Are all three in the same run (or "session")? Or are they in different runs? From your post it sounds like they're all in the same run.
This type of thing is always problematic, because the variable of interest (M, "manner of play") appears to be confounded with time. If there's a control condition, C, then you could compare M1 - C1, M2 - C2, and M3 - C3 (where 1, 2, 3 denote first, second, third). If there's no condition C, then IMHO things are pretty hopeless, because of the confound with time, _especially_ in the case that the three manners of play always occur in the same order.
|