First of all RCP isn’t really to be trusted because they do use the polls that are oversampled for Ds (which is a choice they make on purpose) but also because they’re weeks out of date.
I appreciate the averaging idea, but they’re a large percentage of info no longer true.
Agreed.
RCP is also about making money from its website; therefore, its averaging process causes their indicators to react more slowly to newer polls. Averaging is dilutive (except when you always include an extreme or out”liar” poll in one candidate’s favor.
RCP is great, IF you want to measure respondent sentiment from 3 or 4 weeks ago. Sometimes they include polls that are 6-8 weeks dated.