The first poll of Nevada Democratic caucus-goers in more than a month is out. A survey from the Las Vegas Review-Journal and AARP Nevada, released Friday, shows Sen. Bernie Sanders (I-Vt.) leading with 25 percent, former vice president Joe Biden in second with 18 percent, followed by Sen. Elizabeth Warren (Mass.) with 13 percent, business executive Tom Steyer at 11 percent and former South Bend, Ind., mayor Pete Buttigieg and Sen. Amy Klobuchar (D-Minn.) tied at 10 percent. But the poll, taken from Feb. 11 to Feb. 13, does not fully reflect voters’ reaction to the results of the New Hampshire primary and has a margin of error of 4.8 percentage points.

In other words, it does not tell us a whole bunch.

The Iowa caucuses and the New Hampshire primary should inform our expectations for the Nevada caucuses. First, do not be surprised if there are problems with Nevada’s process, which introduces early voting for the first time. Party officials scrapped a voting app similar to the one used in Iowa, but they are scrambling to come up with an alternative and to train its volunteer captains. What could go wrong, huh? (The New York Times reports: “Nevada Democratic officials announced new details on their plans on Thursday, writing in a memo that they planned to provide all caucus precinct chairs with an iPad and would rely on a calculator and Google forms to tabulate the totals.”)

The volunteers had been trained on the now-unusable app, and the campaigns reportedly have scant information about how it is all going to work. There will be paper backups, as in Iowa. Overall, expect reporting delays or confusion about how it will come together. (You have to feel badly for campaigns struggling to keep alive that are investing heavily in Nevada, who better hope this is not another debacle.)

The second lesson is an evergreen warning: The polls may be of limited value. As the early contests have shown, a debate just days before the voting can significantly change the outcome and will not register (or register fully) in polling. Klobuchar’s outstanding debate performance on the Friday before Tuesday’s primary vote in New Hampshire helped lift her to a close third. In Nevada, the caucus will occur just two days after the Feb. 19 debate. But early caucus voting will take place between Feb. 15 and Feb. 18, so those voters won’t have the benefit of the debate to help determine their pick.

Third, a caucus with a realignment process adds an extra degree of uncertainty. In both Iowa and New Hampshire, Buttigieg and Klobuchar over-performed. In the RealClearPolitcs averages, Buttigieg was at 16.8 percent in Iowa (where the last poll was more than two weeks before the caucus); his actual result was 21.3 percent. Klobuchar polled at 9.7 percent, but finished with 12.7 percent. In New Hampshire, Buttigieg also performed better than polling (21.3 percent vs. 24.4 percent) as did Klobuchar (11.7 percent vs. 19.8 percent). Sanders received about 3 percentage points less of the vote than his final polling average in New Hampshire, while Biden and Warren also dropped.

Either because he does better than others in the second alignment or because he finishes strongly, Buttigieg may do better than polling suggests; likewise, Klobuchar’s momentum may not be entirely evident in polling.

The media’s premature political obituaries for some candidates and their excessive certainty about the candidates’ continued viability may become self-fulfilling prophesies. Late-deciding Nevada voters concerned about “throwing their vote away” (i.e., voting for someone they do not want because someone they do want has be written off by talking heads) may make a difference — if not in the first alignment, then in the second. In short, since the punditocracy knows far less than it lets on, maybe voters should just vote for the candidate they like.

Read more: