Actually 31 is correct. 21 retiring outright. 10 running for higher office. 7 Democrats are retiring outright and 8 are running for higher office. https://www.theatlantic.com/politic...2018-congressional-retirement-tracker/545723/ I don't know what those other 11 refer to, but there are quite a few districts with potential gains for Democrats in those 31. Issa, Ros-Lehtinen, Reichert, and Royce are all from districts Clinton won. Dent, LoBiondo, Poe and Trott all came from districts in which Trump won by less than 10. Current race standings have about 3 Republican seats being Democratic leaning. In addition, they have 16-17 tossups, and about 21 leans (where if something weird happens, which it inevitably does with enough rolls of the die, you have potential pickups). Democrats have 4 tossups and 5 leans. If Democrats split the tossups and capture their leaning Republican held seats, that would be 9 pickups. And then you have the huge disparity in leans. I think the high end of your scale is likely the low end of the potential output. And that is assuming about a +7-8% Democratic electorate. If it is higher than that (and it has been in a number of the specials), the math shifts pretty rapidly. I doubt that they will gain "at least" 3-4 seats this year. Current Senate ratings: https://www.cookpolitical.com/ratings/senate-race-ratings It is a bad field for Democrats, but they have 4 tossups to 3 Republican tossups right now. They do have 4 additional leans. But that would still essentially mean that Republicans would have to hold their tossups in total and split the Democratic tossups while picking up at least one lean. That seems a bit optimistic from your view. And 2020 is an awful field for the Republicans. They will likely gain back Alabama, but they have 22 seats up in 2020 to 11 for Democrats. I don't see how you find that as being "a few more" from a net perspective. Retirements are much higher in Republican seats than Democratic ones. That is actually true.