repeated accusations that refusal t comply with court orders was the reason for dismissal.

My point is that there is no evidence of Mann refusing to comply with any information requests, yet it’s being cited on this website as the reason why the case was dismissed, by numerous people, none of whom can seem to cite exactly where they got that information.

What is the point in speculating? Who can say what the reasons were. There is no evidence that they had anything to do with defiance of a court order. Can we agree on this point, at least? Committing contempt of court and defying court ordered discovery would certainly have warranted mention in the ruling.

And when would the court have done this? They never even got to a hearing.

People in this thread are claiming that it happened, and was the reason for dismissal.

We’ll never know for sure but you’re acting as if “no evidence” means didn’t happen when all it means is unverifiable.

It certainly means you shouldn’t be going around telling everyone that the case was dismissed because Mann defied court ordered discovery.

I would urge you to read the thread, that is indeed what is being alleged, and is commonly alleged every time the Mann v. Ball case is brought up on this website.

And if that’s the reference, then you’ve added “court ordered”.

Because it is the court that mandates that parties comply with discovery. If the court did not cite Mann for refusal to comply, then there was no refusal. Again, the issue is that the allegation that Mann refused to turn over documents and data that is why the case was dismissed is completely, utterly baseless, yet repeated ad nauseam.

Not once did I deny that the dismissal was due to Mann’s delay, but thank you for quoting directly from the ruling and affirming what I’ve said about the case. Cheers.

It took Wahl and Amman a few weeks to replicate Mann’s methods and results. M&M weren’t trying to replicate the methods of the study and compare results, they were trying to reverse-engineer the exact computational steps for the purpose of “auditing.” This is much more labor intensive than replication, and unnecessary. It was also ultimately pointless, because M&M never produced any substantive challenge to the findings of the papers despite years (now decades) of pointless nitpicking.

Your comment contains a multitude of misconceptions and conflation of different issues, so it’s a bit difficult to untangle in a clean way. I’m focusing on the primary issue you raise, but hours could be spent dissecting the various other errors you’re making.

M&M’s claim that the hockey stick is an artifact of non-centered PCA is simply wrong, and has been shown to be wrong by numerous scientists. M&M failed to follow selection rules for determining the number of PCs to retain using their centered-PCA, so they incorrectly retained only the first two components, despite the selection rules dictating that the first five be retained. Using their centered PCA yields the hockey stick when the correct rules are applied (M&M’s reproduction is in blue):

But importantly, the hockey stick shapes emerges even when PCA is not used at all:

The yellow line shows the effect of simply using the individual proxy records directly, without PCA.

The problem with M&M’s critiques has always been that the egregious issues they alleged, which might have brought the robustness of the reconstruction into question, are invariably based on gross errors or misunderstandings by M&M, and any actual issues or errors they uncovered in the analysis have no impact on the results – it’s just pointless nitpicking.

W&A showed conclusively that if you follow the methods described in Mann’s publications, using the set of proxies listed, you obtain the same result. They further applied different methods of compiling the reconstruction and also obtained the same result, providing independent verification of not only the methodology, but the conclusions. A multitude of other studies in the many years since have similarly confirmed the results.

Yes, as I’ve noted, reproducing the hockey stick by following Mann’s documented methods is trivial, as WA showed. This isn’t good enough for Steve M, who insists that unless the reproduction is perfect, it is nothing. It’s like saying, “I followed your directions and got to your house, but unless you show me where every single one of your footsteps landed as you walk, I really haven’t reproduced your route.”

There’s some esoteric academic interest in doing this, I suppose, but it is of no practical value. You don’t need to use principal components analysis at all – the hockey stick shape emerges from using the proxies directly, as I showed above.

They ended up having to “reverse engineer” it, because the list of proxies actuallyused was not the same as the one in the SI of the original paper.

McIntyre is saying he had to retain the first 6 PCs instead of the first 2 to “exactly match” Mann’s line, but retaining two produces the WA graph, which yields the hockey stick. Again, it’s all arguing over nitpicks, that’s what Steve M has spent a quarter century doing. Nothing in his work has come close to overturning the fundamental result of MBH98 or any of the myriad reconstructions since.

This is what your AI is describing as an inability to replicate Mann’s results:

This is a graph showing the W&A reproduction compared to MBH98. McIntyre claims he needs to be able to make the lines identical for his auditing purposes, otherwise replication hasn’t been achieved. Any sane person would look at this and conclude that following the methods described in Mann’s paper yields the hockey stick.

Additionally, some of your AI’s responses are garbled interpretations of the sources it is linking, and you need to follow each and every one of them up carefully for verification. It also seems to be relying entirely on Steve McIntyre’s writings, which yields a completely one-sided and incomplete view of the issues.

This is not an argument about the spatial coverage of MBH98, it’s an argument about what selection procedures were followed in the pCA analysis. But it’s an argument about a nitpick because, as shown above, you get the same result whether you do PCA analysis or not, and W&A replicated Mann’s PCA analysis.

The argument was indeed a technical one, but it hardly “destroyed the paper.” It is a slight nuance that has no impact on the result. I’ve shown above the different between using centered and non-centered PCA in the reconstruction.

Didn’t impact the result in what sense?

In the sense that the paper’s results are identical whether you use centered PCA or not, or even whether you don’t use PCA at all. Mann was presenting a pioneering method of compiling paleoclimate reconstructions, so the discovery of a minute methodological issue is of minor interest to specialists in the field, but has no interest whatsoever to people mainly interested in the outcome of the work itself – i.e. the reconstructed global temperature profile back to 1400.

If you think scientific papers are supposed to be perfect then you are living in a fantasy world, and the presence of minor methodological flaws in otherwise seminal works hardly diminish their impact.

Completely correct. I showed above that the result is identical with centered or recentered PCA, or even no PCA at all. And highly relevant to whether the science is good, because it shows that it is a minor and unimportant methodological detail, not a major factor on which results hinge. So it’s good to have discussions about the best way to approach things, but nobody but the specialists need care much, certainly not to he extent of wasting a quarter of one’s life on it. Imagine if McIntyre had instead spent his time meaningfully contributing to the field?

Not surprising, though. MBH98 proved to be an absolute shambles in terms justifiable science with Mann (et al) refusing to justify the result, data, or anything, leaving it up to others to find his many errors.

Mann justified his methods and results – the paper passed peer review. Mann also meticulously responded to the various criticisms levied by people like McIntyre. And the results have been verified ad nauseam since that time by the multitude of other reconstructions that have been published showing the same pattern that Man found.

Just completely misses the point on science. The right answer for the wrong reason is bad science. You’re considering the conclusion not the workings. 

It doesn’t miss the point at all. If you take any scientific paper and pick it apart to the very bones, you will find minor issues that might have been done in a better way but which have no material impact on the result. It’s good if those things are identified and improved for future works, but no one expected that the paper was completely flawless to begin with, it was presenting a completely new approach to paleoclimate reconstructions, and it was a certainty that the approach would be refined.

The question that matters a quarter of a century later, unless you’re a specialist in the field who’s working on a paper, is whether the fundamental results of the study hold up, and they do. The result that Mann found is the basic pattern of change over the last few centuries.

Mann obstructed all attempts to have his work audited.

“Auditing” research papers isn’t a thing, and Mann was under no obligation to help the effort, although he did, to the extent that such help was reasonable and appropriate. Was he supposed to stop his entire career to focus on aiding McIntyre’s hobby project that has yielded zero substantive results in 22 years? For a paper that was already 5-years old at the time of McIntyre’s initial interest, no less?

The thing McIntyre should have done from the outset was to say, “here’s how I think a reconstruction should be done, and here are my results.” But he’s never attempted to do that, almost certainly because he realizes his results would be the same as everybody else (and what would he have to show for himself then?).

It absolutely is a thing and is what reviewers of the paper can and should do.

“Auditing” isn’t the job of peer reviewers, and the paper had already passed peer review some 5 years before McIntyre became interested in it.

Science is fundamentally forward looking, while “auditing” is fundamentally regressive. Publishing papers (or blogs) whose only reason for existence is to pick apart older papers is deeply unproductive, because at the end of the work you haven’t learned anything new, at best you’ve unlearned something old. This is why when scientists think there’s a better way to do an analysis, they just do the analysis the better way and present their results. McIntyre doesn’t ever do this, because his goal is not to advance human knowledge, it’s to cast doubt and aspersions.

For MBH98, Mann point-blank refused to help McIntyre whatsoever.

Not being overly helpful is not being obstructive. It is not surprising that Mann wasn’t particularly enthusiastic about helping the random layperson implicitly accusing him of fraud. Of course history has shown that the paper wasn’t destroyed and Mann’s results were robust, so his intuition was spot on.

Peer review is not an exhaustive forensic audit of a paper’s computational details and it is not intended to replicate every aspect of the analysis. Peer reviewers are expected to evaluate the paper’s overall soundness and suitability for publication: whether the described methodology (if implemented as detailed) is sound, whether it is relevant to the field, whether the conclusions are supported by the analysis.

McIntyre’s auditing exceeds any reasonable scope of peer review, as it is attempting to exhaustively reverse-engineer the minutiae of every computational details to find faults.

The proper way to verify the validity of studies like Mann’s is via independent replication, not auditing. Wahl and Amman replicated Mann’s results by following the detailed methodology, numerous other studies have used completely independent methodology and different data and arrived at the same conclusions. These approaches have provided vastly more valuable insight, and move the field as a whole forward, than has McIntyre’s 25 years of auditing.

 that means error free.

That is not the expectation at all, because this expectation would be impossible to meet. Peer review is not the final word in quality, it is the minimum barrier. Science by design is self-correcting over time. Future studies will implement new methodologies or apply old ones and make new findings, and improve the methods or supplant them.

Again, I’ll point out the fundamental issue that McIntyre’s work has not produced any new insights into the science of paleoclimate reconstructions. He has spent 25 years looking backward, when his time would have been vastly better served just making a reconstruction using the methods he deems superior and correct, and publishing those results. “Here’s why I think my methods are better and here are my methods and here are the results.” This is how science progresses; it is really how any field of human endeavor progresses.

That belief undermines the idea that peer review produces better papers that can be relied upon as sound science. You don’t mean it because I’ll bet you believe peer review is better than crowd review as happens against articles posted on forums.

I do believe peer review is better than crowd review – little is gained from having a group of non-specialists evaluating a work they don’t understand and can’t contextualize, and no one on the outside can assess the validity of the review because they have no way of evaluating the expertise of the reviewers (you can’t trust that they, at the bare minimum, know what they’re talking about).

This doesn’t mean that peer review is a failsafe mechanism for preventing errors, it is just a minimum barrier than needs to be met before a paper is deemed suitable for publication.

A bold statement that is wrong IMO. McIntyre brought rigor and actual statistics into the field and has been the cause of many changes in practice. If that weren’t true, why is he hated so much by those authors he takes apart?

What changes in practice? I don’t think it’s hard to see why he’s disliked by people he’s spent 25 years attacking in bad faith.

This is nonsense, some in the crowd are scientists who are formally trained in the field and are just as qualified as any of the actual reviewers to comment and some others in the crowd informally know more about the field than the reviewers.

The people who rely on the work having been looked over by competent experts cannot assume that random commenters on a blog are qualified to evaluate the paper. it’s of course possible that someone with competent expertise might see it, but not guaranteed, as it is in peer review.

McIntyre knows more about proxies than most. He’s spent a lifetime investigating them.

Then he’s well positioned to positively contribute to the field by publishing his own reconstruction, rather than hurling rotten tomatoes from the sidelines.

If community review were followed in practice, scientists would be unable to cite the work of their peers, because they cannot know if the work they are citing has been subjected to expert scrutiny. Maybe it has, or maybe it hasn’t, but they would have to go find the random blog thread about the paper and sift through the anonymous comments and find out the identities of the commentors and determine if they had adequate expertise or not. This system would halt the progress of science in its tracks.

That is why peer review is a minimum bar. It lets other scientists know that some people who know what they are talking about have looked the paper over (and I’m being a bit superficial in my description – peer review is typically quite rigorous and thorough).

Auditing is not a productive enterprise. There is nothing inherently wrong with reverse-engineering papers as a hobby, but it contributes nothing to the progress of the field. Again, if McIntyre is such an expert, he is denying the field that expertise by not performing research and publishing the results.