When Republican presidential candidate Sen. Ted Cruz came to the Bronx to campaign in the run-up to the New York primary, ridicule was not far away. After all, what is a Republican candidate doing in the staunchly Democratic, minority-majority northern part of New York City? What is he doing in New York State at all, when Donald Trump is nearly certain to carry the most votes and, according to market-based forecasts, has an 85 percent chance to win more than 50 percent of the vote?

Why would Ted Cruz campaign in the Bronx?

For someone like Cruz to campaign exactly where Republicans are least popular is actually a very reasonable strategy. That’s because 81 of 95 delegates are allocated equally between the 27 congressional districts. While congressional districts are drawn by population size, the number of registered Republicans per district varies widely in New York.

As a consequence, some congressional districts will have their three delegates decided by just a few hundred voters.

Nowhere is this imbalance as stark as in the Bronx, which includes New York’s 15th district, and parts of the 14th and 16th as well.

For this calculation, we predict Republican primary voters by congressional district using a voter file provided by TargetSmart. We count as likely primary voters all registered voters who were registered with the Republican Party as of 2014, given that the New York Republican primary is closed, and who participated in the 2012 presidential primary. We expect as few as 211 voters to turn out in the Republican primary in New York’s 15th congressional district, where Cruz campaigned.

Of course, our prediction is partly based on historical turnout, and turnout this year might be slightly higher. The Republican primary remains very competitive, and we base part of our estimates on the 2012 primary, when Mitt Romney had all but wrapped up the nomination. While New York districts were redrawn after the 2012 primary — for example, the 15th congressional district of today used to be, more or less unchanged, the 16th in 2012 — the 2012 turnout numbers are comparable to our prediction.

No more than 285 Republicans voted in the primary here in 2012. Primary turnout in the low three digits per district is common in New York City. On the highest end, we expect the Republican turnout to be in triple digits, so just a few dozen Republican voters in the Bronx — one restaurant stop’s worth — could be enough to tip the election in that congressional district.

What does this mean for “one person, one vote?”

Clearly, “one person, one vote,” that mantra of U.S. democracy, is not the operating principle in the complex rules of Republican delegate allocation. Below, we chart the vote-to-delegate ratio of all New York congressional districts in relation to the 27th district, located upstate between Rochester and Buffalo.

According to our calculations, the 27th will have the most Republican primary voters of any New York state district. Therefore, that district’s Republican primary voters have the least influence on delegate selection of any district in the state.

Substantively, a Republican primary vote in the 15th district is about 113 times as important as a Republican vote in the 27th. A Republican primary vote in the 4th district, north of Long Beach, is about 31 times as important. A Republican primary vote in the 13th, New York City’s Upper West Side, is about 29 times as important. These ratios are unlikely to change, even if turnout in this year’s primary will be higher, as we assume turnout in all these congressional districts will increase over 2012 somewhat proportionally.

What’s more, the way New York State’s Republican Party distributes its 95 delegates makes votes in the districts with few Republicans even weightier. If a candidate gets 50 percent of the primary vote share in any of New York’s 27 congressional districts, that candidate receives three delegates. If no candidate gets an absolute majority in a district, the winner takes two and the second-place-finisher takes one delegate, as long as both clear a 20 percent hurdle.

Cruz isn’t necessarily aiming to win a district. If he comes in second place in a district with just a few hundred votes, that could be enough to get a delegate — and in this Republican nomination year, every delegate will count at the convention.

Take a look at the 15th district to assess the likely cost of one delegate. We compare the 15th to its Manhattan neighbors, the 13th district and the 14th district, which are also overwhelmingly Democratic and will have few Republican primary voters. In a three-candidate field, we can conceive three different scenarios: a blowout win, in which the majority candidate takes 90 percent of the vote and therefore all three delegates; a razor-thin majority, in which the majority candidate takes 50 percent plus one vote, but gets all three delegates; and a graded outcome, in which the more popular candidate takes 45 percent of the vote, a second candidate takes 30 percent and one delegate, and a third candidate takes 15 percent and no delegates.

In the razor-thin scenario, taking one 15th district delegate would require only 35 primary votes. That is one of the best vote-to-delegate ratios in the country. In the blowout scenario, one 15th district delegate would require 63 votes; winning one delegate in the 14th would require five times as many in that scenario. And there’s another scenario we don’t show in the figure, because we lack the space: To get one delegate in the razor-thin-majority scenario in the more mixed 27th congressional district would require winning 3,974 primary votes, more than 100 times as many as in the 15th under the same scenario.

In other words, Cruz has many incentives to target congressional districts with few Republican primary voters, since winning even one matters more than in any other district in the state.

Cruz could easily have met half the district’s Republican voters on April 6.

When Cruz visited the Bronx on April 6, he stopped at a Dominican-Chinese restaurant favored by Hispanics. Based on our voter file data, we predict more than half the likely Republican primary voters in the district to be Hispanic. Those 110 or so Hispanic Republicans could easily fit in a Bronx restaurant. If Cruz were to win every one of their votes — no matter what his overall numbers in New York State as a whole — he will take three more delegates to Cleveland.

Polls make it clear that Trump will win in New York. He will probably get more than half the vote, and the vast majority of the delegates.

But keep an eye on Cruz’s delegate haul from majority-Democratic districts, particularly the 15th. If he wins, or finishes second, he makes it that much more likely that the Republican convention will be contested. No matter how many votes it took to acquire each delegate, once in Cleveland, each delegate’s vote counts equally.

Tobias Konitzer, PhD candidate in communication at Stanford University

David Rothschild, economist at Microsoft Research, posts at PredictWise.com. Find him on Twitter @DavMicRot.