Statistikhengst
Diamond Member
- Banned
- #21
So you are suggesting or saying the poster is a child? And YOU need to educate him? YOU? Clown trash? YOU spammer?Oh, I see. Then you honestly believe that polls are never manipulated in any way to show bias? No one ever plays with the numbers to show a particular bias towards one side or the other? ............ oh, OK ............. I see your point. And, polls are always random, and never consist of targeted groups or locations? ........... well, OK then, if you say so .....Yea, and the polls are what, about plus or minus 50%?Can't wait to see the PPP Virginia poll that will come out in the next days, maybe even today. Wanna bet that Hillary is at least 10 points ahead of Trump in Virginia?
The damage he is doing is going to show up in polling in the next 3 weeks most intensively. I look forward to seeing all the interesting numbers...
No, +/ 2.8 to 3.0.
Nice try, you get a for effort.
Waidaminnit. You and I, we have never had a real adult conversation about polls, so I will enlighten you.
Any one, single poll can be wrong. Some are very wrong.
But two things happen: an average of all polls, from competing pollsters, usually comes pretty damned close to the truth, something I proved in 2008 and in 2012.
The composite polling averages for the polls in 2012 showed Obama 303, Romney 206, pure tossup 29. Obama won those 29 (Florida).
And the very most accurate pollster of 2012? PPP (D), which nailed 11 of 12 battlegrounds and 22 of the 23 end-polls from all states, including non-battleground states. It only missed NC, where it called a tie, which is automatically a miscall.
Rasmussen, on the other hand, missed SIX of TWELVE battlegrounds and grossly understated Obama's margins in 22 of 22 polls.
Back to Virginia:
PPP's final poll for Virginia, 2012:
Google Sheets - create and edit spreadsheets online for free.
Obama 51 / Romney 47, Obama +4
Actual result:
Obama 51.16 / Romney 47.28, Obama +3.87 (it looks like +3.88, but when you take the raw margin and divide by the raw total, it rounds to 3.87) - which is +4.
PPP absolutely NAILED it in Virginia in 2012.
And PPP's results, from November 4th, 2012, were practically identical to the exit polls as well. This is how good PPP was in 2012.
I analysed EVERY single poll from 2012 and did a mathematical comparison of every end-poll to that actual results, which you can read here:
Statistikhengst s ELECTORAL POLITICS - 2015 and beyond The moment of truth how did the pollsters do
And the excel table, again:
Google Sheets - create and edit spreadsheets online for free.
So, no, it's not what you say. It's honest, hard research, field work, and mathematical computations.
Therefore, based on it's track record, PPP is a pollster we can categorize as trustworthy, and actually, it's mathematical bias in 2012 was slightly to the Right, not to the Left.
So, the two things that happen:
1.) Composite margins tend to be very, very accurate and point to the winner in election polling.
2.) Polling firms build a reputation, a verifiable mathematical reputation, by comparing their end polls to the actual results.
It's that simple.
Better luck to you next time.
I am sorry, you just don't scream loud enough. You are unimpressive.
Have you tried 7-point and lots of colors, perhaps?