I presume he means Light Dependent Resistors.
They could well be the problem Rich, as they may be moving between a low and high resistance with the changes in light he is creating, but not going completely open circuit/closed circuit.
Failing that, the
comm 
port may be wrong, but it sounds more like he can get a signal by shorting, or even by blanking out in test . . so the LDRs are the more likely culprit.
amblipijido - You would be better off using a decent photo-transistor. Preferably one that is described as a "darlington pair" (that really means two transistors in one, doing a sort of push-pull thing (like two rowers on one set of oars in a boat) to get more "ooomph"
I have heard of other folk who struggled to get LDRs to work reliably. - maybe just because they were not ideal for the degeee of light change being evident in their situation.
I have made so many leads up that way for timing I lost count. They almost never fail, I can only remember ever replacing one.
Couple with a compatible light source you will get 100% perfect lap counting.