Watching poll-watching is an entertaining study in psychology.

Let's say there are three leading primary candidates: Antelope, Bear and Cougar. DUer Fox supports Antelope, Goat supports Bear and Horse supports Cougar. One day a poll comes out showing Antelope leads in Iowa by 3%. Fox starts a thread celebrating this poll:

FOX: Antelope is ahead by 3% in Iowa!

GOAT: Meh. That poll is an outlier.

HORSE: Iowa doesn't mean anything. Cougar is ahead in New Hampshire.

GOAT: Which poll says Cougar is ahead? Galumph said Bear is ahead. Try to stay up to date.

HORSE: Galumph's margin of error is more than 3%.

FOX: Iowa is more important than New Hampshire.

GOAT: Bear has a better ground game in Iowa. That poll doesn't mean anything.

HORSE: They polled registered voters, not likely voters.

FOX: All the polls say Antelope takes Iowa.

GOAT: No, they don't. Anyhow, Bear gets the youth vote and the polls never call cell phones.

FOX: Young people just go to rallies, they don't vote. Anyhow, Antelope's rallies are bigger.

HORSE: They don't take polls at rallies.

FOX: No matter what the polls say Antelope wins Iowa and New Hampshire. Count on it.

Plug this conversation into any discussion of polls on DU. It's funny how an analysis of a poll depends on whom the analyzer supports. Confirmation bias is a thing.