|
Puzzle
Apr 5, 2010 6:37:58 GMT
Post by icefisher on Apr 5, 2010 6:37:58 GMT
|
|
|
Puzzle
Apr 7, 2010 14:04:57 GMT
Post by steve on Apr 7, 2010 14:04:57 GMT
Gee Icefisher there is plenty of evidence for a trend These deluded economonists were trying to determine whether temperature was instead governed by some sort of weird stochastic based variation as though the propensity for things to warm up might be determined by the same rules as those that tell us the amount of money old people hide under the mattress. There is not enough data to support the weird stochastic variation. Ask Lucia: LOL! Statistics is exactly how we measure the validity of observations. No statistics and you have no mathematical quantification of the likelihood of a hypothesis being true. So you are saying there is plenty of evidence for a trend but not enough to estimate? One cannot just ignore what statistics tells us because of a claim of too few data. I don't understand this. Whether there are "too few data" is determined by statistics. If statistics points to large uncertainty about a link between two things there may be too few data or there may simply be no link! Here we have the case in which there are too few data to determine whether it fits I(2), so there is definitely too few data to test the link. Hunter, As usual your point is confused by your penchant for emotive memes. But Lucia originally had the view that zero temperature rises till 2015 was not sufficient to reject 2C warming this century. She then came up with the ruse whereby she based her statistical tests starting from 2000/2001 because that was the nominal date for the start of the IPCC scenarios. The flaw is that she made her choice only 2-3 years ago. A bit like betting on the team that is already 3-nil up with 10 minutes to play.
|
|
|
Puzzle
Apr 7, 2010 16:05:45 GMT
Post by icefisher on Apr 7, 2010 16:05:45 GMT
LOL! Statistics is exactly how we measure the validity of observations. No statistics and you have no mathematical quantification of the likelihood of a hypothesis being true. So you are saying there is plenty of evidence for a trend but not enough to estimate? One cannot just ignore what statistics tells us because of a claim of too few data. Steve sez: "I don't understand this. Whether there are "too few data" is determined by statistics. If statistics points to large uncertainty about a link between two things there may be too few data or there may simply be no link! Here we have the case in which there are too few data to determine whether it fits I(2), so there is definitely too few data to test the link."In the real world when you have too few data to determine a items status then you either just automatically choose the more rigorous test or if that is not appropriate you gather more data before relying on any test. Fact is in this world when there is no penalty for being wrong people will rely on anything. You generally don't find that in the world of financial writeups because there are consequences for being wrong. icefisher, Lucia is not at all dismissing VS. She is doing what should be done- thoroughly reviewing the claim. Her thoroughness is why she has rejected the catastrophic AGW movement, and her ethical integrity is why she saw the significant failures of standards and ethics of the AGW promoters exposed in climategate Steve Sez: "Hunter, As usual your point is confused by your penchant for emotive memes. But Lucia originally had the view that zero temperature rises till 2015 was not sufficient to reject 2C warming this century. She then came up with the ruse whereby she based her statistical tests starting from 2000/2001 because that was the nominal date for the start of the IPCC scenarios. The flaw is that she made her choice only 2-3 years ago. A bit like betting on the team that is already 3-nil up with 10 minutes to play."Hey its a good bet! Go with it if you can find a sucker to take the other end. LOL!
|
|
|
Puzzle
Apr 7, 2010 16:41:31 GMT
Post by steve on Apr 7, 2010 16:41:31 GMT
Icefisher,
There is already a perfectly reasonable test being used that shows that it is warming in line with the models.
The proposition is that the temperature is either warming because the earth is being heated by CO2, or the temperature always wobbles about following an I(1) (not I(2) as I said before) distribution and the warming is due to this. Since the statistics are not rigorous enough to determine the latter, and since there is no physical reason proposed for why such an I(1) distribution should be followed, I hereby choose to ignore the latter proposition. That does not stop me from ignoring the former proposition because the propositions are independent.
Given LTCM, ENRON and the credit crunch, thanks very much for the coffee spluttering moment there. But you have it the wrong way around. You are the one that is relying on the dodgy statistical method here.
|
|
|
Puzzle
Apr 7, 2010 16:44:18 GMT
Post by steve on Apr 7, 2010 16:44:18 GMT
Icefisher
It's a good bet if the game ends in 10 minutes.
But if it's the start of the season, and it's Hull beating Man Utd, I'd still bet on ManU being above Hull by the end of the season.
|
|
|
Puzzle
Apr 7, 2010 17:54:06 GMT
Post by icefisher on Apr 7, 2010 17:54:06 GMT
Icefisher, There is already a perfectly reasonable test being used that shows that it is warming in line with the models. The proposition is that the temperature is either warming because the earth is being heated by CO2, or the temperature always wobbles about following an I(1) (not I(2) as I said before) distribution and the warming is due to this. Since the statistics are not rigorous enough to determine the latter, and since there is no physical reason proposed for why such an I(1) distribution should be followed, I hereby choose to ignore the latter proposition. That does not stop me from ignoring the former proposition because the propositions are independent. And before doing any work because of your 2 guesses (assumptions) you have about a 75% chance of being wrong. Did you then factor that into your confidence intervals?
|
|
|
Puzzle
Apr 8, 2010 9:33:28 GMT
Post by steve on Apr 8, 2010 9:33:28 GMT
I can only assume that you paid Standard and Poors to come up with that 75% figure.
|
|
|
Puzzle
Apr 8, 2010 20:48:00 GMT
Post by icefisher on Apr 8, 2010 20:48:00 GMT
I can only assume that you paid Standard and Poors to come up with that 75% figure. Obviously you haven't ever tried to sell that kind of work to Standard and Poors.
|
|