[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

thought experiment

Posted By: Chuck Bower
Date: Tuesday, 24 February 2009, at 5:37 p.m.

In Response To: we need a lot more Al Fallers (Matt Cohn-Geier)

Suppose you could program a bot to play like a specific human, technically. Do that for the entire field at Pitt except for one player -- Al Faller.

Conduct two repeated trial tournaments, one with the real Al Faller against the real players, the second with the real Al Faller playing against the cloned bots. Repeat sufficient enough times to get statistical significance you're looking for. Will Al's performance in the two events be similar, i.e. as many main wins, main runnerups, consolation wins, etc?

Al makes many unconventional plays, some of which are clearly 'wrong', others so imaginative that we may not be sure. Humans hone in on the wrong stuff (and maybe the other plays, too) and think "I should steamroll this guy." Then Al, with some lucky rolls, starts building a lead in the match. The human gets annoyed: "this guy can't play his opening rolls correctly and yet he's beating me!" Cool water --> hot water --> (you know what's next).

Meanwhile Al's just cruisin' along thinking "this is fun!" He just plays his usual erratic (what it appears to some of us, anyway) game.

No doubt in my mind (or anyone I've heard from) that Al was lucky this weekend. (Very likely so were Bob Glass, Neil Kaz,...) There was a game against Michael Edge (unfortunately not recorded so it runs the risk of exaggeration) that was truly "one for the ages". What I'm wondering is if he got any unintended help along the way.

Messages In This Thread

 

Post Response

Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:

 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.