4th and Inches said:
Adriacus Peratuun said:
4th and Inches said:
Adriacus Peratuun said:
4th and Inches said:
KaiBear said:
4th and Inches said:
KaiBear said:
LOL
Perfect
At least you MAGA boys are consistent.
Yet if Harris wins you will be stunned, wondering why all the soccer moms jumped ship .
i know alot of real soccer moms and they voting Trump..
Now crazy ass liberal nuts you would label as a soccer mom like my relatives wife.. yeah, she voting Harris but she voted Biden so thats not an improvement for Harris.
Harris needs to gain voters to offset the ones she is losing in the minorities
Trump is behind even on the FOX poll.
At this point I suspect 95% have already made up their mind on who they are going to vote for.
If Trump wins it's going to be a big surprise.
ignore Fox polling.. complete dogsht
Interesting the guys at the top of that list lean heavy to Trump and the bottom of that list lean heavy to Harris..
Using polling error from one election cycle isn't a great idea. 3-5 cycles, yes.
you'll never guess who was right in the previous cycles too? The top guy has been right both in the potus election and the 22 midterms.
Also, Fox has been repeatedly wrong cycle after cycle
Then why post a chart from only one cycle? Wouldn't it be more effective to post a chart of data from a wider date range?
Even when some data overlaps, small snapshots of data always look manipulative.
do you have a chart that shows data across multiple? Sure, I could go hunt for something like that but i dont have it available. That is the most recent potus election data.
Also, look at how off the high level polls in the RCP and 538 averages were..
We have a ton of proprietary data. What is always clear is that small sample sizes are inherently the most misleading and least to be trusted.
538, 270, RCP…….the benefits and problems of averaging results lives in all of them. Merely to differing degrees.
Polling to predict results is problematic. Polling to find trends is far more reliable.
The best analysts are taking trends and applying them to the adapt of historical election results and then applying data of population movement, population age data/trends, etc. to predict outcomes.
There are some reliable takeaways: polling using college students is garbage, the greater the voter orientation correction required = the higher the likelihood of error compounding, polling collection processes by 75-80% of pollsters is absolute garbage, polling true up procedures by 90% of pollsters is "at least" poor, polling averaging is nonsense, the gold nuggets [if any] exist primarily in the tabs.
For example: suburban Hispanic married women in Arizona b/t 25-50 in age, AA men b/t 35-50 in Philadelphia, union voters by profession/union and geography, and similar.
