Thursday, September 19, 2013

Fake Climate Science Skeptic Finally Willing to Put Money Where His Mouth Is

Pat Michaels is one of George Monbiot's
Top 10 climate change deniers.
Back in January, the CATO Institute's resident climate science fake skeptic Pat "Often Wrong" Michaels wrote in the Moonie Times: "...it’s a pretty good bet that we are going to go nearly a quarter of a century without warming."

Being a life-long gambler who discovered and then lost the best bets ever--politics and climate at Intrade--I contacted him and asked, "How much?"

After much back and forth, I discovered that what he actually meant was that there would be no statistically significant warming from 1997 to 2021. OK, fine. Still a better bet than the don't pass line. So, we agreed to bet $250 on this:
Dr. Michaels is betting on no statistically significant warming (at the 95% confidence level) in the HadCRUTx data for the 25 year period starting in 1997. Scott is betting on at least that much warming.
Pat and I had the bet settled months ago, or so I thought, but I had said it was contingent on posting the bet in a public forum like a blog. Pat disappeared for a while, apparently having forgotten that little bit, so I sent him a reminder that was ignored. Finally, I called him out in the comment section of this Dr. Roy Spencer post, and he responded aggressively, calling me a liar.

Of course, he's wrong about that too. He did back out of the bet, and had refused to answer me until I called him out in a public blog full of his fellow deniers.

But whatever. Now we have our little bet. If he wins, I send $250 to Organization for Autism Research. If I win, he sends $250 to the Climate Science Legal Defense Fund.

Dr. Roy Spencer, official climatologist of
The Rush Limbaugh Show, whose UAH data
shows 0.138C/decade of warming since 1979.
Photo from Source Watch.
In all the flurry of finally getting Dr. Michaels to bet, I have managed to rustle up an offer from Dr. Roy Spencer (official climatologist for the Rush Limbaugh show) for another bet. Here's his first bid on this bet:
I’m also in discussions with Scott over betting on a trend that would be 1 standard deviation below the average model warming, which would be +0.162 deg. C/decade for 1997-2021, compared to the 90-model average of +0.226 deg. C/decade. 
I'm just a gambler looking for good bets, and I happen to have a pretty good understanding of the science. I'm going over the numbers and looking for advice before proceeding with the proposed bet from Roy.

What a shame we still don't have a new Intrade. I could be challenging a bunch of dittoheads to go put their money where their fat mouths are, and it wouldn't be for charity.

8 comments:

rfcNOLA said...

good for you, mr. supak!!
people who only blog or post memes all day just talk but don't do the walk.

lucia said...

I'm curious about this bet as it seem ambiguous.

How will the standard deviation be computed? Will it be computed using "red noise"? Some other ARIMA model? Or will it be computed using the variability across model runs (as in Fyfe, Easterling and Wehrner etc.)?

As the latter seems required for papers discussing whether the difference between model projections and observations is statistically significant, it seems to me that's become the standard definition of standard deviation for the purposes of deeming a trend statistically significant.

Scott Supak said...

Bet with Pat Michaels:

Dr. Michaels is betting on no statistically significant warming (at the 95% confidence level) in the HadCRUTx data for the 25 year period starting in 1997. Scott is betting on at least that much warming.

Potential bet with Dr. Spencer:

I’m also in discussions with Scott over betting on a trend that would be 1 standard deviation below the average model warming, which would be +0.162 deg. C/decade for 1997-2021, compared to the 90-model average of +0.226 deg. C/decade.


So, which bet do you mean? With the bet with Michaels, we're only worried about whether the warming in HadCRUT4 is statistically significant (p < 5).

If you mean the potential bet with Spencer, let me know which one is best?

lucia said...

The bet with Spencer is clear provided you have agreed on the observations data set. The bet with Michaels is not clear. You need to agree on a method for estimating ±95%. There is consensus on the method. Note that Roy wrote "But maybe Pat has a better method of computing the statistical significance than I do…I’ll let Pat and Scott work that out."

Pat used a method in the post here:http://www.cato.org/blog/ipcc-pretty-much-dead-wrong

That only goes to 15 year. but the method could be used to compute 25 years uncertainty intervals for the purpose of betting. A similar method could be applied to AR5 models and so on. You would then uses roughly 1.96* standard deviation on that graph for the 95% confidence intervals.

I haven't computed the odds on this bet for any choice of confidence intervals. But you need to specify the method. If I were betting, you'd be dang sure I'd compute the uncertainty intervals various different ways and figure the odds. But it's not my bet. That said my guess:

To improve his odds, Pat should bet using uncertainty intervals he uses when submitting his articles. That is the ones at http://www.cato.org/blog/ipcc-pretty-much-dead-wrong

To improve your odds, you should propose using 'red noise' over the period of the bet which is what Monckton usees here: http://wattsupwiththat.com/2013/06/13/no-significant-warming-for-17-years-4-months/

Most likely, unless a volcano erupts Monckton's error bars will be tighter than the ones in Michaels papers and will favor you. I will, of course advise Pat to stick to the ones he's used in his papers. Though, I don't correspond with him frequently, and so you might get in touch with him before he learns of my advice and manage to trick him into picking error bars he doesn't generally use.

FWIW: I have no idea why Pat would switch to using "Moncktonian" intervals instead of the ones he usually uses when interviewed on the radio. But if you manage to trick him into using Monckton error bars that will likely improve the odd for you.



Scott Supak said...

I don't want to trick anyone into anything. I want to use fair methods, because while I may not be a scientist, I can most certainly follow a debate, and I know who won this one.

I can also spot a good value proposition.

For instance, as Tamino noted today, we can pick a 15 year period where the trend was much higher, and "skeptics" said it was natural variability. We have a 15 year trend below average, but now it must be the models that are wrong.

This kind of disingenuousness from a debater is a sign of bluffing. It's a tell. They don't really believe their own BS. But I'll give Pat his due. He put his money where his mouth is (well, he publicly promised to do so, we'll see if he pays up). I called what I thought was a bluff. We'll see how his hand plays.

For the record, if anyone else wants to make public bets, I prefer Intrade type prediction markets where we can buy and sell shares on whether that prediction will happen or not.

In fact, if some rich person was paying attention, they could start a new Intrade right now in NJ, NV, or DE, b/c the Obama DoJ has said they won't mess with their new online gambling laws.

JD said...

JD Ohio says:
March 1, 2014 at 11:58 AM

Scott, I am willing to make a bet they will expose the fallacy of your bet. I will bet you that at any point 25 through 100 years from now (you choose the length) that a smaller percentage of people will be living in extreme poverty (if we can find another objective definition of poverty that is fine by me) than are living in it now. The fallacy of your bet is that it assumes that if there is a substantial rise in temperatures that it will cause harm to people.

Your assumption is wrong for two reasons. 1. Technology is advancing so rapidly that by the time that any large rises in temperature could occur, it will be a trivial exercise to manage the increase and potentially CO2. 2. Even assuming that there are no direct ways to manage warming, people will find a way to adapt.

Please note that my focus is on PEOPLE, not what you would consider to be proxies for things that could harm people, such as temperature rises, species extinction, acidification of the ocean et cet. Hansen evoked the Nazi death trains in his ridiculous scaremongering with respect to potential rising temperatures. So, let’s join the real issue– which is whether people will be harmed.

I will warn you ahead of time that this is sucker bet in my favor. Malthus has always been proven wrong because he did not understand technological innovation or the human capacity for adaption. Please let me know whether you are willing to put your money where your mouth is on an issue that really matters.

JD

Scott Supak said...

"fallacy of your bet"

What an odd concept. First of all, the bet with Pat Michaels isn't my bet. It's his. He said it was a "good bet" there would be no warming for 25 years. All I did was ask how much and offer the same bet to other deniers of basic science. If only I could bet on evolution...

But you consider a bet on future global temperatures fallacious, which is just silly. Your point seems to be that whether the planet gets warm or not isn't as important as whether mankind will be better off for it, or at least be able to adapt to it. This seems to suggest that you think the warming is real, ergo your unwillingness to bet on it.

But your counter offer is also bullshit. Ocean acidification is already happening. Pests and diseases are moving north. Soil tilth has been ruined by years of industrial/chemical farming. Fracking is sucking up what little is left of water in areas that are becoming more arid. Not only will we not be able to adapt to these self inflicted wounds, but the carnage will, of course, affect the poor more.

But if a bet can be said to be fallacious, it is yours that really takes the cake. See, the way I see it, the poor will be the most affected. This is also the way every study on the subject has seen it, including the famous Pentagon study.

For that reason, a great many poor people will be dying way before their time due to the continued rape and pillaging of the commons. Less poor people will, of course, mean a smaller percent of the world's population will be poor in the future (unless the affluent break the pattern of lower birth rates among the affluent).

This won't be because fossil fuel techno apologists found ways for them to cope with collapsing ecosystems on which they depend; it will be because we killed so many of them.

So since you won't bet on whether the warming is real or not, you can take your theoretical rapid free-market adaptation to warming on behalf of people who cannot afford to pay for the "solutions" and shove it.

Scott Supak said...

Pat lost the bet. AFAIK, he never paid. Then he died.

What a shocker. A science denier welched on a science bet he lost.

JD (above, who is known as Wilfred et al at Predictit) disappeared because in all our arguments here, or on twitter, or at Predictit, I owned his willfully stupid ass. My favorite was when he put up a web page that he said had 300 peer-reviewed studies that disputed global warming and I randomly sampled 20 of them that weren't reviewed and then found someone had gone through them all and not one of them was.