Those numbers are a wee bit more compelling.
I notice that the 1-set group start out generally weaker, and the 8-set group start out generally stronger. I also notice that there's a small improvement from 1 set to 4 sets, and a huge improvement from either of the above to 8 sets. All things being equal, you'd expect the stronger person to progress at a slower rate, and you'd expect that each successive set would promote less of a training benefit than the previous set (so, the second set might only provide an additional 50% benefit of the first set, the third set might only provide an additional 50% benefit of the second set, and so on). So, I would have predicted that there'd be a much bigger jump in progress between the 1-set group and the 4-set group than the 4-set group and the 8-set group, both due to their starting points and due to the principal of diminishing returns. The fact that this isn't the case raises additional questions about the test subjects and about the benefits of additional sets.
If the 1-set guys were all 70-80kg and the 8-set guys were all 110-130kg (quite possible, given the requirement that they all be able to squat 130% of their bodyweight), we then have to consider how much of the progress occurred due to the workload, and how much occurred due to what was being worked with. If we can confirm that the huge difference in results is due to programming rather than the bodies doing the program, this data would suggest that we re-evaluate our understanding of diminishing returns, at least at an acute level. It would also prompt us to investigate further what happens as the sets go on to cause such an excess in results.
