Then I suggest you go complain to McIntyre. He’s the one who made the charge. Was it not secondary then? It’s OK for McIntyre to raise a secondary issue, but to refute it is an attack on McIntyre–because it’s secondary? WTF.

And what exactly makes refuting McIntyre’s claim “some kind of bludgeon to attack McIntyre”? Any attempt to refute any claim is using a bludgeon to attack the person who made the claim? Double WTF.

on which reasonable people may disagree, that tucking the truncated series under another series makes the truncation less conspicuous

There are many things on which reasonable people may disagree. This ain’t one of them. It would take a most unreasonable person to actually look at the two charts and still make the claim you have just made. The effect of this “tucking” is literally invisible. It cannot be detected by normal human vision. I defy you or anyone else to look at DC’s two versions of the chart and say with a straight face that the difference is detectable.

This isn’t a reasonable disagreement, it’s a knee-jerk defense of McIntyre. I’m thinking you never even looked.

That hurts.

“What I am *that* upset about is not the pixels, it’s your moronic posts.” – CD

What upsets me is that you try to use something on which reasonable people may disagree, that tucking the truncated series under another series makes the truncation less conspicuous, and attempt to use it as some kind of bludgeon to attack McIntyre and defend Mann, when it is clear to everybody, including yourself, that the issue is secondary.

“At least NiV stopped after just one very brief attempt to bring “Mike’s Nature trick” into it. One attempt I can chalk up to an honest mistake, maybe from not reading the posts quite carefully enough.” CD

Sorry, I am passive aggressive. “Passive aggressive” is what control freaks call people who don’t want to be controlled. I didn’t notice under terms of use where it was stated that you get to set the terms and bounds of debate. Maybe you can cite that passage.

They are well-known to often show 20th century growth spurts that are not related to temperature, because there’s no matching rise in the local temperature records where they grow. There are several theories as to what causes the sudden surge. The most common one is fertilisation, either from CO2 or acid rain. When Greybill and Idso collected the American tree ring series that Mann used, they were trying to prove this theory, and deliberately selected strip-bark bristlecones because they were thought to be more sensitive to CO2. They reported in the paper that Mann got them from that they were not temperature-sensitive. Steve McIntyre, on the other hand, has an alternative theory that it is to do with the strip-barking process – when bark is stripped from one side of a tree, the cambium on the other side grows faster to compensate. This was based on his research in the field looking at the highly asymmetric, non-circular rings that resulted – meaning that you could get entirely different-looking series from the same tree, with one showing a surge and another not. I suspect it is a combination of things.

But whatever the cause might be, it’s not temperature. Unfortunately, the statistical method can’t easily tell the difference. It sees a 20th century rise, matching the 20th century global thermometer rise, and it lines the two up. It shifts and scales the tree rings to give the best match in the calibration period, and then the rest of the shifted/scaled series is used for the reconstruction. (If it sees a 20th century fall, it will still match, but now with a negative coefficient that will turn the series upside down.)

Of course, statisticians have a standard way to deal with this problem. They hold back some of the temperature data they have – called the verification period. Then after they have done the calibration on the rest, they check the correlation between the reconstruction and the temperature in the verification period. In other words, they try to use the reconstruction to make a prediction and test it. There are a range of correlation metrics, but the best known is Pearson’s r-squared correlation coefficient – also known as R2.

When Mann calculated it for his reconstruction, it came out about 0.1 or below, which is too small. It showed that the reconstructed relationship was spurious. So Mann searched around to find another metric – an obscure and little-understood metric called reduction of error (RE), and carefully didn’t mention the adverse R2 results. (Apart from for the 1820 step, where the R2 value was a little higher.) Unfortunately, the standard version of RE doesn’t work well on autocorrelated time series, so Mann got the wrong significance threshold, making it look significant when it wasn’t.

If you take a look at the NAS report, or the Wegman report, and search on ‘Bristlecones’, they give all the citations and expert opinions regarding their use.

But since there are relatively few tree ring series that show strong hockeysticks, with bristlecones being one of the most important of them, they pretty much have to include them if they want to get a hockeystick. (At least, they did until Yamal came along.) As they say, “You need to pick cherries to make cherry pie”.

Go read #34, which is everything I initially said about this. Does it sound *that* upset to you?

Then read your #36 where you instantly go off the rails and start with your condescending instructional material about something that is not the subject at hand (and of which I am fully aware, thank you very much).

What I am *that* upset about is not the pixels, it’s your moronic posts.

At least NiV stopped after just one very brief attempt to bring “Mike’s Nature trick” into it. One attempt I can chalk up to an honest mistake, maybe from not reading the posts quite carefully enough.

You, on the other hand, seem willfully obtuse. At one time I thought you were a serious person. You’re not. You’re just another troll.

Sweet Jesus. Do you do this intentionally, or are you just thick?

Here, I will put as many different kinds of emphasis on this as I can:

OF COURSE I DON’T THINK THIS WAS MCINTYRE’S ONLY COMPLAINT. BUT IT WAS THE SUBJECT OF THE DC POST AND IT WAS THE SUBJECT OF MY COMMENT.

McIntyre had two basic problems with the chart.

1. The Briffa line was truncated.
2. The Briffa line was “artfully hidden” under the other lines.

The DC post was not about item 1. My comment was not about item 1.

The DC post was about item 2. My comment was about item 2.

Item 1 is obviously the more important of the two items. This does not mean that we are permitted to talk only about item 1. This does not mean that in talking about item 2 we are unaware of item 1. It simply means that we are talking about item 2.

DO YOU UNDERSTAND NOW?

So, you think that the reason McIntyre wrote his post was only about the couple of pixels? I guess you do. Holy cow. McIntyre’s claim, whatever DC said, was that by artfully burying the line, it was less apparent that the series had been truncated. You did get that that was McIntyre’s argument, didn’t you, even if you do reject it? So, logically, and I know this is hard for you, the only reason those pixels could possibly have mattered was because Mann chose not to display the more recent data which contradicts his hypothesis.

Even DC acknowledges that the pixel thing is only one step in his argument. A flawed step too because he declares a *subtle difference* to be *no difference*, which, whatever his or your opinion, is not true.

Since both papers used Bristlecones, when everybody knows (as a result of the Mann fiasco) that it is strongly recommended that you shouldn’t, one can say immediately that these are not series to take seriously.

What’s the problem with Bristlecones?

Et tu?

The DeepClimate post was not about “Mike’s Nature trick”. Neither was my comment. The subject was McIntyre’s silly complaint that the end of the Briffa line was “artfully tucked under” one of the other lines. This affected one or two pixels and cannot be detected by ordinary human vision.

How friggin’ difficult is this to understand?