"This isn't scientific because we didn't get the results we wanted"
No poll is scientific, not Twitter polls, not polls you put on your website or polls done by Rasmussen. They are all flawed and all use samples instead of the whole population. They are all using skewed sample sizes. the big-time pollsters overload Democrats in polls ALWAYS. Now, as Rush has been saying, they are adding Republicans - BUT - Republicans in suburbs who are mostly GOPe Romney-ites who hate Trump. Soccer moms, Karens, etc who think the GOP should go back to the Bush-era and just be lapdogs and controlled opposition to the Dems but let the Dems get what they want.
any kind of poll relies on a "sample" which can be interpreted to mean anything the pollster wants it to be. If they did a census of a particular area, and everyone who responded was 100% honest about their political leanings, and they sampled based on that breakdown, then it'd be scientific. Making estimates leaves it wide open to have "D+10" or whatever and skew the polls. You're not going to get many that are 100% honest about their politics and even then if you're not polling EVERYONE, it's not accurate.
"This isn't scientific because we didn't get the results we wanted"
No poll is scientific, not Twitter polls, not polls you put on your website or polls done by Rasmussen. They are all flawed and all use samples instead of the whole population. They are all using skewed sample sizes. the big-time pollsters overload Democrats in polls ALWAYS. Now, as Rush has been saying, they are adding Republicans - BUT - Republicans in suburbs who are mostly GOPe Romney-ites who hate Trump. Soccer moms, Karens, etc who think the GOP should go back to the Bush-era and just be lapdogs and controlled opposition to the Dems but let the Dems get what they want.
any kind of poll relies on a "sample" which can be interpreted to mean anything the pollster wants it to be. If they did a census of a particular area, and everyone who responded was 100% honest about their political leanings, and they sampled based on that breakdown, then it'd be scientific. Making estimates leaves it wide open to have "D+10" or whatever and skew the polls. You're not going to get many that are 100% honest about their politics and even then if you're not polling EVERYONE, it's not accurate.
It's only scientific when the results are switched.