If you could change the result of one senate race in favor of both parties in each election cycle

[Kabam]

Not counting special elections and pretending everything happens in a vacuum.

2008: Kentucky (They will flip all in 2014 anyway, so bye Mitch!) and North Carolina (Will flip in 2014 anyway. Dems are in a better position in 2020 with an open seat or an 80+ Dole. And the butterfly effect might prevent Hagan's death.)
2010: Wisconsin (Bye, Ron Johnson. Better chances to win in 2016 and 2022.) and California (Boxer retires in 2016 anyway and automatic flip in 2016.)
2012: Arizona (Carmona is probably better than both Sinema and Berkely) and Florida (Flips in 2018 anyway. Maybe Dems can find a better candidate there.)
2014: Alaska (Given Peltola's recent success, Begich has good chances to hold on in 2020. And it prevents Sullivan, who is hard to defeat.) and Illinois (Easy flip in 2020 and someone younger there might be nice.)
2016: Wisconsin (Bye, Ron Johnson [2] and really the only good option for Dems) and Colorado (By far easiest to win back in 2022.)
2018: Texas (Best chances to keep in 2024 and Beto would be a good senator) and West Virginia (Gone in 2024 anyway, with Manchin retiring. Menendez in NJ (the other good option) is a more solid vote for Dems.)
2020: Texas (Probably not much harder to keep in 2026 than North Carolina. Hegar > Cunningham and Tillis might retire anyway, otherwise he is easier to beat than Cornyn. Also thinking Collins will retire anyway.) and Colorado (Should be easier to flip in 2026 than others and Hickenlooper does not really seem to want to be in the Senate anyway.)
2022: North Carolina (Very hard between WI and NC. But Beasley might have better chances to hold on in 2028. And I prefer her over Barnes.) and Connecticut (Blumenthal is getting old and defeating Levy would be so easy in 2028.)