Sunday, September 28, 2014

More ClimateBall at Climate Audit

Steve McIntyre has a new post up at ClimateAudit. It's called "What Nick Stokes won't show you". It's a continuation of the smokescreen about demanding the unselected PC1s be shown with orientation favorable to a hockey-stick interpretation, using a hockey stick index (HSI), rather than as his program produces them. Again pretending that it's about Wegman aligning the orientation, rather than selecting the top 1% by HSI without disclosure.

He's made some pretty outrageous claims about how people here are, well I'll quote: "Some ClimateBallers, including commenters at Stokes’ blog, are now making the fabricated claim that MM05 results were not based on the 10,000 simulations reported in Figure 2, but on a cherry-picked subset of the top percentile. "

I wrote a substantive response to this, soon after the post appeared. It went into moderation - I've posted the text here. About four hours later it disappeared from the queue; I don't know what is happening there. Steve says he'll look in the morning. (It's here).

I reran the code to get some quantitative HSI numbers for the various cases, matching those described in detail here, and pictured here. It's a new run, not exactly matching. Here are the numbers, matching the cases described . For unselected sets, it's the mean absolute value HSI:

Decentered PCA (MBH98)        Centered PCA
Selected 100 out of 10000 by HSI     1.981.60
Not Selected1.610.65

The mean of 1.61 for unselected decentered (MBH) matches the mid-range figure in Steve's post. The difference between 1.98 and 1.61 made by selection may not seem so great, but these are like t-values. And it shows when centered but selected has almost the same mean HSI. The undisclosed selection is about as effective in creating HS appearance as decentering.

Emphasising the compression of the t-like HSI scale, centered unselected, which shouldn't have any HSI effect, and don't seem to, still show a mean (absolute) HSI of 0.65.

Anyway, below the jump I'll show various plots relevant to Brandon's contention that they should be oriented.
Here is the original Wegman tableau:

It is the selected version, with a mean HSI of about 1.98. I contrasted it with an unselected version. Brandon modified this to show the plots with positive HSI:
Unselected, as produced by MM05 program As reoriented by Brandon

He says that it makes a big difference. Judge for yourself. But is it a legitimate difference? It makes any kind of trend look like a HS contribution, by turning it up.

Here's another comparison - Brandon's reoriented MBH unselected versus centered but selected.

Centered PCA, but with 100/10000 selection as in M&M05 Decentered, unselected as reoriented by Brandon

As you'd expect from the above table, even with reorientation, they look similar. But the left has no decentering. The HS is created just by selecting the 100 top HSI out of 10000.


  1. I'd like to know how McIntyre has measured the temperatures over the last 2000 years to get those hockey sticks?

  2. Nick, Odd that Steve M is going after you on the HSI claim when he made no comment on DeepClimate's pointing out the HSI selection in Steve's code. I had the impression that since DeppClimate's discovery both he and McKittrick had gone totally quiet on the whole "red noise produces hockey sticks 99% of the time" line (I'm quoting McKittrick,). The sudden resurrecting of this claim is very strange. Can he really be so anxious to save the CEI's and Steyn's skins?

    I see Steve has replied to you, noticeably completely avoiding your challenge of how Wegman exactly reproduced one of M+M's cherry picks, instead going on at great length about t-square tests !!! This looks like a whole new smokescreen, though maybe I'm missing something: maybe I'm rather dim, but Steve's posts can be very hard to follow despite my acquaintance with the hockey stick "controversy". Do others experience the same difficulty?

    1. Bill, I've discussed this in a number of places, but given that McIntyre provided the code that DeepClimate used to understand Fig 4.4 of Wegman, you can hardly accuse McIntyre of hiding anything on this one.

      As I pointed out to Nick and others, where I to develop a new index that is supposed to catch a feature that I see in data, I would naturally plot examples with the high value of the index to ensure that my new index properly captured this feature of the data.

      I wouldn't publish this graph, and it does not appear that McIntyre did either.

      Anyway this a is perfectly reasonable check to do.

      If I wanted to understand the distribution of PC1 hockey-stick-indices generated from Mann's short-centered versus the standard centered PCA, I would then plot histograms of the hockey stick index to see how much bias, if any, was conferred by the short-centered PCA algorithm.

      As is seen by McIntyre and McKitrick Figure 2, it is clear that Mann's algorithm does produce a bias in PC1. That's the valuable figure, not what I consider the "throw-away" Figure 4.4.

      The t-test issue was brought up by Nick and McIntyre is responding to it. Not a smoke screen.

  3. Carrick
    "I wouldn't publish this graph, and it does not appear that McIntyre did either."
    It seems they submitted it to Nature. The sample of 100 is built into the structure of the code - I think so that they could read back for graphics. The selection seems fairly embedded too, though I guess it could have been an addition.

    But in some ways the more serious aspect was putting the selected PCs on the SI, without notification. I think this is how Wegman got caught. He claimed to have reproduced the results, but in fact they just came from this file. Though if he had re-run the code, it would still have had the selection.

    1. Carrick,
      I checked, and I've been mis-remembering about Nature. I was thinking of their "Comment on MBH98", where they described the red noise process and said "Ten simulations were carried out and a hockey stick shape was observed in every simulation.". No graph appeared, but I think they must have submitted one, as Ross M quotes a referees response. Whether it was selected, I don't know.

    2. It doesn't appear to matter:

      WIth Mann's short-centered PCA, they all look hockey-stick like.

    3. > they all look hockey-stick like.

      Are you suggesting they all indicate buy signals, Carrick?

    4. "It doesn't appear to matter:
      With Mann's short-centered PCA, they all look hockey-stick like."

      Well, some ladies go to much trouble enhancing breasts that already looked breast-like.

      On checking again, I mis-remembered my mis-memory about Nature. They did actually submit the GRL paper to Nature, and there does seem to have been a panel of 10, seen at least by them.

  4. You might also like:

  5. Nick Stokes = down and out

  6. How does the scale on the charts above compare to MBH98?

  7. Manual pingback:

    Interestingly, my later response was caught in Judy's moderation device:


    > It was disputed in the excerpt I blockquoted above.

    “It” refers to a red herring the Auditor constructed to feign having addressed Nick’s argument, which was

    As you’d expect from the above table, even with reorientation, they look similar. But the left has no decentering. The HS is created just by selecting the 100 top HSI out of 10000.

    The Auditor’s red herring about orientation only deflects from the fact that the whole argument rests on a peculiar notion of hockey stick.

    The “it” does not seem to be what eadler2 said.