Indecision 2012

Perspectives on our world and our universe, how it works, what is happening, and why it happens. Whether by a hidden hand or natural laws, we come together to hash it out, and perhaps provide a little bit of education and enlightenment for others. This is a place for civil discussion. Please keep it that way.
Forum rules
1) Remain civil. Respect others' rights to their viewpoints, even if you believe them to be completely wrong.
2) Sourcing your information is highly recommended. Plagiarism will get you banned.
3) Please create a new thread for a new topic, even if you think it might not get a lot of responses. Do not create mega-threads.
4) If you think the subject of a thread is not important enough to merit a post, simply avoid posting in it. If enough people agree, it will fall off the page soon enough.
User avatar
collegestudent22
Redshirt
Posts: 6886
Joined: Fri Jun 08, 2007 10:02 am
Gender: Male
Location: Gallifrey

Re: Indecision 2012

Post by collegestudent22 » Tue Nov 13, 2012 6:00 pm

The Cid wrote:
collegestudent22 wrote:The accuracy of the numbers is not in question here - the prediction that Obama would get X electoral votes is a separate issue, and relies on - mainly - an assumption of continuity that tends to (but does not always) hold. The idea that a candidate has a certain percentage of winning the election, however, is nonsense.
The number you're pointing at is one of many numbers the site in question runs. They run these on the right edge of the site. To the left of them is an extensive blog that explains the logic, math, and margin of error behind those numbers.
Yes, and my argument is that some of the numbers are utter nonsense, logically. The math and margin of error merely obfuscate this logical fallacy of assigning numerical probability - a notion of frequency statistics - to non-repeatable events.
I don't see what there is to argue about. The sabermetrician got it right. His electoral college projection was pretty much dead on.
Did Obama win 85% of the repeated 2012 elections? Because these kinds of numbers are the ones I am disputing - the "chance of winning" and such things. It is one thing to analyze the polls and say something like "I guess that out of the total amount of electors 45 millions will exercise their franchise, 25 millions of whom will vote for Roosevelt." It is wholly another to then say "I estimate Roosevelt's chances as 9 to 1."

"For the comparison is based on a conception which is in itself faulty in the very frame of the calculus of probability, namely the gambler's fallacy. In asserting that Roosevelt's chances are 9:1, the idea is that Roosevelt is in regard to the impending election in the position of a man who owns 90 per cent of all tickets of a lottery in regard to the first prize. It is implied that this ratio 9:1 tells us something substantial about the outcome of the unique case in which we are interested. There is no need to repeat that this is a mistaken idea." And yet, I keep having to repeat it.
collegestudent22 wrote:(Or Vegas odds, as it were, which are not determined by analysis, but what people are willing to bet.)
Um, no. That's not how Vegas determines their odds. The original odds are always determined by analysis by professional handicappers who, were they not extremely good at this, would not be professionals anymore.
Right. Sorry. I was thinking the prediction market system instead. I contend, however, that the analysis of the handicappers is subjective, inherently non-numerical in the sense of "chance of winning", and based upon their own reasoning, which could be wrong. Further, if their opinion differs significantly from that of the gamblers, the odds would shift significantly by the interactions of the bets.
So you're arguing with ampersand about meteorology
The information I obtained was direct from NOAA. It isn't so much me arguing with amp about it, but the (other) professionals.

To quote: "the correct way to interpret the forecast is: there is a 40 percent chance that rain will occur at any given point in the area." This is basically an estimate based on the frequency of rain coming with the expected pattern, combined with the frequency estimate of the area. It is, however, not a unique combination of factors that is being analyzed - it differs in form from the "chance of winning" idea, as it is class probability of a similar manner to insurance classification.
Frédéric Bastiat wrote:And now that the legislators and do-gooders have so futilely inflicted so many systems upon society, may they finally end where they should have begun: May they reject all systems, and try liberty; for liberty is an acknowledgment of faith in God and His works.
Count Axel Oxenstierna wrote:Dost thou not know, my son, with how little wisdom the world is governed?

ampersand
Redshirt
Posts: 7404
Joined: Wed May 23, 2007 11:43 pm
Real Name: Andrew Kunz
Gender: Male
Location: Portland, Oregon

Re: Indecision 2012

Post by ampersand » Tue Nov 13, 2012 6:24 pm

You are correct, CS22, that is the strict definition...but you also have to keep in mind most forecast are made for an area which is what is means by "at any given point in the area." So the decision tree is sort of like this: will there be precipitation? If yes, how much of the forecast area will get precipitation? As if you had actually looked further in the discussion you would have seen this:
NOAA wrote:Mathematically, PoP is defined as follows:

PoP = C x A where "C" = the confidence that precipitation will occur somewhere in the forecast area, and where "A" = the percent of the area that will receive measureable precipitation, if it occurs at all.
So... in the case of the forecast above, if the forecaster knows precipitation is sure to occur ( confidence is 100% ), he/she is expressing how much of the area will receive measurable rain. ( PoP = "C" x "A" or "1" times ".4" which equals .4 or 40%.)

But, most of the time, the forecaster is expressing a combination of degree of confidence and areal coverage. If the forecaster is only 50% sure that precipitation will occur, and expects that, if it does occur, it will produce measurable rain over about 80 percent of the area, the PoP (chance of rain) is 40%. ( PoP = .5 x .8 which equals .4 or 40%. )
So what I said about 30% of the area will get rain and 70% of the area won't get rain is basically what's indicated here. Now, if you want nitpick with my simplistic explanation, I suggest you move this to another thread.

User avatar
The Cid
Redshirt
Posts: 7150
Joined: Fri Nov 05, 2004 12:23 pm
Real Name: Tim Williams
Gender: Male
Location: The Suncoast

Re: Indecision 2012

Post by The Cid » Tue Nov 13, 2012 6:44 pm

collegestudent22 wrote:Did Obama win 85% of the repeated 2012 elections? Because these kinds of numbers are the ones I am disputing - the "chance of winning" and such things. It is one thing to analyze the polls and say something like "I guess that out of the total amount of electors 45 millions will exercise their franchise, 25 millions of whom will vote for Roosevelt."
That's not what the "chance of winning" is there for. Think of it as margin of error. "We calculated all of these numbers, and this is what they suggest will happen in the popular vote and electoral college. However, because most of our numbers came from polling, there is a chance that the polls are systematically wrong. This is the likelihood of that happening." Were the numbers that stood roughly an 85% chance of being accurate actually accurate? They absolutely were.
collegestudent22 wrote:I contend, however, that the analysis of the handicappers is subjective, inherently non-numerical in the sense of "chance of winning", and based upon their own reasoning, which could be wrong.
You are correct to suggest that what handicappers do is not an exact science. There is a great deal of method involved, and numbers, and inside knowledge of the thing they are handicapping, but it is subjective in a way.

However, more often than not, they're right. Handicappers kick the crap out of gamblers, which is why even people who know an awful lot about football struggle to beat .500 against the point spread. I guarantee that one could, if they wanted to, calculate a rough estimate of a degree of confidence in handicappers. If I had to guess, it'd probably be about...oh...85%.
Image
Hirschof wrote:I'm waiting for day you people start thinking with portals.

User avatar
Arres
Redshirt
Posts: 2064
Joined: Thu May 12, 2005 4:38 am
Location: Pomona, Ca

Re: Indecision 2012

Post by Arres » Tue Nov 13, 2012 11:25 pm

collegestudent22 wrote:That's just something that could happen, and you can be more or less certain that it will, based on your understanding
Think of it this way:

I am COMPLETELY CERTAIN: 100%
Everything in between: My closeness to one of the two ends: X%
I am COMPLETELY CERTAIN IN THE OTHER DIRECTION: 0%

There is nothing wrong with attempting to define HOW certain you are. Your own argument says you can do this.
Image
Sheldon wrote:For the record, I am waaaay an adult. Like, super-way.
The Ponynati said:You cannot escape us. You cannot stop us. Soon all the world will bow down to the power of ponies.
The Cid wrote:...the text message is the preferred method of communication for prepubescent girls. Bunch of grown men sending digital paper airplanes to each other. Give me a break.

bagheadinc
Bay Harbor Butcher
Posts: 7928
Joined: Wed Jul 12, 2006 6:25 pm
Real Name: Matthew
Gender: Male
Location: Fruitland, MD

Re: Indecision 2012

Post by bagheadinc » Wed Nov 14, 2012 3:54 am

Guys, just give up. CS22 is obviously a genius who knows everything. I mean, he totally read something on a site that kind of sided with him, so he must be right.
Image

User avatar
collegestudent22
Redshirt
Posts: 6886
Joined: Fri Jun 08, 2007 10:02 am
Gender: Male
Location: Gallifrey

Re: Indecision 2012

Post by collegestudent22 » Wed Nov 14, 2012 5:29 am

Arres wrote:There is nothing wrong with attempting to define HOW certain you are. Your own argument says you can do this.
But the numerical idea is just metaphor. To say I am 80% certain is to say merely that I am mostly convinced. To say then that I am 70% certain of something else is only to say that I am reasonably certain, but less certain than the other thing. The numbers are nothing but a metaphor for comparison. They actually have no real meaning - just as it would be mere wordplay and metaphor for me to say that activity X provides me 10 units of "enjoyment", while activity Y only provides 8. All this means is that I enjoy activity Y, but I enjoy X more - the numbers are meaningless.
bagheadinc wrote:I mean, he totally read something on a site that kind of sided with him, so he must be right.
And studied probability and its application in detail for my degree, and read thick books on these sorts of things in my free time (yes, that's eccentric - deal with it).

To quote from one: "A probability value, initial or derived, can only be tested by means of a statistical experiment, i.e. by means of a sufficiently long sequence of observations."
Frédéric Bastiat wrote:And now that the legislators and do-gooders have so futilely inflicted so many systems upon society, may they finally end where they should have begun: May they reject all systems, and try liberty; for liberty is an acknowledgment of faith in God and His works.
Count Axel Oxenstierna wrote:Dost thou not know, my son, with how little wisdom the world is governed?

ampersand
Redshirt
Posts: 7404
Joined: Wed May 23, 2007 11:43 pm
Real Name: Andrew Kunz
Gender: Male
Location: Portland, Oregon

Re: Indecision 2012

Post by ampersand » Wed Nov 14, 2012 8:47 am

What did you get your degree at from the University of Wyoming, perchance, CS22?

bagheadinc
Bay Harbor Butcher
Posts: 7928
Joined: Wed Jul 12, 2006 6:25 pm
Real Name: Matthew
Gender: Male
Location: Fruitland, MD

Re: Indecision 2012

Post by bagheadinc » Wed Nov 14, 2012 10:29 am

collegestudent22 wrote:To quote from one: "A probability value, initial or derived, can only be tested by means of a statistical experiment, i.e. by means of a sufficiently long sequence of observations."
If only there were a source of data where people chose between Romney and Obama to provide some sort of sequence of observations which they could base their calculations on. Hey, I have an idea for the next election, why don't we poll people for this information numerous times before election day.
Image

User avatar
The Cid
Redshirt
Posts: 7150
Joined: Fri Nov 05, 2004 12:23 pm
Real Name: Tim Williams
Gender: Male
Location: The Suncoast

Re: Indecision 2012

Post by The Cid » Wed Nov 14, 2012 2:50 pm

collegestudent22 wrote:To quote from one: "A probability value, initial or derived, can only be tested by means of a statistical experiment, i.e. by means of a sufficiently long sequence of observations."
So you're telling me that margin of error is a number that makes no sense to you and should not exist by the definition of probability.

To review: you do not believe in margin of error. It it not a real number. If I'm incorrect about this, please let me know how the number in question differs from a margin of error, being that that is exactly what it is.
Image
Hirschof wrote:I'm waiting for day you people start thinking with portals.

User avatar
collegestudent22
Redshirt
Posts: 6886
Joined: Fri Jun 08, 2007 10:02 am
Gender: Male
Location: Gallifrey

Re: Indecision 2012

Post by collegestudent22 » Wed Nov 14, 2012 3:26 pm

ampersand wrote:What did you get your degree at from the University of Wyoming, perchance, CS22?
Computer engineering. A very statistics intensive program in some of the higher classes.
bagheadinc wrote:Hey, I have an idea for the next election, why don't we poll people for this information numerous times before election day.
A poll is not an election. Also, they are unique events that change over time - that's kind of the whole purpose of tracking polls.
The Cid wrote:
collegestudent22 wrote:To quote from one: "A probability value, initial or derived, can only be tested by means of a statistical experiment, i.e. by means of a sufficiently long sequence of observations."
So you're telling me that margin of error is a number that makes no sense to you and should not exist by the definition of probability.

To review: you do not believe in margin of error. It it not a real number. If I'm incorrect about this, please let me know how the number in question differs from a margin of error, being that that is exactly what it is.
A margin of error is no different from the basic probability value. In fact, it basically says "if we test this, we may find that we were wrong by X%". If we claim an 85% probability, and a 5% margin of error, this means that it might actually be anywhere from 8 to 9 times out of 10 the event occurs. This still must be tested by repeated experiment - the calculated margin of error could be wrong as well. (This is a big part of having a computer system that can classify objects on statistical criteria.)
Frédéric Bastiat wrote:And now that the legislators and do-gooders have so futilely inflicted so many systems upon society, may they finally end where they should have begun: May they reject all systems, and try liberty; for liberty is an acknowledgment of faith in God and His works.
Count Axel Oxenstierna wrote:Dost thou not know, my son, with how little wisdom the world is governed?

bagheadinc
Bay Harbor Butcher
Posts: 7928
Joined: Wed Jul 12, 2006 6:25 pm
Real Name: Matthew
Gender: Male
Location: Fruitland, MD

Re: Indecision 2012

Post by bagheadinc » Wed Nov 14, 2012 4:53 pm

collegestudent22 wrote:A poll is not an election. Also, they are unique events that change over time - that's kind of the whole purpose of tracking polls.
Exactly, that IS the whole point of tracking polls. To track public opinion of the candidates. They are calculating how likely people are to vote one way or another based on past performance.

For some reason you stuck on this one definition of probability, when in fact there are quite a few types interpretations of probability. No matter how much you say it isn't, forecasting is a type of probability.
Image

User avatar
The Cid
Redshirt
Posts: 7150
Joined: Fri Nov 05, 2004 12:23 pm
Real Name: Tim Williams
Gender: Male
Location: The Suncoast

Re: Indecision 2012

Post by The Cid » Wed Nov 14, 2012 6:51 pm

Student, just go to Silver's methodology page and read the damned thing, would you?

And tell us exactly what you're arguing, because at this point I'm starting to become confused. Are you arguing that the very act of trying to forecast the election is nonsense, or just this one number that you can't be bothered to read into at all?

Edit: Is Nate Silver a witch? A website investigates.
Image
Hirschof wrote:I'm waiting for day you people start thinking with portals.

User avatar
collegestudent22
Redshirt
Posts: 6886
Joined: Fri Jun 08, 2007 10:02 am
Gender: Male
Location: Gallifrey

Re: Indecision 2012

Post by collegestudent22 » Thu Nov 15, 2012 2:46 am

bagheadinc wrote:
collegestudent22 wrote:A poll is not an election. Also, they are unique events that change over time - that's kind of the whole purpose of tracking polls.
Exactly, that IS the whole point of tracking polls. To track public opinion of the candidates. They are calculating how likely people are to vote one way or another based on past performance.
Yesterday's poll means nothing today. Opinions have changed, people have different information, and so forth. You could feasibly have a poll today that says Obama will get 80% of the vote, and another one tomorrow that says he will get 20%. Hell, if you aren't absolutely careful, you could get both on the same day in the same area.

And they aren't calculating anything - they are obtaining data to extrapolate to say "if the election were today, then the percentages would be X% and Y%".
For some reason you stuck on this one definition of probability, when in fact there are quite a few types interpretations of probability.
Probability is a certain relation of frequency, or it is meaningless and arbitrary. If I told you that a dice had a 50% probability of coming up as 6 on the next roll, you would call me nuts - not just roughly certain of the next roll.
No matter how much you say it isn't, forecasting is a type of probability.
Not of unique, once-in-all-history events. Obama did not have a "90% chance of winning". Such a statement can only be metaphor for "I think it pretty certain Obama will win."
The Cid wrote:Student, just go to Silver's methodology page and read the damned thing, would you?
I'm aware of the methodology. I could give you a methodology for "2+2 = fish", and that wouldn't make it a sensible statement. He says things like "For instance, a candidate with a 7-point lead, in a race where the standard error on the forecast estimate is 5 points, will win her race 92 percent of the time." This is utter nonsense. She would either win, or not. There is no "92 percent of the time". Each election is different. Each election that meets those criteria differs in various causal aspects that prevent such a probabilistic element.

In short, you cannot treat human beings, along with their choices and actions, as probabilistic machines. Not even in large numbers. The very system of polls has its own issues as a result of this methodological treatment.
Are you arguing that the very act of trying to forecast the election is nonsense, or just this one number that you can't be bothered to read into at all?
I'm saying that such a forecast is just an educated guess, and putting numbers on it (like "90% chance") other the forecast that a numerical quantity will be a certain amount is merely confusing metaphor at best.
Frédéric Bastiat wrote:And now that the legislators and do-gooders have so futilely inflicted so many systems upon society, may they finally end where they should have begun: May they reject all systems, and try liberty; for liberty is an acknowledgment of faith in God and His works.
Count Axel Oxenstierna wrote:Dost thou not know, my son, with how little wisdom the world is governed?

User avatar
Deacon
Shining Adonis
Posts: 44205
Joined: Wed Jul 30, 2003 3:00 pm
Gender: Male
Location: Lakehills, TX

Indecision 2012

Post by Deacon » Thu Nov 15, 2012 3:01 am

http://abcnews.go.com/blogs/politics/20 ... as-failed/

In thinking about the champions of liberty, his lesson is a bitter one: "History has shown that the masses have been quite receptive to the promises of authoritarians which are rarely if ever fulfilled."

Please, let's change the political discourse in this country from deciding which master we prefer to what we can do to limit the power and authority of those masters. When we no longer depend on them, when our first question is what we can do about something rather than demanding to know what they are going to do about it, then we can shrug off the harnesses that yoke is to the benefit of the State and the few that control it.

"The best chance for achieving peace and prosperity for the maximum number of people worldwide is to pursue the cause of liberty. If you find this to be a worthwhile message, spread it throughout the land.”
The follies which a man regrets the most in his life are those which he didn't commit when he had the opportunity. - Helen Rowland, A Guide to Men, 1922

bagheadinc
Bay Harbor Butcher
Posts: 7928
Joined: Wed Jul 12, 2006 6:25 pm
Real Name: Matthew
Gender: Male
Location: Fruitland, MD

Re: Indecision 2012

Post by bagheadinc » Thu Nov 15, 2012 2:59 pm

collegestudent22 wrote:Yesterday's poll means nothing today. Opinions have changed, people have different information, and so forth. You could feasibly have a poll today that says Obama will get 80% of the vote, and another one tomorrow that says he will get 20%. Hell, if you aren't absolutely careful, you could get both on the same day in the same area.
Apparently they do. That or, as Cid pointed out, Nate Silver is a witch.
collegestudent22 wrote:And they aren't calculating anything
Yes, they are.
collegestudent22 wrote:Probability is a certain relation of frequency, or it is meaningless and arbitrary. If I told you that a dice had a 50% probability of coming up as 6 on the next roll, you would call me nuts - not just roughly certain of the next roll.
That is the classical definition of probability, not the only definition.
collegestudent22 wrote:Not of unique, once-in-all-history events. Obama did not have a "90% chance of winning". Such a statement can only be metaphor for "I think it pretty certain Obama will win."
Once again, it's not that he has a 90% chance of winning, but there is a 90% chance voters will choose him like they did in the various polls. And yes, people do change their opinion from poll to poll and from the polls to the election. If they didn't, he'd be able to give you an answer with 100% certainty.
collegestudent22 wrote:In short, you cannot treat human beings, along with their choices and actions, as probabilistic machines. Not even in large numbers. The very system of polls has its own issues as a result of this methodological treatment.
Yes, you can apply probability to humans. It's done all the time. As I've mentioned before, there is more than one definition of probability.
collegestudent22 wrote:I'm saying that such a forecast is just an educated guess
It is an educated guess. It's educated by past performance. And I'm going to say this again, because you seem to keep skipping over it, NOT the past performance of either Obama and Romeny winning a presidential election, but of how people will vote.
Image

Who is online

Users browsing this forum: Common Crawl (Research) and 0 guests