This piece misleads. It begins with McConnell’s outdated 18% approval rating, then acknowledges he “clawed” it up to 39%, but omits that McConnell’s approval was much higher than McGrath’s, and omits the poll that had McConnell approval at 44%. 1/ rawstory.com/2020/12/why-th…
2/ “Likely voters have a mixed opinion of Senator McConnell, with 44 % saying they have a favorable opinion of him, while 46 % say they have an unfavorable opinion of him. Amy McGrath receives a negative 34 - 47 percent favorability rating. “ - Oct. 2020 poll.qu.edu/kentucky/relea…
3/ It also omits that the exit polls predicted a strong McConnell victory. Yes, McConnell won by more than the polls predicted, but this information should not have been omitted from the piece since the piece begins w/ a discussion of the polls.
4/ Exit polls. tdmsresearch.com/2020/12/14/202…
5/ The other numbers in the piece may be worth looking at, but I haven’t double checked them. I’d like all races to be audited. It’s why I pushed for the #SAFEAct which wld have required that, but Rs killed it. axios.com/gop-senator-el…
6/ If someone is going to use polls to suggest that an election outcome was surprising, it isn’t ok to cherry pick them.
7/ Maybe just don’t use them at all? Or show why we shouldn’t rely on them? But no cherry picking.
8/ Nate Silver gave McConnell a 96% chance of winning. If someone is going to use polls to suggest that cheating occurred, it’s not ok to just use the ones they like w/o explaining why they are discounting the others. projects.fivethirtyeight.com/2020-election-… 
9/ Better screenshot.
10/ I’m not a huge proponent of polls at all. But if someone is going to use them to suggest that an election outcome was either legitimate or fraudulent, it’s not ok to use only those polls that support their argument and omit those that don’t.
11/ Note: I have not fact checked the rest of the piece. And I’d still love to see the race audited and ES&S investigated.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
