There are multiple factors that make it unlikely for a third party to win, and "first past the post" is a lesser contributor.
Of far more concern are ballot access laws that make even the formation of a third party a colossal waste of time, and partisan redistricting that even makes it impossible for the second party to win. I won't even delve into ridiculously insecure electronic voting tabulators, because I'm too likely to devolve into paranoid conspiracy nonsense from there.
Eliding over all that, the spoiler effect is what forces the two major parties to make changes to their platform in your locale. If a party believes that your vote for another party caused them to lose an election, they will certainly make an effort to kiss your ass at least once in the next cycle, becoming more like the party you voted for. It is easy to argue that voting for either of the two major parties means that you don't want them to change their platform or policies.
Voting for the party in power means you are happy with the status quo. If you are not happy, and you find the second party to be more objectionable, voting for a third party that puts you between them and the party in power on the political spectrum is the correct move. If that party is scared enough by votes bleeding from them to that party, they will shift in that direction, perhaps enough that you will be happy voting for them again.
You're getting into game theory territory. At present, it might be the case that voting for the third party helps to change the platforms of the dominant parties, but if this idea becomes well enough known to significantly impact voting behavior, then the dominant parties will likely pick up on it and adjust accordingly. That is to say that if they see (through polling, exit interviews and statistical analysis) that this is a significant effect, it also likely implies that the people who employ this strategy are significantly less gettable by platform changes, because a party that can taylor itself to the issues they feel strongest about will always be more attractive than the two parties that have to satisfy everyone. So you might see a Nash equilibrium develop that nullifies the effect of this strategy.
The current equilibrium is stuck at candidates pretending to be further from the political barycenter to win their party primary, then pretending to be closer to the barycenter in the general election. Any person holding a consistent position over the entire elections cycle--which is to say anyone that has non-negotiable personal principles--is eliminated from consideration. Campaign promises are rarely honored. Nothing of genuine importance--like debt, systemic unemployment, inflation, or even just passing a budget bill for the year--is ever seriously considered.
If a new non-optimal Nash equilibrium is reached, that's fine with me, because the one we have now is absolutely terrible, in my opinion. perhaps it will inspire new political strategies that do not leave huge segments of the populace effectively disenfranchised.
Of far more concern are ballot access laws that make even the formation of a third party a colossal waste of time, and partisan redistricting that even makes it impossible for the second party to win. I won't even delve into ridiculously insecure electronic voting tabulators, because I'm too likely to devolve into paranoid conspiracy nonsense from there.
Eliding over all that, the spoiler effect is what forces the two major parties to make changes to their platform in your locale. If a party believes that your vote for another party caused them to lose an election, they will certainly make an effort to kiss your ass at least once in the next cycle, becoming more like the party you voted for. It is easy to argue that voting for either of the two major parties means that you don't want them to change their platform or policies.
Voting for the party in power means you are happy with the status quo. If you are not happy, and you find the second party to be more objectionable, voting for a third party that puts you between them and the party in power on the political spectrum is the correct move. If that party is scared enough by votes bleeding from them to that party, they will shift in that direction, perhaps enough that you will be happy voting for them again.