By the way, for the scientifically inclined out there, average player numbers for Target for Today (one 12-hour frame) and the various 4-frame scenarios are not exactly the same statistic in the following way. The way I create these statistics is to look at the "players" stat in the log for each frame, and take the average over all frames.
In the case of the one 12-hour frame, I am getting the total number of players who took off at least once in that single frame.
In the case of the 4-frame events, I am getting the average number of players who took off at least once. The total number of players who took off at least once would be greater than the average but not as large as four times the average. It is greater than the average because the exact same set of players will not be present each frame -- some new players will be coming in from time to time while some former players are not in that frame. It is not equal to four times the average, though, because not *all* players are different frame to frame. However, I would not be surprised if the total number of players in a 4-frame event were 1.5 times the average.
So, if you want to compare a single 12-hour frame to multi-frame events in participation, you need to multiply the multi-frame average participation by a factor, perhaps something like 1.5.