As you can see, the dots are bunched around the most probable value and taper off in either direction—in the same way that the the area under a bell curve decreases away from the center. s, the distance from the blue line to Mars divided by the size of the error, is 3.7, giving a probability of 10,000:1.
Notice that the scale here is 500,000 km, so this is zoomed out by a factor of 5 from the 9 January picture. Two things have happened in the fortnight. First, the position of the blue line has changed a little. More importantly, the size of the error was a lot bigger two weeks ago. Back then the error was large enough so that the distance from the blue line to Mars divided by the error was only 2.2, giving a probability of 25:1.