A very high proportion of seats in any given election are, for "natural" or gerrymandering reasons, regarded as "safe seats" for their incumbent or (if the incumbent isn't running) party. These can all be ignored, unless you're just wanting to mess with a few polling places in safe Democratic districts to further a narrative of election chaos. For actually directly changing outcomes, they're irrelevant.
A seat that is likely to flip is probably going to be a relatively close race.
There are a bunch of public and private sources you can use, and databases both parties have already compiled, to find out which polling places are likely to be overwhelmingly visited by one party's voters over the other.
A majority of one in the House, and a tie in the Senate, is all the White House needs to mostly prevent Congress from messing with them much (though larger margins are better).
Combine all these facts and they have more than enough ICE agents to have a huge effect on the outcomes. They only need to show up in a handful of specific places to completely change the course of the next couple years.
A seat that is likely to flip is probably going to be a relatively close race.
There are a bunch of public and private sources you can use, and databases both parties have already compiled, to find out which polling places are likely to be overwhelmingly visited by one party's voters over the other.
A majority of one in the House, and a tie in the Senate, is all the White House needs to mostly prevent Congress from messing with them much (though larger margins are better).
Combine all these facts and they have more than enough ICE agents to have a huge effect on the outcomes. They only need to show up in a handful of specific places to completely change the course of the next couple years.