I have always thought it more or less established fact that the games were better late at night (Danish time, GMT+1). The logic is that at around midnight everyone in the western world will NOT be at work, and that sleeping schedules is less of a hindrance for poker playing than work schedules. Also the European players who are on roughly the same local time as me will supposedly play worse late at night for various reasons.
What is lacking is some hard proof that this is actually fact, and not just an example of the standard human mindset that the grass is always greener on the other side, or in this case, that money flows more freely when I can't play 'cause wifey wants me in bed.
During my marathon, I did not have the feeling that the games were much better at night time, and surely my results are an indication of the opposite. My day time sessions (before dinner break) produced a win rate of 3.6 BB/100, while my night time sessions produced a win rate of 0.22 BB/100. Obviously those results are not proof of anything, for example a break even session on Sunday night instead of a -17k session would have made the two win rates roughly equal.
If you have some solid evidence one way or the other, I would be pleased to hear of it. Unfortunately Poker Tracker 2 was not much help in this regard. Right now I am working with PT3 to see if there is functionality there to help me (I am pretty sure there is, but I am having some techical problems and am in contact with support).
Fellow blogger Joxum asked for showdown statistics during the marathon. I went to showdown on 26.4% of the hands were I saw the flop, and I won 56% of those showdowns.