Quote (hydrogod @ Sep 13 2016 09:37am)
I've heard you work there, so if you stop trading paychecks for pies you'd be good😁
Isn't huff post extremely liberal? I always thought of rcp as being less biased
I don't keep track of Huffington Post editorials or reporting, the aggregate is completely separate from it and is the only public feature that I'd suggest people use. It gives the user full domain over inputs and is incredibly transparent about its methodology (especially with which firms it includes and why) -- and crucially, you have the option of deselecting firms if you want, in order to create a customizable aggregate of your own.
The RCP aggregate by comparison isn't transparent, and isn't unbiased. It deliberately chooses arbitrary survey date cutoffs in order to skew the numbers towards Republicans, if a polling firm publishes multiple data sets (like numbers among All Adults, Registered Voters, and Likely Voters) it will deliberately pick the one most favorable to Republicans, and it deliberately includes Republican polling firms without doing the same for the Democratic equivalent. It also just includes fraudulent firms that should be excluded. Now that the HuffPo product is available, the RCP aggregate really is an incredibly-shoddy product.
Quote (AiNedeSpelCzech @ Sep 13 2016 09:35am)
I'll stick with Math Jesus in any case. I mean, if you want to talk about a founder with a dog in the race, I can't imagine HuffPo being all that much better. :P
If you're referring to Nate Silver then that's really a separate issue. People are free to trust his commentary, his modeling. I would even recommend him with some caveats. But what he provides is a model, and commentary based off of that model; the HuffPo Pollster and RCP averages are simple aggregates. They're directly comparable, and HuffPo simply has the far superior product. People have gotten use to RCP, I suspect it will eventually become outdated.