That is why you should probably follow the link, since Silver tends not to cherry pick polls to work backwards to a predetermined end, and he uses actual data sources rather than some foggy 30+ year-old memory.
Also, if you were there you should recall the third candidate that was polling near 10% John Anderson. Well maybe you don't, he often gets overlooked. A quite centerist GOP candidate whose polling support ended up breaking heavily to Reagan rather than Carter when the votes were cast. There is nothing like that in this election.
P.S. To spoil the surprise, because lets face it you don't have the courage to follow the link and read it, the only election where polling this far out from the election didn't nail the popular vote winner was the infamous Dewey vs Truman. A good deal of that anomaly was created by a shift in communication/demographics combo that polling firms didn't realize and that heavily favoured Dewey in the polling. There might be something like that this year, the switch by households away from landline telephones to cell phones only. But that favours Romney in the polling, not Obama.
EDIT: Anyway, enough poking the bear. I'll put it in my calendar to drop by here on Nov 8th to see how you are doing.