Over at Daily Kos, Mann says that the Nomads (this is a new candidate name for the Hockey Team) have moved on and that RegEM, as in Rutherford, Mann et al , is the new sheriff. At the end of the day, it’s hard to see how the RegEM method avoids any of the BàÆà⻲ger and Cubasch criticisms as a Daily Kos reader has pointed out (and Mann has yet to answer).
But today I’m going to visit another curiosity about Rutherford, Mann et al .
There’s some very very curious relationships between Rutherford et al. and MM03. Rutherford et al.  was submitted to Journal of Climate on July 23, 2003, at around the same time as MM03. As outlined below, I’ll bet dollars to doughnuts that the file "pcproxy.txt", the file at Mann’s FTP site (in Rutherford’s subdirectory) to which we were directed and now supposedly the "wrong file", was used in the original submission of Rutherford, Mann et al.  and was fixed up after MM03 pointed out problems with it.
The proxy files presently at the SI for Rutherford, Mann et al.  are identical to the proxy files archived with the MBH98 Corrigendum. They are undoubtedly consistent with the final version of Rutherford, Mann et al. . They are related to the MBH98 data, but the format is consistent not with MBH98, but with the later publication. Finally, as I speculated before, Rutherford, Mann et al  use all the original screwy PC series.
Screwy PC Series in Rutherford, Mann et al 
I’ll deal with these matters in reverse order. Early last year, based on the written description in the Rutherford prepring, I speculated that Rutherford, Mann et al  used all the screwy PC series in calculations supposedly showing that the screwy PC problems did not "matter" as follows:
It may seem impossible to believe that calculations using the original flawed PC series of MBH98 are now presented as evidence that the calculation errors do not "matter", but the evidence is incontrovertible once you wade through Rutherford et al. . The data set (as used in Rutherford et al. ) is not archived, so one has to rely on the descriptions in the text. The article itself states that 112 indicators were used in the AD1820 step and 22 indicators in the AD1400 step. These are exactly the same number of indicators as used in MBH98. There is no statement in Rutherford et al.  that the PC series were re-calculated to correct the erroneous PC method. Thus, at this point, the evidence suggests that the Rutherford et al.  dataset is identical to the MBH98 dataset. I’ll check this if and when the Rutherford et al. dataset as used is archived, but there’s not a shred of doubt in my mind on this point.
John A. is not blocked at Rutherford’s SI and archived the directory for me about 2 weeks ago and I’ve now taken a quick look at it. I’ve checked the above point and confirm that Rutherford, Mann et al  did not fix their flawed PC method, but used precisely the same data as MBH98 itself – flawed PCs and all. So how this shows that the PC problems did not "matter" is a bizarre argument even by Hockey Team standards. Of course, it’s been peer reviewed by Andrew Weaver so being screwed up presumably doesn’t matter.
Nature versus Nurture
Here’s another curiosity. The MBH98 data version archived at Rutherford’s website http://fox.rwu.edu/~rutherfo/supplements/jclim2003a/data/mbh98/ is exactly the same as the data version archived at Nature in connection with the MBH98 Corrigendum. http://www.nature.com/nature/journal/v430/n6995/suppinfo/nature02478.html . This applies even to the temperature dataset anomalies-new.dat, which goes from 1854 to 1993. (Rutherford, Mann et al. say that they used the HadCRU (not HadCRU2) dataset going form 1856-1998, but archived something different. Neither MBH98 nor even the Corrigendum correctly identified the data version archived as anomalies-new.dat. There’s probably another version floating around; it’s the Hockey Team and they’ve "moved on".)
The proxy versions archived in the Corrigendum were supposed to be the versions used in MBH98. While they are related to the versions used in MBH98, they are actually the versions used in Rutherford, Mann et al. . If you look at the form of the proxy data sets archived for Rutherford, Mann et al /Corrigendum , the data is collated for each step – see the files data1820.dat, data1800.dat, etc. The form and nomenclature of these files appears to be consistent with the file calls in Rutherford’s program, but do not work with the file calls in MBH98 programs archived last summer. So what was archived at Nature was not the data as used in MBH98, but the data as used in Rutherford, Mann et al . The data is obviously related but why wouldn’t they simply archive it in the form that they actually used in MBH98? I still can’t replicate their results exactly (nor can Ammann and Wahl, despite their claims to have "exactly" replicated MBH98). I doubt that much turns on this, but I can assure you that any business auditor would not be content with Hockey Team arm-waving and would insist on ironing this out. Little discrepancies are sometimes just that, but sometimes they are telltales for bigger problems. You never know what lies underneath a stone.
Now let’s reflect a little on an old question, which keeps getting re-opened by the Hockey Team – pcproxy.txt, the "wrong file". It sure looks to me like this was associated somehow with Rutherford’s Journal of Climate article.
Originally in April 2003, Rutherford directed me to a location on Mann’s FTP site containing a file pcproxy.txt, dated in August 2002 (about 8 days after the FTP site was set up on July 31, 2002). There was also a Matlab version pcproxy.mat and Rutherford uses Matlab (Mann used Fortran). A few months later, on July 23, 2003, Rutherford submitted the first version of Rutherford, Mann et al  to Journal of Climate. At around the same time, the phrase "pcproxy" occurred in a graph harvested from Rutherford’s website in 2002 by the Wayback Machine as reported here and shown again below. The figure shows that Rutherford is dealing with MXD data as well as MBH98 data – just as in Rutherford, Mann et al  originally submitted in July 2003.
The form of pxproxy.txt is exactly compatible with the Matlab formats of Rutherford’s other work. So doesn’t it seem like a teensy bit likely that Rutherford actually used pcproxy.txt – warts and all?
After publication of MM03 in October 2003 (so the timing is about the same), Mann said that the data version [pcproxy] had been prepared especially in response to my alleged request for an Excel spreadsheet. We’ve discussed this sordid story before, as Mann’s story was a complete fabrication (this has not prevented Crowley from publishing the fabrication in Eos). Obviously it wasn’t prepared for us (and we noticed the problems and re-collated the entire data set), but why was it prepared and how was it actually used?
I’ll bet dollars to doughnuts that Rutherford’s original submission to Journal of Climate (dated July 23, 2003 just prior to MM03) used pcproxy.txt, collation problems and all (and that what Rutherford linked us to was the data collation that he was using for that study.) Then when MM03 pointed out the collation problems in pcproxy.txt, Rutherford and/or Mann realized his oops and deleted the pcproxy file. Mann then commenced his disinformation about us using the "wrong" data, when, in fact, it was Rutherford, Mann et al in their 2003 submission to Journal of Climate that used the "wrong file"
I’ve discussed on many occasions how MM03 has stood up – see MM03 Scorecard here. All the big criticisms have stood up – especially about the incorrect PC series, although we were then unable to explain why the MBH98 PC series were wrong (but did so in MM05- GRL). The only points that might not have applied to MBH98 (and we acknowledged this promptly were the collation errors, which we speculated might have been introduced by Rutherford at a later date.)
Here’s where Rutherford and Mann get a little cheeky. Rutherford, Mann et al. [Journal of Climate 2005] stated:
It should be noted that some falsely reported putative “errors” in the Mann et al. (1998) proxy data claimed by McIntyre and McKitrick (2003) are an artifact of (a) the use by these latter authors of an incorrect version of the Mann et al. (1998) proxy indicator dataset…
You have to watch their language carefully. This is a little more craftily worded than Mann’s original allegations to David Appell (and some of his later public statements) and does not say that all the MBH98 errors identified related to pcproxy – only "some" of them and does not claim that we used incorrectly collated data in our calculations (although they slyly spread this disinformation elsewhere).
The irony here is that I’ll bet that the so-called "incorrect version" was the version used by Rutherford, Mann et al. in their July 23, 2003 submission to Journal of Climate. After MM03, they coopered up their calculations with a re-collation of the data. I’ll bet that the files presently archived in the Nature Corrigendum did not exist prior to MM03. In fact, they should be thanking us for pointing out a problem with the data that they had used in their Journal of Climate submission, although I won’t hold my breath waiting for a thank-you note.
BTW, in an earlier post, I pointed out that Rutherford, Mann et al. completely failed to discuss any of the problems raised in our 2005 papers. I wrote to Andrew Weaver about this, asserting that this constituted a lack of full, true and plain disclosure and ended up with very unsatisfactory answers as to why Rutherford, Mann et al. should be permitted to avoid discussion of PC problems which they knew to exist. Despite this lack of full disclosure, Weaver did not intervene and allowed a very incomplete record to be published. This ended up having some further effect as Houghton, in his testimony to Congress, credited Rutherford Mann et al.  with supposedly discussing our 2005 articles even though this was untrue. It is frustrating that Weaver did not require full disclosure and that the authors Rutherford, Mann, Bradley, Hughes, Jones, Briffa and Osborn failed to provide it.
The New Sheriff
Mann now says that they’ve been using RegEM for over 6 years. There’s strong, verging on conclusive evidence, that the collation problems that Mann denied as existing as to MBH98 have affected their RegEM collations. So there are 4 years of calculations before MM03 which could be affected by splicing and collation problems – all of which are additional to the PC problems and bristlecones. In this case, not only did they move on, but they deleted the files when they de-camped.