NASA Lecture on ExoPlanets The Study of positional error in Astronomy. Just based on the graphic (which is just a camera shot of a projector in lit room). Apparently, most new planets are just math abberations. They are calling positional errors that are less than 1 sigma from the mean position (with 2 or 3 observations) planets.
I don't know about you, but If I tried to set up a manufacturing process with that level of confidence (minus 2 degrees of freedom based on what I saw.) I could expect my new process to fly like a lead pipe.
In this same lecture, she alluded to what's causing this fuzzy math. At NASA, . . . We like to issue press releases . . . pressure on a young guy . . . interesting numbers. . . . funding problems. It appears NASA is playing games just like ENRON was. They do what they have to to produce news for their customer (Congress.)
If you disregard the fuzzy math, the lecture is still fascinating look at IR telescopy. It really does increase visibility into space by a multiple of 1000's. Just like sound travels better in water than light, but light travels better in the air than sound, IR travels better in space than visible light.