BS: 65% vs 35% of a Chance

Subject: BS: a 65% vs 35% Chance
From: meself
Date: 06 Nov 16 - 02:06 PM

It sounds kind of scientific - but what does it mean? (Reference to Hillary's vs Trump's 'chance' of winning). So, all you brainiacs: what, if anything, does 'a 65% chance' - or is it '65% of a chance' - mean?

Subject: RE: BS: 65% vs 35% of a Chance
From: DMcG
Date: 06 Nov 16 - 02:29 PM

That's a really good question. These things really refer to something tried out hundreds or thousands of times and what proportion have a specific result. That's fine for coin tossing or the chance of a poker hand winning but it doesn't really have much meaning for a single event like this specific election. Please spare us from running it another thousand times ...

Subject: RE: BS: 65% vs 35% of a Chance
From: Donuel
Date: 06 Nov 16 - 02:35 PM

I know what it means statistically in a literal sense but as a guardian of the esoteric, I would ascribe another meaning which is the number 65% representing a majority of people who are neither social psychopathic nor influenced by social psychopaths.

That being said, Hillary still needs a re-education beyond the overly simplistic gates of power and the notion of the banking system must control the helm of the world, navigating through war and peace.




not a brainier
but my number
is         
42
yoo hoo!

Subject: RE: BS: 65% vs 35% of a Chance
From: meself
Date: 06 Nov 16 - 03:16 PM

That's what I thought!

Seriously - I get how it makes sense for coin tosses and poker hands - and maybe even as a formula for placing bets at horse races - but for someone's 'chance' of winning an election?! Are they figuring in things like, 'the odds that Hillary will use a racist slur and alienate half her supporters are 1000 to 1' and 'the odds that she will collapse with exhaustion and alienate 27.8% of her supporters are 100 to 1' and 'the odds that Trump will change his mind on Muslims and gain 22% more votes are a million to 1' - or what?

Subject: RE: BS: 65% vs 35% of a Chance
From: gnu
Date: 06 Nov 16 - 04:01 PM

Trump is a piece of garbage. 110% for sure. The US citizens? The minority comprised of the ignorant and racist twits will not deter decent Americans from electing HRC. Will it be so for the Congress and Senate? THAT is the big question. Imagine Bernie Sanders as Senate majority leader... I have a dream!

Subject: RE: BS: 65% vs 35% of a Chance
From: McGrath of Harlow
Date: 06 Nov 16 - 05:21 PM

Of course if go in for the many world's notion, it's the ratio of universes where Hillary wins and Donald wins.

Of course except for one tiny little part those universes there'd be no difference whatsoever.

Subject: RE: BS: 65% vs 35% of a Chance
From: Doug Chadwick
Date: 06 Nov 16 - 05:23 PM

If you backed a horse that had 35% chance of winning, you would expect to get odds to get odds of 13/7 against - i.e. bet $1 in the hope of winning $1.85

If it has a 65% chance of winning, it would be an odds on bet of 7/13 and your $1 would only net you 54c.

This, of course, dose not take into account any trimming of the odds to make sure the bookmaker always make a profit.

DC

Subject: RE: BS: 65% vs 35% of a Chance
From: DMcG
Date: 06 Nov 16 - 07:00 PM

I think, Doug, the question was not what do the odds mean in terms of payout, but how do people come up with those odds in the first place. For a significant number of horse races where the bookies can balance an overpayment in one race with an underpayment in the other, by taking the past experience of the horses and its competitors into account they can come up with odds that suit their business.

But in this case we are talking of a single race, where Trump in particular is so different to previous candidates it is by no means certain past voting patterns are any guide at all.

Subject: RE: BS: 65% vs 35% of a Chance
From: meself
Date: 06 Nov 16 - 07:35 PM

For horse racing, I assume there is a fixed set of factors that are taken into account: the horse's record, age, condition of tracks raced on, who the jockey/driver is, etc., and then it is assumed that nothing else will affect the horse's performance.

But in the case of an election, the variables are almost endless.

Subject: RE: BS: 65% vs 35% of a Chance
From: Rapparee
Date: 06 Nov 16 - 11:41 PM

If you have flown 5,000,000 miles, as a passenger or a crew member, do your chances of being in a crash on your next flight increase?

No, as every flight is a discrete event.

What the pollsters determine -- assuming an unbiased set of questions to be asked -- is who will win the election (or whatever) at any given moment. A series of these, done through a period of time, will provide a median and mean which can give guidance but cannot infallibly predict the outcome of an event.

Enough of the median and means, combined with a statistical analysis of additional factors (economic, societal, etc.) can give an approximation of any event. Approximation, not certainty.

Subject: RE: BS: 65% vs 35% of a Chance
From: Stilly River Sage
Date: 07 Nov 16 - 12:03 AM

PLEASE - no more election threads

Use the ones that are already in place. Any more will be combined with existing threads.

Thank you.

Subject: RE: BS: 65% vs 35% of a Chance
From: Joe Offer
Date: 07 Nov 16 - 12:59 AM

I think it's worth looking at the odds set by bookmakers. They're professionals, and they stand to lose a lot of money if their calculations come out wrong.
If the odds are 65 to 35, it doesn't mean that one candidate will get 65% of the vote - I think we're all clear on that. I think the odds are slim that Hillary will win by 30 percentage points. I'd say the odds are good that Hillary will win by 4 to 8 percentage points. I wonder what odds the bookies would give me on that.
I'm not a betting man - I tend to lose...and somehow, I got through college without taking a course in statistics.
-Joe-

Subject: RE: BS: 65% vs 35% of a Chance
From: meself
Date: 07 Nov 16 - 01:31 AM

All very interesting, and I suppose it might seem to make a bit more sense if these percentages were announced as being linked to odds-setting by bookmakers, rather than as if those percentages mean something in and of themselves. If you told me there was a 100% chance that I would get hit if I back out of my driveway, I would know that I shouldn't back out of my driveway - but if you tell me there's a 65% chance I'd get hit, I would find that virtually meaningless. That would mean no more than that if I stay in the driveway, there's no chance I'll get hit, and if I back out, I'd better use caution, because there is a chance I'll be hit. But I really wouldn't know what the difference between a 65% chance and a 35% chance meant, or how much more or less cautious I should be ... !

Subject: RE: BS: 65% vs 35% of a Chance
From: DMcG
Date: 07 Nov 16 - 03:20 AM

What you are talking about there is how you should modify your behaviour given the odds. If you are actually placing a bet, it should affect you. For anything else - like skipping voting to go and visit your aged aunt as you have been promising to do for years instead of voting - the odds are too close and I would say they should not influence your behaviour.

Subject: RE: BS: 65% vs 35% of a Chance
From: Donuel
Date: 07 Nov 16 - 10:17 AM

I think bookmakers have more significance than coin flippers.

For Example the surprising results of the Brexit vote was due to the favorable polls to stay that came from London polls which were highly skewed to stay because of the many bookmaker-Investment Bankers centered in London. People expected the country to follow the smart money.

Switch to Washington, and land line phone polls vastly favor the status quo. No one expects rural America to turn out an unprecedented vote. If Trump was serious he would have leased 10o,000 buses and pick up trucks for rural America to get folks to cast their vote.


PS
Some of the religious right have come off looking like Family value hypocrites, more than usual.

I am not expecting surprising results but I do expect chaos and recriminations.

Subject: RE: BS: 65% vs 35% of a Chance
From: Bee-dubya-ell
Date: 07 Nov 16 - 11:32 AM

Bearded Bruce, please check your facts. There is NO federal prohibition against anyone with a criminal record running for office, much less one against someone under indictment. State laws vary, but no such federal law exists. If still in doubt, click here.

Subject: RE: BS: 65% vs 35% of a Chance
From: Rapparee
Date: 07 Nov 16 - 11:38 AM

Every throw of honest dice, every honest card deal, makes each a discrete event. In a game where "used" cards are discarded from active play the odds of getting a certain card are reduced as the number of "active" cards left becomes smaller. If at the beginning your chance of getting the jack of diamonds from a 52 card deck is 1 in 52, it becomes 1 in 51 if one card is taken from play, 1 in 50 if two cards, etc.

Being wrecked as you back out of your driveway is a 50-50 event: you will be or you won't. Each backing out is a separate event. However, if you back out in spite of seeing a drunken driver careening down the street, then the odds of getting smacked rise dramatically -- and it's your fault, not that of statistics, but your odds of backing out in such a situation are still 50-50 (you will or you won't).

The odds of a howling blizzard in July (in the Northern hemisphere) is 50-50 because there will be one or there won't be one.

And you can bet on all of this!!

Subject: RE: BS: 65% vs 35% of a Chance
From: akenaton
Date: 07 Nov 16 - 12:37 PM

Joe says that moderation here is not politically motivated, why then was my last post removed?.....Because it referred in an uncomplimentary way to Mrs Clinton?

The point of the post was to illustrate how the results can be distorted by publishing "odds" or percentage of chance.

Subject: RE: BS: 65% vs 35% of a Chance
From: McGrath of Harlow
Date: 07 Nov 16 - 12:39 PM

Wise bookmakers don't go so much by guessing the results, they follow the betting so that they will win whatever happens, and that is reflected by the odds as they change. So it's basically a poll of the changing expectations of those placing bets.

When bookmakers rely on guessing the results themselves, that's when they are liable to come unstuck.

When Jeremy Corbyn got elected leader the first time, the bookies had to pay out big for the few people who'd backed him at the start, but I doubt if they lost, because of all the people who had backed the losers who looked so much more likely. The same will apply if Trump wins.

Bookies very rarely lose money.

Subject: RE: BS: 65% vs 35% of a Chance
From: akenaton
Date: 07 Nov 16 - 12:50 PM

Bookmakers seldom lose Mr McGrath 60 years of Greyhound training and owning have taught me that :0)

The creation of "False favourites" is just one trick in their little black books. Percentage of chance is another way of manipulating the punter....bookmakers don't care who wins the race or the election as long as they have control of the odds.

Subject: RE: BS: 65% vs 35% of a Chance
From: DMcG
Date: 07 Nov 16 - 01:07 PM

Hmm, Rap, just because something may or may not happen - a binary event - does not make it 50:50. I will or will not die in a skydiving accident, but I can assure it is far more likely I die some other way.

Subject: RE: BS: 65% vs 35% of a Chance
From: meself
Date: 07 Nov 16 - 01:49 PM

Actually, the reason I had backing out in the back of my mind is that, when I was looking at houses a few years ago, one property I was considering would have had me backing out onto a busy street - I figured that if I bought the house, I would be backing out into traffic once or twice a day for the next twenty or thirty years - and the chances or odds were that sooner or later, I would get hit. Similarly, I recently changed the route I take to the grocery store, thinking that if I were to keep taking the narrow street with limited visibility at the corners, I would have a greater chance of being involved in some kind of accident than if I took a 'safer' route. But in either case, how the 'chance' could be reduced to a mathematical formula that could give a meaningful result expressed as a percentage is beyond me - as in the election, there are so many variables that it would be impossible to take even most of them into consideration.

Maybe I'm missing something, but I can get the idea that bookmakers can come up with odds on someone winning an election, but I don't get the notion of someone having a certain percentage chance of winning - even if it is just another way of saying what the odds are ... which I guess it is ....

Subject: RE: BS: 65% vs 35% of a Chance
From: Donuel
Date: 07 Nov 16 - 02:37 PM

Donald Trump has an election clock that runs backwards to election day at his home in Florida. 35% of Omen readers fear that it was ill timed for Donald to bring his clock in for a cleaning.
MsNBC



65% of Soothsayers agree Hillary will switch to flash paper post it notes rather than use e-mail.
GNN



Nate Silver says its close.
ZBS


The DOW is up 300 points today based on good news for Hillary
WSJ

Subject: RE: BS: 65% vs 35% of a Chance
From: meself
Date: 07 Nov 16 - 02:45 PM

Yes, but those percentages make sense - out of a total number of respondents, you have a certain number saying yes to a question, so it is obvious what the percentage in itself means (the implications are another thing, of course). But a percentage chance ....

Subject: RE: BS: 65% vs 35% of a Chance
From: Long Firm Freddie
Date: 07 Nov 16 - 04:45 PM

The New York Times uses an elections model (Upshot) to arrive at a percentage chance for each candidate. The page also gives a brief explanation of how other models work.


NYT

LFF

Subject: RE: BS: 65% vs 35% of a Chance
From: Donuel
Date: 07 Nov 16 - 05:52 PM

Great NYT article eh?
The last graph reminds me of PLINKO where Clinton has 2 chips and Trump has one.


They forgot the state of Russia :-)

On FOX news a Trump staffer responded to this article with "You can't believe everything you read in the NYT".

Subject: RE: BS: 65% vs 35% of a Chance
From: Rapparee
Date: 07 Nov 16 - 09:18 PM

Those are binary events: you may or may not die in a skydiving accident. The chances are remote if you do not skydive, but you could be hit by someone or something falling on you. When you view life (or whatever) as a binary event much changes.

However, life is not such. I find it interesting to blow people's minds by reducing the weather or something to an either/or scenario.

However, statistics and permutations CAN tell us something about overall trends: so many people will probably die in traffic accidents, for example. They cannot tell us who these people will be; they can tell us a certain percentage will be driving drunk.

Subject: RE: BS: 65% vs 35% of a Chance
From: meself
Date: 07 Nov 16 - 09:32 PM

So that NYT article makes it clear - I think! - that the percentage chance is derived from the results of polls - with no other variables taken into consideration (what is the chance that there will be a blizzard all through the north on election day? for example).

Subject: RE: BS: 65% vs 35% of a Chance
From: DMcG
Date: 08 Nov 16 - 03:14 AM

I agree about the binary event, Rap. It was just the association between that and "50:50" on which I demurred.

But your point about "the statistics cannot tell us who" is spot on. It is totally fallacious to reason from the statistical to the particular: if people who jog live x years longer than those who don't according to the stats it is fine to take up jogging but it is not fine to assume you will live x years longer.

And when we say one candidate is x% more likely to win than another we are following that fallacious reasoning.

Subject: RE: BS: 65% vs 35% of a Chance
From: BobL
Date: 08 Nov 16 - 04:13 AM

85% of statistics are made up on the spot. Including this one.

Subject: RE: BS: 65% vs 35% of a Chance
From: Mr Red
Date: 08 Nov 16 - 04:15 AM

Hmm.
When we cross the road, the chances of being hit are (say) 0.1%
But we look both ways.

99.9% v 0.1% is a level that could be ignored usually. But the consequences are way out of proportion. What level of binary probability we regard has to be predicated on the imagined result.

Politicians? Well it is 6 of one and half a dozen of the other in many respects, so maybe 52/48 is notso worrying. Or maybe it is the subject of endless debate and not a few threads within this parish.

And a whole swagbag of irony as the UK PM discusses trade and the Indian opposite talks immigration. Or am I being post ironic? (pun intended)