538 founder Nate Silver said the major "X-factor" in predicting the 2018 midterm elections is turnout and that is determined by the "Trumpian news cycle." He also said we might see "an October surprise or two" before November 6, in an interview on ABC's 'This Week.' <blockquote>GEORGE STEPHANOPOULOS: OK, let’s look at the House as well... You see a seven in nine chance – six in seven chance the Democrats would control, one in seven chance Republicans keep control. So there it comes down to 85 percent chance that Democrats win control. That sounds a lot bigger than it is, right? NATE SILVER, 538: Yes, look, I mean if you were running a business and I told you there’s a 15 percent chance or a 20 percent chance that you key supplier won’t make it’s delivery, you would treat that as a very tangible, real-world risk and you would do things to hedge against it. The thing about the House is that you cannot circle 23 districts where you say oh I know for sure Democrats will win these. Maybe 10, 12, 15 look very likely. However you have a field of maybe 80, 90, 100 potential pickups, mathematically probably the dice come up good enough for Democrats in those districts. But like, there are not a lot of guarantees and the House is very much fought at a district by district level. STEPHANOPOULOS: And only a handful of seats where the Democrats could lose the seat. SILVER: And that’s why – I mean it’s – it really is the mere image of the Senate where Democrats have so much exposure in the Senate. All these incumbents, you know, all these very red states, just a reverse of that in the House where Democrats are kind of in a no lose situation almost literally in the House where they might have four or five seats they could lose versus 100 GOP seats in play. Not a lot of guarantees, but that’s why we show like a very wide range, anywhere from a 20 seat gain, the Democrats have a disappointing night which is not quite enough, up to 50, 60 seats if the turnout is – is very high. STEPHANOPOULOS: What are the biggest X-factors that increase the uncertainty in your model? SILVER: It is turnout, I mean it’s – turnout is always difficult for pollsters to forecast and the fact that you have a lot of districts that have not had competitive races in a long time, turnout’s even more difficult to forecast there than in a state like Florida for example. And we have two weeks to go, you know. I would not put it past us to have – for us to have another October surprise or two in the era of Trumpian news cycles. </blockquote>