Voting in the Democratic Primary contest is getting ready to heat up, but the contest itself will be more-or-less over shortly after it begins. By the end of March, fully two-thirds of the delegates to the national convention will be awarded. Barring a major collapse in his poll numbers between now and then, Bernie Sanders is likely to win the lion’s share – perhaps even a majority – of these delegates. The results from Saturday’s Nevada caucuses certainly did nothing to suggest otherwise. At that point, denying him the nomination becomes an extremely tricky proposition.

If you are among those who believe Sanders is the strongest candidate Democrats could run against Donald Trump, then this is all for the good. If you think the outsider is a general election problem for the party, you may be wondering how it came to this, especially in an election where there seemed to be so many promising candidates. Particularly perplexing is the decision of the Democratic candidates to hammer Michael Bloomberg – who won’t be on the ballot for two more contests – at the most recent debate.

Here are a few thoughts:

1. Megafields are different in an era of weakened gatekeepers.

For most of the past 50 years, primaries have unfolded as follows: A group of 10-12 (give or take) candidates declare their desire to run for president. As they meet with party officials and big money donors, the “invisible primary” unfolds, where the party decides who it would like to see become the nominee. By the time Iowa holds its caucuses, there are only three or four viable nominees, and after New Hampshire, just one or two. Not every race has followed this template, but usually the races are effectively over by mid-March, even if all the candidates don’t know it.

The 2016 Republican and 2020 Democratic primaries have been different. In both instances, the field emerging from Nevada has been roughly the size of a typical Iowa field.  The field will likely narrow further after South Carolina, but there is a reasonable chance we will have four or five serious Democrats competing on Super Tuesday.

These sorts of fields are fertile soil for factional candidates to use their basic level of support to take root in the primary field. Trump probably started out with the firm support of around 20-25% of the Republican Party. But because the other votes were spread out over multiple candidates, that 20-25% produced strong showings in early states. Once the candidate wins the early races, it allows him to take hold as a legitimate candidate and then proceed to broaden his appeal.

That’s what happened in 2016, and it’s what is happening in 2020.

2. The collective action problems in large fields are real.

In 2016 there probably was a path to stopping Trump from becoming the nominee, but it involved an anti-Trump alternative forming early on. Even by Super Tuesday, Trump trailed most of his Republican rivals in head-to-head exit polls. But this created problems for the anti-Trump forces. Each candidate had a plausible claim to the nomination, and each candidate thought that if he outlasted his opponents, he would win the one-on-one. This contributed to candidates such as Jeb Bush, Marco Rubio and John Kasich all staying in the race well past the point where they had a strong claim to the nomination, preventing the anti-Trump vote from consolidating in the process. In a universe where Rubio wins Iowa and comes in second in New Hampshire, or Ted Cruz finishes stronger in New Hampshire, the field potentially winnows, and you have a different outcome.

A similar effect is playing out on the Democratic side. Pete Buttigieg won Iowa under some counts and finished second in New Hampshire. He has a legitimate claim to lasting until Super Tuesday. Bloomberg’s strategy is predicated on a late surge. Joe Biden likewise has stated that he plans on spring-boarding from a strong showing in South Carolina to Super Tuesday. Those three candidates all have a legitimate claim to staying in until Super Tuesday. Massachusetts and Minnesota also cast ballots on Super Tuesday, giving Amy Klobuchar and Elizabeth Warren an incentive to stay in so that they might win crucial delegates for a potentially brokered convention.

In other words, even though it might be rational in the aggregate for candidates to drop out, at an individual level, candidates are incentivized to stay in. Without strong party actors intervening, there’s no reason to expect this field to narrow quickly.

3. The Internet changes everything.

For most of the 1990s and 2000s, political observers were prone to write “the Internet is going to change everything.”  But for most of those decades, there were only glimmers of what was to come. George W. Bush’s fundraising, the Howard Dean campaign, the rise of the blogosphere, the fall of Trent Lott – all of these things seemed to signal radically different politics lurking just around the corner.

The future is now. With online fundraising (with an assist from the rise of Super PACs), candidates are able to bypass the traditional party gatekeepers to a degree that was unthinkable in the 1990s. At the same time, the rise of sites like Facebook, Twitter and YouTube allowed non-traditional candidates to spread their message without having to wait for a call from “60 Minutes.”

This isn’t to say that the traditional gatekeepers have no role anymore, only that it is greatly diminished. This has allowed candidates like Sanders and Trump, but also like Tulsi Gabbard, and Ron Paul, to extend their campaigns. In 2000, John McCain was forced to suspend his insurgent campaign for the presidency because his money ran out. In 2020, that would not be an issue.

4. Unthinkability bias.

Finally, beyond the enhanced fundraising and core of strong supporters, candidates like Trump and Sanders have benefitted from what I dubbed “unthinkability bias” in 2016. This is like a strong version of confirmation bias, where people set their prior probabilities of an event occurring to zero and refuse to update their priors because they just can’t conceive of the possibility coming to pass. In the summer of 2016, I wrote a series of pieces suggesting that Trump could win the general election; these were met such derision and invective that I took a month-long break from Twitter. In late 2019, I wrote a piece suggesting that people were underestimating Sanders’ chances of becoming the Democratic nominee. While the response was less angry, it was still met with a degree of skepticism that seems unwarranted today.

This is harmless in and of itself, but it has the potential to transform the trajectory of races. By writing off Trump and Sanders in the summer and fall before the election year, candidates allowed their candidacies to strengthen and their core bases of support to grow. In addition, by attacking each other, the more “traditional” candidates collectively weakened themselves, diminishing themselves compared to the insurgent.

Of course, there is still a lot of time left in 2020, and Sanders isn’t the nominee yet. But if the above is true, he’s hardly the last insurgent candidate we might expect to see take off. Pundits, candidates – and above all, party power brokers – need to adapt to the new reality, or this will repeat itself in future years.

Sean Trende is senior elections analyst for RealClearPolitics. He is a co-author of the 2014 Almanac of American Politics and author of The Lost Majority. He can be reached at Follow him on Twitter @SeanTrende.