>So if anybody could, why the fuck are we giving all
>the credit to one guy whose model none of these laymen
>can explain?

because he did it and others didn't. he did it first.

>Trust me, this is the type of shit I get PAID FOR

umm..good for you?

>But I'll be damned if I conflate my fake expertise with
>actual expertise

what? like what exactly are you accusing silver of?

>Bwahahaha THERE IS ONLY ONE data point that matters:
>
>THE END.

ok. sure. according to you. (oh, and he was better at the end).

>We, BY DEFINITION, DON'T KNOW HOW THE RACE would have gone
>at any point, because the reliability of the polls changes
>DURING the polling season.

yes, that's why he gave probabilities and notes that forecasting can change as we don't know the future. but in another post, you hate on him using probabities, so...lol.

>RCP did well at the end.
>
>Silver did well at the end.
>
>BOTH argued that one poll cannot be the focus.

but SIlver did better and made the case better.

>And again: THE REASONS that Silver's poll was "better" cannot
>be explained by anyone who isn't a mathematician or
>statistics
>professor.

so? so every statistical model now should be something laypeople can explain in detail? wtf?

>"In the meantime, even Silver may need to tip his cap to
>someone who seems to have done an even better job at
>prognosticating the final presidential election results. In a
>blog post today, dailyKos founder Markos Moulitsas noted that
>he had predicted exactly the final Electoral College vote
>totals and reported an average margin in the swing states that
>was less than Silver's or that of any aggregator. Moulitsas'
>methodology? A savvy but seemingly manual reading of
>last-minute pre-election polls. Maybe the quants don't have
>all the answers after all."

HELLO ANECDOTAL EVIDENCE! YOU JUST DID IT TO ME! BWAHAHAHAHA.

>WHOA!!! So study the last-minute pre-election polls, and get
>data more accurate than Silver or RCP!?!?!

one guy did. ONE GUY. YOU JUST DID IT. YOU JUST USED AN ANECDOTE AGAINST STATISTICS. LOL/

>WOORD!?!?!?

how many got it really wrong doing the same thing?

>RIGHT!!! And RCP STILL GOT THE RESULTS BASICALLY RIGHT!!!

yes. basically right but not as right.

>I'm talking about all these people that he's "educated"
>
>None of them can tell me what is superior about his model

that's not what they're learning, nor is it why they read his blog, nor is it what his blog is about. sorry that disappoints you that he isn't using 538 to teach people bout regression methodology.


-----------
It's only funny till someone gets mad. Then it's hilarious.