The first U.S. presidential poll to use modern statistical methods was a Gallup poll in 1936. But the first known presidential straw polls date to 1824—over a century before.

These early straw polls were regional and informal. Newspapers reported the results as information about local opinions rather than possible predictions about how the national election might play out.

Local straw polls continued throughout the 19th century. Then in 1916, The Literary Digest launched a national presidential poll. The magazine’s methods were flawed, but for five consecutive presidential races, the winner of The Literary Digest poll was also the winner of the actual election.

This streak ended in 1936, when George Gallup predicted Franklin D. Roosevelt would win reelection and The Literary Digest said he wouldn’t. Gallup’s poll was a victory for statistical survey methods, and paved the way for modern presidential polling.

The 1824 presidential election was unique in many ways. The caucus system that U.S. politicians had thus far used to nominate candidates had fallen out of favor, and politicians struggled to adopt a new nomination process (the national party convention didn’t debut until 1831). The country’s first party system had also fallen apart, so the only candidates to choose from were Democratic-Republicans—and there were four of them running at the same time in the general election.

In this chaotic situation, white male voters began to take straw polls to see who others were voting for. They took these polls at all types of gatherings: militia assemblies, grand juries, Fourth of July celebrations and elections for other offices. We know about these polls because newspapers reported on their results. In contrast to today, newspapers didn’t frame these results as predictive of the national election.

“I think people took this as interesting and informative, but not in any sense definitive,” says Tom W. Smith, director of NORC’s Center for the Study of Politics & Society, who has written about the 1824 straw polls for The Public Opinion Quarterly.

It would’ve been hard for any poll to predict what happened in the 1824 election, anyway. Andrew Jackson won the most popular and electoral votes; but because he didn’t win the majority of electoral votes, the U.S. Constitution dictated that the U.S. House of Representatives choose the president. The House elected John Quincy Adams, who had come in second in the popular and electoral votes (the other runners-up were William H. Crawford and Henry Clay).

Although newspapers reported on the 1824 straw polls, it doesn’t appear that they organized any themselves. As the 19th century progressed, newspapers began conducting their own polls, both as self-promotion and a way of gathering information for readers. Still, these straw polls remained regional, and news publications don’t seem to have framed their polls as electoral predictions.

This changed in the early 20th century, when The Literary Digest introduced a national presidential poll that gained a reputation for correctly naming the electoral victor.

In 1916, The Literary Digest conducted a national straw poll about the presidential election. The poll favored incumbent president Woodrow Wilson, who won reelection later that year. Four years later, the weekly magazine sent out postcards for another national presidential poll. Once again, the winner of The Literary Digest poll—Warren G. Harding—also won the real election.

The Literary Digest’s predictive streak continued for three more elections, as its polls correctly anticipated the presidential victories of Calvin Coolidge in 1924, Herbert Hoover in 1928 and Franklin Delano Roosevelt in 1932. The polls gained the magazine a lot of attention, including from a man named George Gallup.

Gallup got his start in public opinion polling while working at an advertising firm. In 1932, he conducted a poll to see if his mother-in-law, Ola Babcock Miller, had a chance of becoming Iowa’s first female secretary of state, and it correctly predicted she’d win. A few years later, he founded the American Institute of Public Opinion, which later became the Gallup organization.

In 1936, Gallup publicly challenged The Literary Digest’s polling methods in a bid to promote his own. He conducted a presidential survey predicting FDR would win reelection, which clashed with The Literary Digest’s poll predicting Kansas Governor Alf Landon would beat him. The public challenge paid off when FDR won the election in a landslide. The victory helped promote Gallup’s methods while discrediting those of The Literary Digest, which folded a couple of years later.

Gallup’s organization continued to conduct polls on public opinion and presidential preferences, running into some issues along the way. One of the most infamous hiccups was in 1948, when Gallup’s company—along with other polling firms—predicted that New York Governor Thomas E. Dewey would beat incumbent president Harry S. Truman.

With the polls in agreement, the Chicago Daily Tribune felt confident settling on “Dewey Defeats Truman” for the early morning headline the day after the election, leading to the famous photo of a victorious Truman holding up the erroneous newspaper headline.

Despite similar “surprise” victories over the years, presidential polls have continued to be a major part of election news, and also become a tool political parties use to test the strength of their own candidates.