Donald Trump heads into March like a lion, leading polls and looking to emerge with a delegate count that may put the Republican nomination away. Marco Rubio and Ted Cruz are taking their whacks at him more aggressively than before – and even if they can’t nudge Trump out of the race, they can extend the nominating contest until it falls under the arcane rules of the party convention.

All these turning gears have the more speculative wondering about whether a serious third option could appear on the November ballot for the first time in 20 years. Trump supporters want their guy to have a spot if they feel the GOP finds some kind of black magic to nominate someone else; Republicans fear the Trumpocalypse and don’t want to have to write in Mickey Mouse against Hillary Clinton. These are both significant audiences, so an independent candidate seems like it could make some waves. Michael Bloomberg has been the biggest name to consider an independent run so far.

So could it happen?

The major problem is logistics, as Ballotpedia’s page on Presidential ballot access makes clear. While the major political parties pretty much have a free spot on each state ballot, running an independent bid means petitioning 50 separate state election authorities. Signature requirements range from 1,000 in Idaho to nearly 180,000 in California. (Thresholds for getting on primary ballots tend to be easier, and petition signatures can be supplanted by filing fees.)

Getting on all 50 ballots means collecting over 900,000 signatures. But wait, there’s more: As anyone who has handled ballot access can attest, fake and invalid signatures are a major problem. People who sign petitions may not be registered to vote, or they may use a fake name, or they may violate some other arcane rule (such as Nebraska, where a signatory must not have voted in either party’s primary). Campaigns generally try to capture at least twice as many signatures as needed for just this reason, so the real magic number is about 1.8 million signatures.

The first deadline for access is in Texas, where a candidate needs about 80,000 signatures by May 9. That’s significant, because an independent offshoot of the Republican primary would surely look to Texas as an opportunity to pull support. And though a smart operation might cherry pick friendly states to focus efforts there, such a plan requires much advance data work. Either way you slice it, a third party effort has to start almost immediately.

The candidate would almost have to be a self funder, or have access to a very generous fundraising network; they would also have to have a good amount of political savvy to build the organization necessary for the task. Most of the current candidates couldn’t pull off the optics of positioning for a third party run while also running for the nomination, but Trump could probably get away with it. If Trump is the nominee, Mitt Romney, Carly Fiorina, and Jeb Bush are in the sweet spot of the money/strategy Venn diagram, though it’s tough to imagine they would do more than split votes and toss some close states to Clinton (or Sanders).

Rick Perry has floated the idea that he’s open to a second crack at the nomination at a contested Republican convention, but he offers a compelling case as an independent candidate as well. Winning Texas and maybe a handful of southern, western, and midwestern states could disrupt Electoral College totals enough to push the race to the House of Representatives. Another, center-left independent (like Bloomberg or Jim Webb) would make that outcome even more likely.

It makes sense why the prospect of a candidate beyond the two major parties holds considerable sway this cycle. Yet, the election laws in place greatly discourage it. Beyond smaller third parties and failing an indictment, Americans are likely stuck choosing between the two candidates who emerge from the party conventions this summer.