• This is a political forum that is non-biased/non-partisan and treats every person's position on topics equally. This debate forum is not aligned to any political party. In today's politics, many ideas are split between and even within all the political parties. Often we find ourselves agreeing on one platform but some topics break our mold. We are here to discuss them in a civil political debate. If this is your first visit to our political forums, be sure to check out the RULES. Registering for debate politics is necessary before posting. Register today to participate - it's free!

Maybe the Polls are Rigged?

Not sure. Have you given up nitpicking? ;)

Whether or not the individual races were statistically dependent matters to your point.
 
Whether or not the individual races were statistically dependent matters to your point.
My OP acknowledges that these races are not exactly like coin flips. That limitation still does not explain how every race rated a toss-up -- more than two dozen -- went only one way.
 
My OP acknowledges that these races are not exactly like coin flips. That limitation still does not explain how every race rated a toss-up -- more than two dozen -- went only one way.

Earlier you said this:
... yet the GOP won every single one of those 27 races. How can that happen? I realize it's a bit simplistic to liken a close race to a coin flip, but the etymology of the term "toss-up" does come from the act of flipping a coin, so a 50/50 chance is clearly what is meant by the term. The odds of flipping a coin 27 times and having it come up heads 27 times in a row is 1 in 134,217,728 (i.e. 2 to the 27th power). A quick search of the web tells me those chances are roughly 268 times more remote than the chances of any one of you reading this message being struck by lightning this year.

So are we really to believe Cook was this wrong merely by chance?

You did quantify the simplicity of your assumption, but your 1-in-2^27 lower bound was profoundly useless. At that point you might as well have said that the probability that that would occur was at least 0%.
 
Whether or not the individual races were statistically dependent matters to your point.
I don't think so. I believe "toss up" means their prediction is in the margin of error and they can't make a call. Statistically, it seems quite odd that every single one the 27 races went one way.
 
I don't think so.

It very much matters. Do you know what statistical independence is?

I believe "toss up" means their prediction is in the margin of error and they can't make a call. Statistically, it seems quite odd that every single one the 27 races went one way.

Of course. The question is what led to this bias. But figuring that out is very likely to get mired in yet another issue to politically fight over. :confused:
 
It very much matters. Do you know what statistical independence is?

How would they be dependent? 27 different races from 27 different districts across the country.
Of course. The question is what led to this bias. But figuring that out is very likely to get mired in yet another issue to politically fight over. :confused:

Their models suck and their methodolgies suck.
 
I realize it's not a perfect comparison. The point stands: 27 "toss-ups" all going in only one direction by mere chance? That's a bit much to believe.

Your problem is only looking at numbers. Where were the 27 "tossup" races? Did they include the seven seats that flipped from blue to red? Numbers mean nothing if you don't know where they came from.
 
Ok, I know when the Republicans say "the media's polls are rigged!" it sounds like mindless Trumpism, but more and more a rational mind has to agree there is something to these complaints.

A case in point, The Cook Political Report bills itself as "... an independent, non-partisan newsletter that analyzes elections and campaigns for the US House of Representatives, US Senate, Governors and President as well as American political trends." Their studies are cited widely in the media as "non-partisan." (a quick Google will confirm this).

Here's the thing, on November 2nd, the day before the election, they published a report labeling 27 House races as "toss ups."

https://cookpolitical.com/ratings/house-race-ratings

... yet the GOP won every single one of those 27 races. How can that happen? I realize it's a bit simplistic to liken a close race to a coin flip, but the etymology of the term "toss-up" does come from the act of flipping a coin, so a 50/50 chance is clearly what is meant by the term. The odds of flipping a coin 27 times and having it come up heads 27 times in a row is 1 in 134,217,728 (i.e. 2 to the 27th power). A quick search of the web tells me those chances are roughly 268 times more remote than the chances of any one of you reading this message being struck by lightning this year.

So are we really to believe Cook was this wrong merely by chance?



"more a rational mind has to agree"


lol

the pollsters are 538 were wrong too. All the pollsters were wrong about this election.

We need to have a rational mind look at this.


.
 
Earlier you said this:


You did quantify the simplicity of your assumption, but your 1-in-2^27 lower bound was profoundly useless. At that point you might as well have said that the probability that that would occur was at least 0%.
By saying it was "simplistic" in the OP I acknowledged the comparison had limitations.

Say, you might be the right person to ask this question. Is "anal-retentive" properly spelled with a hyphen?
 
I don't think so. I believe "toss up" means their prediction is in the margin of error and they can't make a call. Statistically, it seems quite odd that every single one the 27 races went one way.
Based on many of the responses here, it doesn't seem odd when you've got left-wing media bias to defend.
 
Your problem is only looking at numbers. Where were the 27 "tossup" races? Did they include the seven seats that flipped from blue to red? Numbers mean nothing if you don't know where they came from.
27 of 27 means something, even if there were contiguous districts, shaped as perfect squares, and all beautifully bisected by the Mason Dixon Line.
 
"more a rational mind has to agree"


lol

the pollsters are 538 were wrong too. All the pollsters were wrong about this election.

We need to have a rational mind look at this.


.
Getting the election trend wrong is one thing. Batting .000 in close House races is quite another.
 
I think 60 or so posts into this thread it's been successfully demonstrated that, for some, any evidence suggesting the existence of liberal media bias is -- no matter what -- to be treated as an apostasy.
 
How would they be dependent? 27 different races from 27 different districts across the country.

If you don't understand what the implications of statistical dependencies are on poll results, then I'm afraid you're not qualified to proceed any further in this discussion. 🤷‍♂️
 
Based on many of the responses here, it doesn't seem odd when you've got left-wing media bias to defend.
Getting it so wrong, now two Presidential election years in a row, they've outted themselves, as has the press, as partisan actors. They've screwed themsleves as no one but the loons wil given them any serious consideration.
 
Last edited:
By saying it was "simplistic" in the OP I acknowledged the comparison had limitations.

That was a severe understatement.

Now demonstrate that you understand the effect of statistical dependency on election results.
 
If you don't understand what the implications of statistical dependencies are on poll results, then I'm afraid you're not qualified to proceed any further in this discussion. 🤷‍♂️
Deal. You're just too smart for the rest of us. Must be lonely.
 
If you don't understand what the implications of statistical dependencies are on poll results, then I'm afraid you're not qualified to proceed any further in this discussion. 🤷‍♂️
27 of 27. Anyone who thinks that is mere chance playing out cannot betaken seriously.
 
Complete conjecture, but let's agree that polling methodologies differ. Some might have a tendency to skew the results to the right, others to the left. If, for example, the NY Times sees that variability and prefers purchasing the polls that show liberal candidates performing better and you're a data provider who needs the NY Times as a client, what might you do?

If you look at the large media outlets -- i.e. the ones who are likely to spend the most for polling data -- I would argue they skew left. Of the major television networks, only Fox clearly leans right. CNN, NBC, ABC, CBS, MSNBC all lean left. For newspapers, it's not even close. The only truly national brand that doesn't lean left is the WSJ. All the other major dailies are moderate or far left (WP, NY Times, Chicago Tribune). The conservative newspapers are all second tier, IMO, the NY Post, The Washington Examiner, etc. Their smaller circulation means they are less interesting customers to content providers.

Perhaps polling data providers are skewing their results to the left in order to satisfy customer demand.

Oh please! If you want realistic polling information, you should look at the Polls page on Real Clear Politics and take the aggregate. Add that to the fact that the networks you cited also always announce polls other than their own - (FOX is fond of quoting Rasmussen, who seem to consistently skew much further to the right, which included having Mitt Romney winning against Obama right up until the election. BTW, the NYTimes consistently prints Rasmussen data too - pretty much shooting to hell your entire first paragraph.) The aggregate polling data includes names like Quinnipiac, Monmouth, Reuters/Ipsos, Neilson Brothers, Rasmussen, and many many universities who did local polling in their own states which concluded essentially similar data. As for the WSJ, they lean decidedly right - because they're also OWNED BY RUPERT MURDOCH! And frankly, exactly how influential do you think newspapers are nowadays?

The other glaring factor you seem to fail to get is that there's a huge difference between leaning left, and being opposed to Trump's emerging proto-fascism. The list of CONSERVATIVE outlets that slam Trump is long and deep and getting longer and deeper. The conservative newspaper New Hampshire Union Leader, for example, endorsed Biden-Harris, and it was the first time they'd strayed from endorsing the Republican candidate IN 100 YEARS! Nothing about voting Biden-Harris means skewing left. In fact, if anything, Biden has always been slightly right-of-center. That's why he's being targeted by the left.
 
Ok, I know when the Republicans say "the media's polls are rigged!" it sounds like mindless Trumpism, but more and more a rational mind has to agree there is something to these complaints.

Needs more tinfoil.
 
Checking with 2,500 people in a state with a population of over 10M has always seemed kind of silly to me. Add to that the fact that most people don't even know who their representatives are; expecting a valid prediction is dumb.

That's because you don't understand either statistics or probability.

Does is seem dumb to you that if there are 365 days in a year and you had 35 people (that's 1 person for every 10 days [or 10%] you would have an 83.2% chance that two of those people would have the same birthday?

I mean, 83.2 is MORE than 8 times as much as 10, isn't it?
 
Yes, but again, one would expect the errors to fall on both sides in something like equal measure. Why not here?

What the early analysis appears to indicate is that the "Biden Percentages" were pretty accurate.

What seems to have happened is that the "Undecided Percentages" included a substantial portion of "Shy Trump" voters. They weren't about to say WHOM they intended to vote for and the actual vote breakdown appears to indicate that they split MUCH more decisively than in the past.
 
Back
Top Bottom