atomic_fungus (atomic_fungus) wrote,

#6820: This is how you know the science is settled.

Michael Mann lost his defamation lawsuit.

Here's the basic rundown. Michael Mann published a paper stating there was clear and irrefutable proof that global warming was real, man-made, and happening now. It included his now-infamous "hockey stick" graph.

The people who don't just uncritically accept the assertions of the global warming crowd looked over the paper, the methodology, and the data that was available, and cried "foul". They asked for source code and unadjusted data and Mann said, "No."

Using the published version of the program, the compiled one, people found out that if you fed just about any data series into the program, the resulting graph would come out shaped like a hockey stick. This was pointed out to all and sundry, getting them labeled "science deniers" and worse.

Someone said that Mann's paper was a fraud. Mann--very foolishly--sued for libel.

Here's why it was foolish: to prove that you did not libel someone, all you need to demonstrate is that what you said is either true, or a matter of opinion. In order to prove that you have been libeled, you need to show that what the other person said about you is in fact false. The person who has the facts on his side wins. Simple, right? if you are perpetrating a hoax, and someone says so, it's the height of stupidity to sue that person for libel. Because then you need to stand up and show the court how your hoax is actually not a hoax. In Mann's case, the natural move for the defendant is to say, "Well, then, let's see the source code and the unadjusted data that supports your claim," and then you must put up or shut up.

When Mann brought the suit there was a lot of scoffing from the folks who think anthropogenic global warming (AGW) has not been proven at all (self included) for this exact reason. It was a bad idea fo him to sue because the defendant's move was obvious and Mann wasn't going to let the court have his source code and data; he couldn't, because then it would be a matter of public record and everyone could see it, and the jig would be up.

And of course that's what happened. Case filed, processes served, and the defendant's first move was to subpoena all of Mann's code and data. And Mann refused to provide it. And kept on refusing.

For eight years.

For most of a decade Mann has stalled and refused and delayed, until the court had enough of it, dismissed the case, and awarded legal fees to the defendant. Mann says he's going to appeal, but let's face it: the defendant's recourse here is simply to tell the appeals court, "Well, we made an eminently reasonable request and he fapped around for eight years and never complied with the court." Go right back to asking the code and data, which Mann cannot provide and have any hope of winning.

Because when you come right down to it, Michael Mann's entire "hockey stick" nonsense is nothing but fake science. Just like the entirety of AGW is.

And the real-world data backs that up. If you look at the graph it becomes obvious that all the models--or nearly all of them--are wrong. If you look at the past century or so, you see a pretty steady rise in atmospheric CO2 but temperatures are not increasing along with them, at least in 75% of that period.

Climatology is the only "science" where a theory which predicts reality 25% of the time is proven.

But for me, the biggest fail of AGW is how the models are built. Solar insolation is one number. Albedo--the reflectivity of Earth--is one number. The effects of clouds on heat retention is one number. The effect of humidity is one number. Globally, so that warming in the artic circle is treated the same as warming at the equator, even though they have vastly different numbers for solar insolation, humidity, and surface albedo.

The other dirty secret is that these programs must be run repeatedly because they never give quite the same answer twice, and so the runs must be tuned to come out right. This and that variable get tweaked, the other constant gets adjusted, and run it again. Some runs are rejected because their answers are obviously wrong (like the ones that lead to a five hundred degree surface temperature, or minus three hundred) but then they also reject the runs that don't show warming. And after a while they have enough runs accumulated showing warming that they can say what they've been saying: "The science is settled!"

The problem is that computer models can't prove anything. Most people don't get that; a computer can be programmed to say whatever you want it to say. Computer programs are only as unbiased as the people are who write them.

I mean, there's a famous example, so famous I don't even need to show the picture. "DEWEY DEFEATS TRUMAN" went the computer prediction.

We're better now at writing programs than we were then, but it remains true that computers cannot and do not magically produce "the right answer" without careful programming. The example I cited was not a problem with programmer bias (I don't think) but a simple failure of the algorithm.

But when you think there is a certain right answer to the question you're asking, and you're the one writing the program and tuning it to work right, you're not going to accept any output that goes against your belief.

You'll discount negative results as bugs.

* * *

The Democrat-media complex is now working on manufacturing a recession in order to beat Trump next year. I just have to wonder how well they'll do, considering that the public is a lot more skeptical of the press than it was in past years.

* * *

I should go outside and cut the grass, but I'm not feeling it. Problem is, if I don't do it today, I need to do it on a weekday. Get up early and cut it.

Maybe I'll do that.

  • #7558: Yeah, I thought that sounded kind of strange.

    What if they held an insurrection and nobody came? Wednesday night Mrs. Fungus was telling me all about how the news said there was going to be a…

  • #7557: Whose fault, exactly?

    Kid is ranked 62 out of 120 with a GPA of 0.13. What's his mother have to say? He didn't fail, the school failed him. The school failed at their…

  • #7556: Yakisoba night!

    I don't get to make it very often, but I saw a really nice piece of round steak at the store the other day, so I bought it. 1-1.5 lbs beef (round…

  • Post a new comment


    default userpic

    Your reply will be screened

    Your IP address will be recorded 

    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.