Not fully. That's why he wasn't prepared for the Republican avalanche.
Silver is a compiler who takes polls at face value and has a very sophisticated method of blending those poll numbers into an accurate percentagewise state of the race, if the polls are accurate.
Frankly, Nate Silver has taken money out of my pocket. Prior to his arrival, political betting odds were crude and wonderfully vulnerable. Favorites were simply too low if they had a consistent small edge in the polls. I could bet those candidates at 2/3 (-150) odds when they should have been closer to 1/8 (-800). Silver ruined that.
However, 2014 was a rare cycle in which Silver obviously wasn't prepared for the twin mingling of a Democratic president with a sustained low approval rating in the low 40s, combined with midterm realities in which too many of our key demographics don't show up. As a handicapper for 30 years I was well aware what would happen. The polls wouldn't depict how devastating it would be. I posted here Monday night that I had absolutely no energy because it was going to be a disaster since the party with the wind at its back would benefit from a late surge and pick off the vast majority of the toss up races.
We wouldn't have been in nearly as much jeopardy if Obama had maintained approval rating just a bit higher, like his 45 or 46 throughout summer and fall 2010. Once he's in the low 40s month after month it's identical to Clinton in fall 1994. That was the first midterm I bet politics heavily. The poll consensus meant nothing. The betting odds actually shifted somewhat to Democrats in the final days in 1994 because Clinton's approval rating ticked upward. I sensed it was meaningless and it was. Elections are dictated by foundational opinions, not late flimsy tide. That tide can carry a primary but not flip a general election if all the indicators point the other way for months and years.
Bush in 2006 had a disastrous approval rating, lower than Clinton in 1994 or Obama this year. Republicans aren't in as much peril in midterms because their older demographics participate dependably. We've become a party that specializes in presidential election years.
Polling can be rubbish. I've emphasized here since fall 2002 that Georgia and Alaska polling is incompetent, invariably overstating the support of the Democrat. This year we had high hopes in Georgia and Alaska. What a joke. The more things change, the more they stay the same.
Nationwide the polling tends to be very accurate other than atypical cycles like this one in which the most important variables were tilted in one direction only. Again, I think my betting background helped prepare me for it. Once actual money is at stake you're forced to pay attention. Nate Silver follows betting but he's not dependent on it for his bottom line, as I am. Therefore he just accepts the polls and happily inserts them in his program, minus proper scrutiny.