Looking at history, one person choosing to vote is statistically useless if the goal is to influence the outcome of a presidential or congressional/senatorial election in the USA.

The chance of one vote being the difference between a tie and a win for a candidate has been approximately zero. And in the extremely rare times when there has been such a one vote difference, there have been re-counts that conclude with a margin that is larger than one vote [0][1].

Mathematically/statistically this makes intuitive sense. One might ask: what's the chance that a vote ends in a tie when there are 10, 100, 100,000... 10 million voters? As the voting pool increases, the chance decreases.

The most common objection is along the lines of, "if everyone decided not to vote because their vote didn't matter, then no one would vote!" Yep, that's correct... but that doesn't describe reality: clearly people vote a lot.

Voting because one enjoys it, gets satisfaction of some sort from it, feels that they are doing their "civic duty" and discharging a responsibility, etc... that's fine. But don't vote because you think you're gonna change the outcome.

[0]https://www.nber.org/system/files/working_papers/w8590/w8590... "THE EMPIRICAL FREQUENCY OF A PIVOTAL VOTE"

[1]https://archive.ph/ThMlo NYT: "Why Vote?"