Our best bet: Wait a couple more weeks to see where the dust settles. Our now-cast, for example, which projects what would happen in a hypothetical election held today, has Clinton winning the popular vote by 8 percentage points. My guess is that that will tick down a couple of points in the coming days. Our polls-plus model, which accounts for convention bounces and so discounts some of Clinton’s recent surge, projects her to win on Nov. 8 by 4 points. And our polls-only model, which basically takes the polls at face-value, projects her to win by 7.

But it’s also possible that Clinton’s strong numbers aren’t solely the result of a fleeting post-convention afterglow. As my colleague Nate Silver pointed out on Friday, Trump’s recent struggles — his attacks on the Khan family and feuds with Republican leadership, for instance — could be inflicting more durable damage to his chances. Trump is the least-liked major party nominee in modern history. Perhaps the conventions and their aftermath, when many voters presumably tuned into the 2016 race for the first time, established a new equilibrium. Perhaps this is 1988 all over again, with the parties reversed.