• While waiting for Axel's "proof", I conducted my own experiment.

    From MK@21:1/5 to All on Sun Dec 12 01:38:19 2021
    Today all conditions were right for me to play a session of 100 games
    with the intentions of trying to do well against XGR++ in Ironman mode,
    playing as consistently as I could, to defy the so called "skill" using the cube more aggressively than ever before, along the lines of what Axel
    is experimenting with, raccooning all opening beavers and even some
    later ones in the game.

    I played in one sitting, so to speak, over 9.5 hours with about 2.5 hours
    total breaks for eating, etc. but not doing anything else that would take
    my interest/focus away from the experiment.

    Considering that I picked higher than many defult settings for XGR++,
    I would think that, with about 4.2 minutes per game average, this would
    be considered pretty fast playing. In fact, I think most of the time I was waiting for the bot th move.

    In the end, I won 262 vs 188 points, for a total of 450 even. I don't know
    what is typical average game value in the gambling world out there but
    I think 4.5 would be higher than typical?? I don't know how many games
    I won. I will tabulate and post to my site later. I tried to win more games than in my previous experiments also, even cashing early early, etc.

    I broke my own cube error record with 40.75! :) With my checker error
    of 8.89, my overall error of 13.38 netted my +86 point while I should
    have lost -194. That means I won 280 points more than my error rate
    deserved. I am very satisfied with my achievement. Along with the
    two other 100-game sessions I played during the last month or so,
    producing similar results, I feel I have proven well enough that the
    "cube skill" is indeed a pile of elaborate, fanciful bullshit. I am now
    quite confident that I keep replicating my results comfortably.

    I will do my own kind os analysis of the session, tabulate some data
    into a spread sheet and post it to my site, along with the session file,
    player profile, etc.

    MK

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From MK@21:1/5 to All on Sun Dec 12 01:57:28 2021
    On December 12, 2021 at 2:38:21 AM UTC-7, MK wrote:

    Today all conditions were right for me to play a session

    I just noticed a big oops in my experiment. I won the
    opening roll 98% of the time. :)

    I used a dice file with a million rolls from Bgblitz website
    with my simple dll to read from it. While testing I edited
    the file to have the higher number on the right in order to
    see if the sequences were matching what XG displayed
    with the higher number always on the right. Then I forgot
    and kept using the same file for my experiment.

    I suppose, instead of throwing away the whole session, I
    can play another 100 games reversing the rolls and letting
    XG start 98% of the time... With that, there should be a lot
    more beavers/raccoons. (I was wondering why there were
    so few.)

    Bummer. :(

    MK

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)