I'm currently helping develop a rhythm game, and as part of the scoring system, I'd like to make scores and ranks based more on accurate pattern recognition rather than hitting notes very exactly on time as is the case for many games of this variety.
In many cases, the margin of error for the best judgment in rhythm games is so small that it seems to be a matter of chance more than skill that a certain judgment is counted between "excellent" and "good" (for example, one game has the margin as small as ±0.011 s).
Of course, some people are better at hitting notes accurately than others either naturally or through practice, but I would like to find some actual research into what the base neurological limitations to rhythmic precision are and to what degree practice can actually improve it.
As a baseline benchmark for the best judgment, I was thinking of looking for some research into how accurately people can follow the tick of a metronome - for example, by measuring the average deviation of somebody tapping a button to a regular beat after about 100 beats or so when they've gotten used to it.
A Google search didn't return anything useful, probably because the search terms draw up so much other, irrelevant content. Is there any prior research on this subject? If so, where can I find it? If not, how could I conduct such an experiment myself?