Read on Twitter
Holy crap, I& #39;m floored by how bad this article is. These political scientists should probably have talked to their political science colleagues who work on network decision desks before writing this garbage. Do they think the networks are flying blindly into this election? https://twitter.com/guardian/status/1313795667937103872">https://twitter.com/guardian/...
The former @APSAtweets presidents who signed on to it should be ashamed to have attached their name to an article that& #39;s at the worst intersection of incredibly important and poorly researched. No, not poorly researched. Completely unresearched.
Their proposal of how to do things right on election night is *literally* exactly how it does work on election night. Every major news network has an independent decision desk team, made up of political scientists, data scientists, pollsters, and election scientists.
They take a swipe at exit polling. But no media organization calls close races off of exit polls. They& #39;re attacking a straw man, because they didn& #39;t talk to anybody who actually works on an exit poll team or decision desk.
And, again, many of the people who work on exit poll teams or decision desks are *literally their polisci colleagues whose offices are down the hall from them*
This isn& #39;t to say that we& #39;ll definitely know who won by the end of election night. If it comes down to a state or two we might not. But this article implies that news orgs are stupid enough to put their reputation at risk by not putting the right people in charge of calling races
You know how I know this article is bad? They literally never use the phrase "decision desk" anywhere in it. Again, what they propose is already how it already works.
For a *much* better take on how things will go on election night, you should read @cstewartiii & #39;s recent op-ed. He did his research and actually talked to folks who work on decision desks before publishing it. https://www.wsj.com/articles/actually-we-will-know-a-lot-on-election-night-11600959867">https://www.wsj.com/articles/...
/end rant
