Comments on: Both Sad and Hilarious http://www.coyoteblog.com/coyote_blog/2009/09/both-sad-and-hilarious.html Dispatches from a Small Business Sun, 12 Feb 2012 09:39:47 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Michael http://www.coyoteblog.com/coyote_blog/2009/09/both-sad-and-hilarious.html/comment-page-1#comment-23662 Michael Fri, 25 Sep 2009 18:05:28 +0000 http://www.coyoteblog.com/?p=9121#comment-23662 I get dinged by the spam filter if I try to post html, but at climate audit page id ?page_id=354, Steve lists presentations, commentary and correspondence of all the materials he has on the hockey stick I get dinged by the spam filter if I try to post html, but at climate audit page id ?page_id=354, Steve lists presentations, commentary and correspondence of all the materials he has on the hockey stick

]]> By: EscapedWestOfTheBigMuddy http://www.coyoteblog.com/coyote_blog/2009/09/both-sad-and-hilarious.html/comment-page-1#comment-23659 EscapedWestOfTheBigMuddy Fri, 25 Sep 2009 17:06:15 +0000 http://www.coyoteblog.com/?p=9121#comment-23659 Well, it has been a while since I read up on the subject, but originally Steve was however given the input data set by the original authors (Mann et.al), but did not use the same code. He tried to reconstruct the method from the paper (as task which is always harder than you'd think), and his reconstruction made a hokey stick of a generated-to-be-null test data set. Mann et.al disputed that McIntyre's approach was what they had done. I never read a claim that they could feed in null data sets and get null results, which I would have expected to put the whole issue to rest. This really bothers me BTW, so if someone knows of a paper on the matter, please point me at it. There was also some argument over the quality of the data set that used and over some of the preprocessing applied to the data, and over the method of presentation, and over the conclusions. Well, it has been a while since I read up on the subject, but originally Steve was however given the input data set by the original authors (Mann et.al), but did not use the same code.

He tried to reconstruct the method from the paper (as task which is always harder than you’d think), and his reconstruction made a hokey stick of a generated-to-be-null test data set.

Mann et.al disputed that McIntyre’s approach was what they had done. I never read a claim that they could feed in null data sets and get null results, which I would have expected to put the whole issue to rest. This really bothers me BTW, so if someone knows of a paper on the matter, please point me at it.

There was also some argument over the quality of the data set that used and over some of the preprocessing applied to the data, and over the method of presentation, and over the conclusions.

]]> By: Michael http://www.coyoteblog.com/coyote_blog/2009/09/both-sad-and-hilarious.html/comment-page-1#comment-23644 Michael Fri, 25 Sep 2009 03:36:27 +0000 http://www.coyoteblog.com/?p=9121#comment-23644 I wonder how Steve McIntyre got his hands on Michael Mann's hockey stick data and algorithm? Steve was able to show Mann's algorithm would create a hockey stick with any randomly generated data set. He went on to show Mann cooked the historical data set he used to create the hockey stick. I wonder how Steve McIntyre got his hands on Michael Mann’s hockey stick data and algorithm? Steve was able to show Mann’s algorithm would create a hockey stick with any randomly generated data set. He went on to show Mann cooked the historical data set he used to create the hockey stick.

]]>