That's to be expected; imagine the asteroid is projecting a cone in front of it that represents where we think it might go. Then take a cross section of the cone at the point where it meets earth. You can get a rough estimate for the odds of a hit based on how much area of that cross section the earth takes up. As observations and data come in our predictions get more accurate and the cone gets narrower. The circle of where it might be as it gets close to earth shrinks but earth doesn't so a higher percentage of the circle is covering earth hence the odds go up. But earth is not in the center of the circle so it'll shrink to the point where part of earth isn't covered by the circle which will cause the odds to suddenly start dropping as the circle shrinks past earth. This will be the case with every object that we spot that is going to be a near miss. The odds will climb up and then drop down.

I know most people hanging out in a space community probably don't need that explanation. But I've seen too many people elsewhere who seem to view the changing odds as a sign of unreliability and uncertainty.