Good Nate Silver piece on pollsters potentially "herding" this cycle.

In other words, releasing results that are similar to others.

Small quibble though- polling violates the central limit theorem, so in reality these odds aren't as bad as they are here. But yeah, there looks to be some herding going on.

If I want to be optimistic about Kamala Harris, one explanation is that pollsters don't want to miss on Trump again and so aren't releasing their polls that look good for Harris.