This is Part 4 in a 7-part series about world poker champion and cognitive scientist Annie Duke, Ph.D., whose insights follow each question below. To read from the start of the series, see Part 1.

Jenny Grant Rankin: How is the remaining uncertainty problematic?

Annie Duke: It’s a recursive problem. Suppose I've got this adverse signal: I learned new stuff, and it's not good. Now I have to decide whether to stop, but I'm still making that decision under conditions of uncertainty. Because if I stop, I don't ever get to know, “Well, maybe I could have turned it around. Maybe it actually would have worked out well.” From where I’m currently standing, I have to make an estimate of how the situation is going to turn out if I stay, compared to if I stop and go on another path.

We see this in firing all the time, where someone will say, “I've got this person in seat, but they stink.” I’ll say, “Why aren't you firing them?” and they’ll say, “Well, I've put some time into them, and they're........

© Psychology Today