As far as gubernatorial races go, Wednesday was a wild ride.

Independent candidate Eliot Cutler vowed to keep campaigning but conceded he’s a longshot in the three-way race. He’d understand if voters who had been sticking with him shifted allegiances at this late stage, he said.

As groups allied with Democrat Mike Michaud’s campaign spotlighted Cutler-turned-Michaud supporters, Cutler lost one of his highest-profile backers. Independent Sen. Angus King effectively followed Cutler’s advice Wednesday and announced his support for Michaud.

Wednesday’s wild ride might appear a bit like chaos. But turning to political science, there are a few principles that capture it all:

The voting calculus is rendering its final calculations. It was clear from the start of this race that Maine voters’ decisions in November would be about more than simply their preference for Michaud, Cutler or Republican Gov. Paul LePage. In a three-way race decided by a plurality of the votes, an entire “calculus of voting” applies, as political scientists have written through the years. The strategic voting that ensues as a result is an effort on the voter’s part to prevent the victory of her least favorite candidate, even if it means her chosen candidate doesn’t get her vote.

Polls are where the parties were at. Political party labels typically serve the purpose of coordinating voters — at least in a conventional, two-way election. The “D” or the “R” next to the candidate’s name normally sends a clear signal. But not when there’s a credible third-party challenger. In this situation — and especially after living the results from Maine’s 2010 gubernatorial election — voters who don’t want to see LePage re-elected instead use the polls as their coordinating mechanism. “Opinion polls can … coordinate voter choices, but only if they convey a clear message,” Bowdoin College political science professor Jeffrey Selinger wrote in a BDN OpEd in August. Recent polls that are consistently showing little movement for Cutler — or even downward movement — appear to be coordinating the choices of anti-LePage voters.

Polls aren’t just the party. Polls are king. The mechanisms typically responsible for coordinating voter preferences — campaigns and, to a lesser extent, high-profile endorsements — are looking to the polls for guidance. Michaud’s campaign has looked to the polls throughout the election season to make the argument that Michaud is the guy who can beat LePage. Cutler’s campaign tried to do the same for its guy. And on Wednesday, King, the senator, allowed the polls to be king when he decided to shift his weight from Cutler to Michaud. He made an argument for voter “utility,” another political science principle that explains strategic voting. “[W]hat informs our decision to vote for the governor of Maine should be the belief in a system of government where every vote counts,” King wrote in the statement announcing his change of heart.

There can only be two. The strategy Maine voters appear to be employing in the final days before Election Day ensures the Duvergerian Equilibrium applies again. In 1951, French sociologist Maurice Duverger wrote that an election system in which victory requires a plurality effectively limits the political system to two dominant parties. The Duvergerian Equilibrium, as this concept was later called, says, essentially, that there can only be two truly viable candidates — as in, LePage and Michaud — even in a race with three or more.

If not my guy, please don’t let it be that guy. Do voters care more about allegiance to their first choice, or do they care more about keeping their last choice from prevailing? If it can only be a race that’s effectively between two, voters have a choice to make about how they deploy their utility.

While many voters have expressed support for Cutler they’ve remained wary of actually voting for him for fear of re-electing LePage. That dynamic didn’t change throughout the campaign. The result is the two-person race we have today.