Entering the Republican Iowa caucuses on Feb. 1, Donald Trump looked unstoppable. Topping every poll in Iowa by a margin of seven points, his victory seemed well assured. Only six hours later, his domination in the polls evaporated, ending in a humiliating three-point defeat to Ted Cruz. This 10-point difference is enough to make Trump’s defeat the largest polling error since Barack Obama’s 2008 loss in New Hampshire. Was this defeat a case of short-term polling abnormalities or does it betray a structural instability in the Trump campaign?

Miscalculation of polls, especially in early-voting states, is not an uncommon occurrence. This phenomenon can be attributed to low voter turnout rates, which often fall below 20 percent. Nevertheless, in the aftermath of the Trump campaign’s unraveling, political number crunchers immediately went into election assessment mode. Among the most prominent explanations has been Trump’s absence in the final Iowa debate, along with the lack of a ground game to drive voters out.

Despite the search for answers, all theories fail to provide a satisfying explanation for the ten-point error in polling. In fact, it remains dubious whether the two excuses are valid at all. After Trump’s absence in the Iowa debate, there was no decline in his polls both nationally and in New Hampshire. As for the lack of a voter turnout, the exact opposite happened—with caucuses topping projections from even the most optimistic estimates. Additionally, Trump won the vast majority of first-time caucus-goers, a tested sign of voter enthusiasm.

The answer may be much simpler: Trump is the lovechild of a Republican splinter group, and will fail to expand his support to include a broader base. Trump’s candidacy is much more like that of Pat Buchanan, an ultra-conservative who previously served as the White House Communications Director. Despite leading in the polls by large margins, he was upset in the Iowa caucuses by lesser-known Bob Dole, who would go on to become the Republican nominee in 1996. In this way, despite all the sensationalism, factional candidates thrive within the fragmented environment, until the establishment consolidates behind a single candidate. While divisive candidates may seem unstoppable in a crowded field, once moderates and the base rally around a single candidate, the splinter group, once a formidable opponent, is tiny in comparison to the much larger centrists.

The media focus on Trump has provided him with 24/7-cable news coverage that any candidate would kill to have. Unfortunately, this support has plateaued. Polls in the early stage can offer the illusion of being the most reliable analysis, but until Trump’s support among Republicans nationally can rise above 50 percent, the attention is more hype than actuality.