To expand on what Keith and Floyd are saying:
Practical considerations:
In a world where leadout flex and friction did not matter, you could put the bellcrank any place where it would fit as it traveled through its range. (In fact, some early control line planes hooked the leadouts to the elevator horn and put pulleys in the fuselage or wing to help them turn the corner. This worked, kinda, but was really done to dodge Jim Walker's patents, rather than for any functional reason).
What matters to the flying of the plane is the position of the leadouts on the wing, the amount of friction, bind, springback, throws vs. line motion, etc., in the control system itself, the whole shebang being able to withstand the pull of the aircraft on the lines, and the CG position.
That having been said, the more that you ask the leadouts to bend on their way from the bellcrank to the ends of your lines, the more that you are flexing the leadouts and the more friction you are creating between the leadouts and the leadout guides. Similarly, the way that you mount the bellcrank with respect to the flap or elevator horn affects the friction, springback, and throw ratios of the control system.
In general, bellcranks are placed so that the link to the outside world (usually the flap horn on a stunter, or the elevator horn) is a straight shot without opportunities to bind. You generally see them placed as forward in the wing as they will go without making the wing construction inconvenient (which is why they're often a bit behind the CG). You will sometimes see bellcranks located at a tilt when viewed from the side of the plane: some people feel this gives a more favorable geometry with respect to the flap horn (and will often couple this tilted bellcrank position with a flap horn that has a dog-leg in it).
Social considerations:
There are some people who are convinced that the above is not true. They will tell you that where you put the bellcrank in the wing changes how the plane flies because of the direction that the leadouts pull on the bellcrank. The notable myth is that the bellcrank must be as close to the CG as possible, and moving it back destabilizes the plane. You must choose yourself whether to believe them -- but keep in mind if you do that they are going against some pretty basic principals of physics. The position of the bellcrank in the plane does matter for the reasons I gave, but it's not going to make the plane stable or unstable unless there's friction, spring, or bad control geometry involved.