Jay Cobb explains at National Review Online how an overly crowded Republican primary field helped pave the way for Donald Trump‘s nomination.

There are many theories for Trump’s rise, but the simplest is the truest. Celebrity populism combined with media support helped him win a divided field. He started from a base of around 20 percent, a number similar to past fringe candidates. He was then lifted to front-runner status with his media advantage. Unlike previous front-runners, he never came under attack in a way that would have exposed his vulnerabilities, as is happening right now.

The failure was in not taking him down before the primaries started and in failing to unite behind an alternative front-runner. Had the party united to defeat Trump, he could have been defeated. But each candidate focused on his own interests and hoped to be the last person standing. The assumption, which turned out to be wrong, was that Trump could not defeat a candidate face to face and would be doomed once the field narrowed.

Writing in the Investor’s Business Daily, Stephen Moore compared this to the “prisoner’s dilemma,” a concept drawn from game theory. The prisoner’s dilemma is usually described as follows. Two prisoners are facing medium sentences, let’s say five years each. Each can turn evidence for additional crimes that will give the other a longer sentence, say, ten years. The prisoner who turns evidence will be rewarded with a three-year reduction in his sentence.

What each prisoner realizes is that regardless of what the other does, it’s worth it to snitch. If the other keeps quiet, snitching will reduce your sentence from five years to two. And if the other does snitch, you can reduce your sentence from ten years to seven. Since each prisoner acts in his own self-interest, they will both turn evidence, and both will end up with seven-year sentence; whereas if both had kept silent, both would be better off.

However, if the two have a way to communicate or if they have a code of conduct that they trust, they can break the pattern. When the Cruz and Kasich campaigns coordinated late in the primary process, Marketplace described it as an example of repeat playing, in which the players start to understand that it’s in their best interest to cooperate.

Ideally, members of a political party should have an overriding shared interest in ensuring that their party wins. Barring that, they should at least take their own interests to heart. After all, they are not playing blind; there are channels of communications that allow the players to make deals. So why couldn’t they communicate earlier on for a better result?

The answer is that the prisoner’s dilemma is not really an accurate comparison. In the prisoner’s dilemma, working together benefits not only the whole but also each individual. In 2016, the Republican party as a whole lost, but it’s a more complicated situation for the individual candidates. Many of the candidates did in fact choose the winning strategy for themselves.