I side with Wayne Woodlief over Howie Carr on the question of whether a independent gubernatorial campaign by Tim Cahill helps or hurts incumbent Deval Patrick in next year’s general election. If Cahill finishes third but gets a substantial number of votes, Patrick is probably toast.

It seems logical that multiple candidates can split up the anti-incumbent vote and allow an unpopular pol to stay in office, but that hasn’t been the case at the national level. The strongest third-party presidential candidates of the past 50 years were all involved in elections that threw out the incumbent party. In 1968, George Wallace got 14 percent and the Democrats lost; in 1980, John Anderson got 7 percent and Jimmy Carter was unseated; and in 1992, Ross Perot polled 18 percent and George H.W. Bush was evicted. We haven’t had many strong third-party candidates for governor in Massachusetts, but the last one (Christy Mihos in 2006) obviously hurt the GOP’s attempt to hold onto the office.

My hunch is that third candidates run strongest among voters who lean toward the party in power: They’re disappointed or underwhelmed with their party’s candidate, but they’re not willing to cross party lines. They don’t necessarily like “throwing their vote away,” but they don’t agonize about it. As it happens, they have a general dislike for one major party but also dislike the individual in charge of the other party.

In contrast, the people who really dislike both the individual and the party in power are not going to waste their ballots. They’re going to rally around the opponent who can win. So, in 1992, Ross Perot did best among Republican-leaning voters who didn’t like Bush, while people who hated both Bush and the Republican Party stuck with Democratic nominee Bill Clinton. In 1980, John Anderson snared a lot of Democrats who were cool toward Carter, but voters who didn’t like Carter or the Democrats rallied around Reagan.

Based on history, I see almost no chance of a competitive three-way race for governor next year that Deval Patrick could win with as little as 35 percent of the vote. My prediction is that Cahill has until the Republican primary next September to eclipse Charlie Baker and Christy Mihos to become the alternative to Patrick. If he fails to do so, almost all the voters who really dislike Patrick are going to unite behind the GOP nominee (part of the bounce from his primary victory). In that scenario, Cahill will then either fade completely (good news for Patrick) or end up getting anywhere from 6 percent to 18 percent of the vote — which will come out of Patrick’s hide.