Thread by @TSBigMoney

Post debate nat'l polling through the weekend after the debate, excluding regular weeklies

6/26: 7 polls, 6155 respondents

7/31: 6 polls, 5687 respondents

9/15: 7 polls, 7265 respondents, including a massive Ipsos not released

Most recent: 2 polls, 1774 respondents

🤔🤔🤔

The Ipsos poll that was sat on (and was basically accidentally leaked through a tweet of an Excel chart) showed Bernie improving over last debate and Warren losing ground, right as a media narrative about Warren running away with the nom began. And now we have crickets

On Friday, Sanders was about to pass Warren in Scott Rasmussen's Harris X poll, even after filtering their release to "Definite Voters" only, which heavily benefitted Warren and Biden. Then the next day they said they would no longer poll the race bcz nothing is happening

I try to avoid being conspiratorial, Berners are pretty sensitive to the seemingly bottomless media bias against the candidate. But how do you explain Ipsos spiking a huge survey, just to rush out a notably smaller release a few days later instead showing a very different result?

Assuming Ipsos's normal A/RV split, you can figure that was about 1369 Dem+leaner RVs. The vote breakdown would have been Biden 20%, Sanders 16%, Warren 13% if you combined the Sept 16-20 and 23-24 results. A week later, normal sized poll, Sanders is back in 2nd. Come on

Now 3 polls, 2204.

Biden +2, Sanders +3, Warren -2 since end of Sept.

Bernie won the debate, so, media blackout

emersonpolling.reportablenews.com/pr/october-nat…