Good evening again, Frank! > Good evening again, Steve! > > Steven Thompson wrote to Lowell C. Savage... > > > So, that's my two cents, tell me yours... How would you work to > eliminate > > the total sell-out of our most precious right, our right to vote in a > > democratic, free election? > > I'm wondering, honestly, just how 'honest' elections, even local > elections, really are anyway. Lowell made some remarks earlier > on that elections should be viewed in much the same way as > lotteries are viewed; and I have to disagree.
I'm sorry Frank, but you're confusing what is being analyzed with the tools used to analyze it. Your "evidence" that the elections weren't "honest" (you chose the word "honest", not me) consists of apparent statistical anomalies. I've compared them to other things (like lotteries and having an even-dollar bill at the grocery store) that are perhaps more common experiences and (I was hoping) easier to grasp. That doesn't mean that the various phenomena are "the same." It only means that we can use the same sorts of mathematical tools to analyze them. Two plus two equals four, whether we are talking about 2 couples going out to eat, or talking about front and rear vehicle tires or explaining the term "four eyes." The "two plus two" bit is a mathematical method or tool. The examples are applications of that tool to real life. Statistical methods are the same sort of thing--only more complicated. > I know, I raised the bait to some extent here when I wrote what > the 'odds' might really be for a string of events occuring all at > the same time. When I wrote that however, it was not intended to > be identified with a lottery event, but rather trying very hard > to define the probabilities of certain circumstances happening, > all at the same time, in a local election in District 1, in > Idaho. I probably should have guessed that Lowell would raise > the bar and compare the results with a state lottery or something > along these lines. Yes, I understood what you were doing and I was trying to point out that the probabilities you were looking at were very small relative to some other things. > The difference here is that political choices require, yes, a > choice. It's not a random gamble. The answer, on the ballot, is > supposed to be, or at least entail, a principled 'choice' on > political alternatives. Well, we use statistical methods to analyze a lot of "mass human behavior" and it works quite well. Whether we are talking about murder rates and the effects of various laws, polling data about the President's "favorability rating" or whether the public holds used car dealers, attorneys or politicians in the most contempt. Every one of these things involves a "choice" by individuals. Yet we know so little about the causes of such choices that our best tool for analyzing them is to use statistical methods. > When such choices are so equally divided as they were in Bonner > County, that leave a lot of open questions, some of which I have > entertained earlier tonight. Ah, when we now start talking about choices made by people, we should consider how close so many political races are. Sometimes Democrats win, then later Republicans, then Democrats, etc. The numbers seem to hover around some sort of rough parity. Perhaps then, the real surprising thing is that an exact split seems to be so rare--even in a portion of a district. (And as I pointed out, your "multiple of ten" thing basically amounts to a 1/100 chance.) If there were more than 100 third-party candidates across the nation in races where their district covered two counties, then we would expect that one of them would experience the sort of result you did. It's quite reasonable to expect that there were more than 100 such candidates. Therefore, we should only start getting surprised if there are several candidates who had your situation. Lowell C. Savage It's the freedom, stupid! Gun control: tyrants' tool, fools' folly. _______________________________________________ Libnw mailing list [EMAIL PROTECTED] List info and subscriber options: http://immosys.com/mailman/listinfo/libnw Archives: http://immosys.com/mailman//pipermail/libnw
