For political junkies like me, it's easy to go a little (or a lot) crazy during the excruciatingly long presidential campaign season -- becoming joyful when a chosen candidate (Clinton, in my case) is ascendent in the polls, and despondent when a downturn happens.
Here's my advice about how to handle this:
Follow the best poll analyzer around, FiveThirtyEight, Nate Silver's operation. After he left the New York Times, Silver and Co. branched out into analyzing sports and other areas.
But politics, not surprisingly, is his emphasis at the moment. I follow FiveThirtyEight on Twitter and enjoy his highly competent take on what's happening in the Clinton-Trump battle.
He regularly reminds people that single polls aren't meaningful, like the recent CNN poll that showed Trump with a 2-point lead nationally. Looking at averages of many polls over time is the way to go.
Currently FiveThirtyEight gives Clinton a 67% chance of winning, almost exactly the same as in early June. There have been ups and downs for the candidates, but no overall movement in the past three months.
I'm also a fan of prediction markets, where people lay bets on political and other sorts of contests. I don't follow prediction markets very closely, so had to do some Googling to find a good example to share.
PredictWise popped up. This is the web site's pitch:
PredictWise reflects David Rothschild’s academic, peer-reviewed, research into prediction markets, along with polling and online/social media data. The backbone of predictions on this site are market-based, generated from real-money markets that trade contracts on upcoming events. While we have great respect for: polls and online/social media data, the topline predictions reflect data that are more accurate, flexible, and update in real-time.
Below is the current presidential chart.
You can see that the prediction markets are more stable than the FiveThirtyEight model. FiveThirtyEight had the race tied in late July, while at that time PredictWise had it 70-30 in favor of Clinton.
At any rate, there's two months left before the November election.
Quite a few ups and downs likely remain for both presidential candidates. Like I said, it's easy to get overwrought about temporary blips in the long term trend. The time to worry is when your favored candidate is on a steady downward trend close to election day.
Once the presidential debates begin, I suspect we'll see some significant bumps for the candidates. The debates will give people a chance to see Clinton and Trump on the same stage, talking about the same issues. I'm predicting Clinton will benefit from the debates.
But I'm not betting on it. I'll be watching PredictWise as a detached observer.