Terence Mills does not believe his “forecast” and other hits

Before I go any further, here’s some hot bummping action from today:

Right, I’m glad I’ve got that out of my system. It was a glorious afternoon for it. And there will be more tomorrow; Christ’s get a shot at Kings, Pembroke get their’s at Jesus which could be exciting; and in the women’s world Jesus probably take the headship from Christ’s. I’d put money on that, if anyone cares to bet.

* No, Terence Mills does not believe his “forecast” says James. Where do these idiots come from? Essex it appears – say no more. ATTP was there first (but RT wins. No, not that RT, the other one:-). Thanks for NS for the pufferfish.
* Bernie Sanders Is Right On Wall Street: Just Not Right On How says Timmy, desperately trying to point out why Yet More Regulation isn’t the answer. Pearls before swine, I fear. BTW, did you know that Timmy is a professional Bernie Sander conspirator?
* Top 10 Tricks to Find That Temperature Trend You Always Dreamed Of by Inferno.
* Global surface warming continues without pause contrary to denier claims says Sou, which i kinda the bleedin’ obvious, but I thought I’d throw it in to burnish my sadly tarnished credentials.

17 thoughts on “Terence Mills does not believe his “forecast” and other hits”

  1. Nick et al –

    You’re misreading the Mills forecast if you think that. Mills’ forecast starts at Jan 15, meaning we now have 13 more months of data. In a clear case of being too clever by half, Gavin Schmidt overlaid Jan-Dec 15 for HADCRUT4 onto the Mills forecast and made the not-so-surprising discovery that the sharp spike from the El Nino-induced record warm of Dec. 15 broke outside of the Mills ARIMA 95% confidence boundary.

    Schmidt hastily declared the model had failed on this single data point, all the while omitting to mention that he left out Jan. 16….which, sure enough, brings the series back on track with Mills’ model. To make matters worse for Schmidt, Mills already anticipated this in his comments on the release of the model in noting the outlier effects that would likely occur because of the El Nino spike.

    Far from a breakdown in Mills’ forecast, 1 single outlier out of 13 observations is in keeping with a 95% CI boundary. So basically Schmidt only demonstrated that he doesn’t understand how to read an ARIMA output.

    [Ooh, flamebait. The details hardly matter; the model is so obviously drivel -w]


  2. but I thought I’d throw it in to burnish my sadly tarnished credentials.

    I thought we’d established that they’re at least as good as Bob Carter’s 🙂


  3. Maybe I should have added that ‘on track’ only applies to the ARIMA(0,1,3) model. For the other forecast based on segmented fits, HADCRUT4 needs to go down more than 0.1 degrees more over the next few months to get to the upper bound of the 95% confidence interval.


  4. LOL. Annan’s betting track record isn’t exactly a stellar one, though IIRC he has a habit of blaming its shortcomings on ‘bad luck’ followed by moving the goalposts of the original bet.

    [You’re missing the point. JA has an excellent track record of offering bets; and various people – the current idiot, Lindzen – have a track record of eating their words when pushed to put money behind them -W]

    But let’s propose another one instead.

    I predict HADCRUT4 trend line will remain within the 95% confidence boundaries of the Mills model through 2020, excepting a tolerable number of outliers consistent with a 95% CI.

    [That’s not a bet, that’s a prediction. It would be an offer of a bet if you were prepared to put money on it. Are you? -W]


  5. Well, then, Brent, surely you would dare to bet against James? Do note that he has lost one, but is well on track on winning another (10,000 dollar!) bet. He’s also well on track to win the bet against Lindzen, but we’d have to wait another 8 year before that one concludes.

    Who knows, maybe you can get him to accept a bet with your own premise (do make sure the select the right model, though).


  6. Funny. You think Annan’s track record should be evaluated not on the bet that he publicly lost, but on a bet you think he’s going to win 8 years from now. Your gambling approach is much like your climate modeling: ignore the past record while insisting that everyone accept your claims about the distant future in the present as “true.”

    If Annon wants to propose terms on the Mills ARIMA he’s free to do so. I’m simply predicting that HADCRUT4 will remain consistent with the 95% boundaries he has projected through 2020.

    [Words are cheap. JA put up money; you’re running away when challenged. Are you prepared to put money on your “prediction”? -W]


  7. Brent, you are focusing on a single bet with a mere 4-year window (made in 2007, prediction for 2011 – interestingly, Annan actually would have won the bet with the revised HADCRUT4 which has much better global coverage), and ignore the bet with the Russians (he’s well into winning that – let’s see if they ever pay that 10,000 dollar)

    Oh, and Annan actually also won a bet against Chris Hope, but that was just a bit of friendly fun (although it did make him 666 pounds richer – Hope made 1000 pound from Rudge though, and should in principle also get the same amount from Ian Plimer, but to the best of my knowledge he’s not honoured his bet yet).

    And of course the Lindzen bet was actually never formalized, since Lindzen wanted a 50-1 pay-out. That’s how ‘sure’ he was of his bet!


  8. I’m not sure what kind of casino you are running, but gamblers usually do not have the luxury of either (A) changing the goalposts on a bet they lost or (B) claiming victory on a bet that not only still has 8 years left on it but was never agreed to in the first place.

    So yes, given Annan’s track record of doing both of those things, placing money on the table before him seems to have much in common with purchasing speakers out of a van at a convenience mart. A bet itself may not be unreasonable, but he would have to propose terms of its execution that lack their current reputation as something one might expect to find at Donald Trump University.


  9. Brent,
    What has what you’ve said got to with setting up a bet now. Annan has a reputation for actually paying out, so that’s all that should matter. There’s no reason not to trust that he would honour the bet. I actually think what you predict would make an interesting bet as – I think – you actually have some chance of winning. Stoat seems keen, so what of it?


  10. Everett Sargent points out at Eli’s:

    Review status
    A revision of this discussion paper was accepted for the journal Atmospheric Chemistry and Physics (ACP).

    R.I.P. no more.

    [Well spotted. On a quick look, I didn’t see the final text up. Do you know if it is? -W]


  11. ATTP, Brent clearly is a chicken *and* can’t read. Already twice I pointed out the bet Annan has with the Russians (date for payout: 2018), and he ignored it, twice.

    Has Annan claimed victory over Lindzen? Not at all. He only pointed out that Lindzen was so insecure about the bet that he required a 50-1 payout. Has Annan claimed victory over Whitehouse? No he didn’t. He paid out, and only lamented his choice of the surface record to use for that short-term bet.


Leave a Reply to Hank Roberts Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s