Cook: Pickup of 20-45 seats, most likely 30-40. Daily Kos Elections: Pickup of ~ 30 seats. 538: Dems are 'pretty clear favorites.' Odds of control: 80.0%/87.4%/84.1%, depending on model. Although maybe we were starting to see a late break to the left? And a lot of polled races have been within a few points, making it easy to imagine a very long night for Dems, or conversely a total blowout.
By the same token, that same kind of error in Democrats' favor could mean a gain of 50 or more.
The GOP still has a 'systematic polling error' chance to hang on to control, though - and we all remember the last time one of those errors happened. What do the prognosticators say? Unlike in some past years, they're broadly in agreement - Democrats are likely to get maybe about 35 seats. Contesting far more seats than usual, holding an substantial 8.5 point advantage on the generic ballot average, and with eye-popping amounts of small dollar donations flowing in, the Democrats may have what it takes to finally wash over the gerrymandering wall the GOP erected in 2010. Buoyed by historically low presidential approval ratings, successful 2017 state elections in Virginia and New Jersey, and a string of surprise victories in special elections, the Democrats are feeling good (well, as good as Dems ever feel) about getting at least the net 23 pick-ups they need to take control of the House.
