zx has the right idea. Mechanical advantage is the same thing as leverage, which is all about moment arms. Mechanical advantage is achieved by linkage closer to the hub on the servo arm and out on the end of the control arm.

I pick up a Hitec standard short servo arm, and measure the distance from the rotor hub to the linkage holes: ~7, 10, 13 mm. I pick up a plastic ARF control horn and measure from the base to linkage holes: ~10, 15, 20, 25 & 30 mm. Add half the thickness of the control surface to the distances on the control horn--call it another 5 mm for convenience. So my choices at the control surface from hinge line to linkage connecting point are 15, 20, 25, 30 & 35 mm.

If I connect my linkage from the 10 mm hole (middle hole) on the servo arm, to the 30 mm hole on the control horn, then I have a mechanical advantage of 3 to 1, or 3:1 as I was taught in high school. The servo arm will rotate three times further in degrees of arc than the control horn will. So 45˚ of rotation at the servo will give 15˚ of rotation at the control surface. But the mechanical advantage will convert 40 inch-ounces of torque from the servo to 120 inch-ounces of torque at the control arm. This is a pretty good setup for sport aerobatic flying.

A 3D flyer would be happier with a 1:1 ratio, 45˚ servo rotation giving 45˚ control surface rotation--which is no mechanical advantage at all. Maybe an extreme setup would be 45˚ servo rotation giving 90˚ control surface rotation, in which case the servo would be on the short end of the power ratio, such that 40 in-oz torque in the servo would produce only 20 in-oz at the control surface. If you wanted more control torque than that, you'd need a high-torque servo.

You can, of course, get more rotation out of a servo than 45˚, so you can get a 4:1 mechanical advantage, or even 5:1 (7 mm at the servo, 35 mm at the control horn). I've had pattern planes where 9˚ deflection on ailerons and elevators was as much as I would want. This gives a somewhat slow but very controllable roll rate, and more than adequately tight loop diameter. So with standard servo I could link from the 7 mm hole on the servo arm to the 35 mm hole on the aileron/elevator control arm and multiply my 40 in-oz torque to 200 in-oz. That, by jiminy, is real mechanical advantage!

Along with the power, you also gain resolution--which for precision aerobatics is just as important as the higher torque. You do lose speed of the control surface reaction, since the servo has to travel farther to get the same deflection, but that has never been a problem for me. If it is, you can spend a bit more for a high-speed servo.