I go along with Erik in this discussion. Light can be easily modulated at much higher frequencies than 100 Hz. Thus you can move entirely away from the power line frequencies of the field lights. It will also turn out that the filter like Charles suggested will be very useful for the filtering of the higher frequency. This would be particularly true if you choose a higher frequency that was not an exact multiple of 100 Hz.

For example if you had a microcontroller operating at 11.0592 MHz you could use a timer or PCA channel to divide by 3000 down to an operating frequency of 3686.4 Hz. That would seem to be a good frequency to use and easy to setup a switch capacitor filter for the receiver filtering. With switch capacitor filters such as...
http://www.national.com/ds/LM/LMF100.pdf
...it is easy to establish the filter characteristics via a clock signal fed into the filter. The above mentioned part can have its center frequency for a notch or bandpass filter set via a 100x frequency clock. This can be attained for the same microcontroller setup by acheiving a divide of the 11.0592 MHz by a factor of 30 down to 368640 Hz which is 100 times faster than the frequency I suggested for the ball emitting frequency.

Good Luck
Michael Karas