Did you pick the winners in the Iowa primaries? How about New Hampshire?

If not, you’re in good company. The renowned Gallup Poll didn’t either. In fact, according to Gallup Editor-in-Chief Frank Newport in a speech last Thursday at Princeton University’s Dodd’s Auditorium in Robertson Hall, Gallup is backing off from the predicting business — no more “horse-race polling.”

Speaking to an audience of about 90 University students and visitors, Mr. Newport emphasized that forecasting is only one of three major contributions that public opinion can make to society and that Gallup has decided that understanding and guidance are much more important

Ever since the founding of the organization in Princeton by George Gallup in the 1930s, there has been great controversy over polls and the role of public opinion in presidential elections. Recently the polls greatly underestimated Justin Trudeau’s election victory in Canada, mis-predicted David Cameron and the Tories’ victory in England and mistakenly called a victory for Donald Trump over Ted Cruz in last week’s Iowa primary.

The change in the Gallup organization’s attitude toward the popular practice of forecasting, or horse-race polling, began after the 2012 presidential election as a result of Gallup’s erroneous prediction of a Mitt Romney victory over Barack Obama, and resulted in a far-reaching review of all of Gallup’s operations.

A sociologist and the author of books titled Polling Matters: Why Leaders Must Listen to the Wisdom of the People (2004) and God is Alive and Well: The Future of Religion in America (2012) and editor-in-chief at Gallup since 1991, Mr. Newport described a long history of problems with polling, citing a recent Frank Bruni New York Times op-ed “Our Insane Addiction to Polls” as supporting evidence.

“It’s a controversial job,” Mr. Newport explained, as he went on to emphasize the importance of understanding (“what’s going on in an election and why, what’s behind the vote, what people are thinking”) and guidance (assessing “the collective wisdom of the people and the positive role it can play in moving a society and democracy forward”) — as opposed to predicting.

The resources of the Gallup Poll, he argued, are much more valuable in contributing to an understanding of the issues and priorities. “We need evidence between votings,” he said, so politicians know “the issues that are most important to the public, to particular groups”

In spite of the high level of interest in forecasting — “Yes, Gallup made it famous. Yes, Americans love a contest. Yes, forecasting can help keep the process honest in cases of possible voter fraud. Yes, it provides a mechanism for decision-making (for example, who gets to participate in the debates)” — Mr. Newport argued strongly for the “tremendous value in spending more time on the understanding and guidance functions of opinion polling.

In a short follow-up speech, Princeton University politics professor Chris Aiken praised the Gallup organization’s great contribution “to get us out of the little world in which we all live” and to “give us a far more honest picture of what people are thinking.”

Survey research and predicting “is a tough racket,” Mr. Aiken said, but he urged Mr. Newport to “get back in the game.” Mr. Aiken emphasized the importance of Gallup’s role in election forecasting, stating, “You can do the game better than the competition can do it. Gallup has the resources and the talent to step-up the level of the game.”

Mr. Aiken contended that recent miscalculations by opinion polls were caused by errors in predicting voter turn-out — that voter turn-out assessments, for example, should have made it clear that Cruz had the advantage over Trump in Iowa.