To the interested reader you are referred further up this exchange http://slashdot.org/comments.pl?sid=3303957&cid=42240095 [slashdot.org].

This entire exchange is a classic example of a Gish Gallop in which one fallacy after another is presented so rapidly it isn't possible for someone to address all of them at once. hsthompson69 will reply to this response with yet more fallacies in the hope of getting the last word. Unless a denialist like hsthompson provides you with an analysis of the full temperature record with a clear null hypothesis and the actual statistics necessary to back up their point I strongly advise extreme caution. If as here they make repeated conceptual mistakes in the maths then you know you are dealing with someone motivated to ignore the facts. He isn't looking for the science. He is looking to trip you up by making you fact check so many points at once the odds are good you will get one explanation wrong simply by a slip of the keyboard.

To the interested reader you are referred further up this exchange http://slashdot.org/comments.pl?sid=3303957&cid=42240095.

This entire exchange is a classic example of a Gish Gallop in which one fallacy after another is presented so rapidly it isn't possible for someone to address all of them at once. hsthompson69 will reply to this response with yet more fallacies in the hope of getting the last word. Unless a denialist like hsthompson provides you with an analysis of the full temperature record with a clear null hypothesis and the actual statistics necessary to back up their point I strongly advise extreme caution. If as here they make repeated conceptual mistakes in the maths then you know you are dealing with someone motivated to ignore the facts. He isn't looking for the science. He is looking to trip you up by making you fact check so many points at once the odds are good you will get one explanation wrong simply by a slip of the keyboard.

To the interested reader you are referred further up this exchange http://slashdot.org/comments.pl?sid=3303957&cid=42240095.

This entire exchange is a classic example of a Gish Gallop in which one fallacy after another is presented so rapidly it isn't possible for someone to address all of them at once. hsthompson69 will reply to this response with yet more fallacies in the hope of getting the last word. Unless a denialist like hsthompson provides you with an analysis of the full temperature record with a clear null hypothesis and the actual statistics necessary to back up their point I strongly advise extreme caution. If as here they make repeated conceptual mistakes in the maths then you know you are dealing with someone motivated to ignore the facts. He isn't looking for the science. He is looking to trip you up by making you fact check so many points at once the odds are good you will get one explanation wrong simply by a slip of the keyboard.

To the interested reader you are referred further up this exchange http://slashdot.org/comments.pl?sid=3303957&cid=42240095.

This entire exchange is a classic example of a Gish Gallop in which one fallacy after another is presented so rapidly it isn't possible for someone to address all of them at once. hsthompson69 will reply to this response with yet more fallacies in the hope of getting the last word. Unless a denialist like hsthompson provides you with an analysis of the full temperature record with a clear null hypothesis and the actual statistics necessary to back up their point I strongly advise extreme caution. If as here they make repeated conceptual mistakes in the maths then you know you are dealing with someone motivated to ignore the facts. He isn't looking for the science. He is looking to trip you up by making you fact check so many points at once the odds are good you will get one explanation wrong simply by a slip of the keyboard.

Your response does not contain the actual math needed to address this point. I asked you to do the actual math necessary or concede and you have failed to do so.

Good day sir.

For anyone else interested in your falsehoods. If the model proposed claimed the six sided dice never rolls three decreasing numbers in a row then a single instance of 3 numbers decreasing is sufficient, and one can select any three numbers one like. This model is more restrictive because it has failed to include a error term. Had one been specified such that 3 decreasing numbers in a row has some low probability then my reasoning would apply (and given that the sequence quote is not random odds are actually pretty decent one would not be able to rule out your model at alpha=0.05).

This entire exchange is a classic example of a Gish Gallop in which one fallacy after another is presented so rapidly it isn't possible for someone to address all of them at once. He will reply to this response with yet more fallacies in the hope of getting the last word. Unless a denialist like hsthompson provides you with an analysis of the full temperature record with a clear null hypothesis and the actual statistics necessary to back up their point I strongly advise extreme caution. If as here they make repeated conceptual mistakes in the maths then you know you are dealing with someone motivated to ignore the facts. He isn't looking for the science. He is looking to trip you up by making you fact check so many points at once the odds are good you will get one explanation wrong simply by a slip of the keyboard.

No sorry, you get one random sample. You lose sir. You want your 10,000 friends to have one sample? No problem, just make sure you do post hoc corrections. You will run out of independent points to sample though so you will need to factor that in to. I look forward to reading your analysis. If you cant do this because you don't have the math chops then concede this point. If you do not, I will ignore you.

If you have up to it feel free to go away rerun the a more appropriate statistical test on the entire data set. If you are looking for periods with anomalous periods of stable temperatures you could do equivalence testing with statistical parametric mapping, that might convince me. If you don't know what that is and you cant propose an appropriate statistical test I suggest leaving this to those of us who have more than a high school education in statistics.

Again keep in mind if you fail to answer my questions I will ignore you.

1) Could the trend over that entire data set have been cooling?
2) Given that we have been predicting global warming since the 70s with consensus emerging in the 80s how is the hypothesis that the temperature will increase not withstood falsification?

You get one random sample. Go ahead, grab yourself a random number generator and let me know. If you do more than one you have to do post hoc corrections. This is basic statistics.

"The idea that a specified falsification must be somehow *randomly* observed, rather than sought out proactively in the data is *really* silly :)"

No, you cant cherry pick, you need a suitable randomised design, you are deliberately misleading people either out of your own ignorance of statistics or because you have an agenda.

You didn't answer my question. If you don't answer it in your next reply I will ignore your response.

You know, there is being dishonest with statistics, and then there is being really dishonest with statistics and whoever crunched those numbers is doing the later.

Give me any time series with some noise on it and I can pick a N where for the past N years the trend has been flat statistically speaking.

A randomly selected 15 year time period with no warming excludes a fair few models, a cherry picked example does no such thing. Talk to me again in 2023 and compare with 2008 and I will listen.

So how about we cut the cherry picking crap and be honest. Over the entire contemporary temperature record (the one taken with thermometers so we don't get into a debate about proxies) what has happened to the Earth's temperature?

Global cooling has never been the consensus, in the 70s most climate scientists were talking about global warming. Seriously just go read a few of the journals and you will see this to be the case. Time magazines report on science is so poor I wouldn't wipe my arse with it and risk the bullshit stain.

Global cooling is covered in physics classes because the mechanism is easy to describe using basic thermodynamics and fluid dynamics. It makes an interesting example. It is also interesting because it shows us that things like snowball earth occur in our simplest models.

Don't know where you got this claim from (Penn and Teller maybe?) but seriously, stop watching the movies, stop reading the blogs, stop watching the mainstream media, turn off the radio, stop reading the newspapers and go pick up the journals. You only need to read about 15 papers to have a pretty strong understanding of the case for anthropogenic climate change and even if you aren't convinced at least you will be making informed arguments criticising the actual science instead of the pop culture version.

3 seconds of googling confirmed that turnout in St. Lucie county was about 70%. Did you fact check this before you posted it?

And at the moment we are at war with Eurasia. By at the moment I of course mean we have always been at war with Eurasia.

Agreed, in fact it is worse than that. Most of my old games now work better under Wine on linux than they do on say Windows 7. I'm not worried about my old games working on Linux, most of them essentially already do. It is the next game that is the problem.

You don't hit back when they are all around you, you hit back later, when you catch the bully on their own, and break both his ankles. And hands. And beat his head so badly he cant see through the swelling in his eyes. Then you do the same to anyone who said anything out of turn while the initial incident was occurring. And if the school let the initial incident go they let this one go too. Or the school send the aggressor and their allies to jail for the initial incident. Either is fine by me, free for all or treat children like adults. Zero tolerance just hands bullies one more tool to use against the oppressed.

Happy to say toughen up. But if we are going to have a free for all in schools lets have a real free for all. If I had been allowed a level playing field with the bullies then bullying would have stopped pretty quick, at least for the six months they spent being fed through a tube. As it was I wasn't allowed to use violence on them (I did, but I was constrained as to how effective a strategy I could employ) because I was 'one of the good ones'. Either treat children like adults where harassment and bullying result in jail time, or let the smart kids settle things themselves with whatever implements they choose and if a few of the bullies get maimed or die, then I guess they should have been tougher.
Personally I think treating children like adults is a much better system. You harass or attack another student? Six months in a specialised boarding school, no parental visits, no release until you meet certain minimum requirements, any behavioural problems extend your sentence.

Interestingly enough if you go back and look at his predictions I suspect you will find he is out performing his own estimates of how frequently he should be wrong. I say interestingly because I'm not convinced his probabilities are off since to evaluate that you need to have a large number of elections and look at how frequently his calls a state wrong, and we have two elections to look at. Two is not a large number.

However there is something even more interesting underlying this uncertainty. I'm not privy to the secret sauce in Silver's analysis, but I suspect the bulk of his uncertainties don't come from statistical error as you have with ordinary polling because he is aggregating so many polls, instead the main source of uncertainty for him is the random bias of the pollsters in totality, which varies a fair old chunk from year to year.

What this effectively means is each year you have to add a random 'the pollsters collectively screwed up this much in their turnout forecasts' factor, which is broadly speaking a constant shift in /all/ your polling averages. What this means is, if Silver is correct in his estimate of how variable this term is (and I see no reason he isn't since as far as I can tell he is basing it on historical trends), at some point he is predicting that he will screw up big time and several of those 20% chances states go the other way will come up at the same time, with his average performance matching his predictions long term. If Nate Silver is right, odds are pretty good one election in the next say 10 he will screw up monumentally.

That said quantified uncertainty is better than idiot pundits. To paraphrase xkcd who had a great strip on this on Wednesday: Maths, it works bitches!