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

BGonline.org Forums

Results: New bot-vs-bot cubeful-vs-cubeless experiment

Posted By: MK
Date: Wednesday, 21 August 2024, at 6:44 p.m.

In Response To: New bot-vs-bot cubeful-vs-cubeless experiment (MK)

These were Experiment 27 and Experiment 28. Here are the results:

In Experiment 27, GnuBG 3-ply checker and 3-ply cube skill, (strong bot), played 20,000 money games against GnuBG 1-ply checker and 3-ply cube skill (weak bot).

Strong bot won 24,247 (51.43%) and weak bot won 22,901 (48.57%) of points, with a total of 47,148 points for both and an average of 2.3574 points per game; strong bot won 10,215 (51.08%) and weak bot won 9,785 (48.93%) of games.

In Experiment 28, GnuBG 3-ply checker skill, (strong bot), played 20,000 money games against GnuBG 1-ply checker skill (weak bot).

Strong bot won 13,100 (50.82%) and weak bot won 12,679 (49.18%) of the points, with a total of 25,779 points and an average of 1.2890 points per game; strong bot won 10,137 (50.69%) and weak bot won 9,863 (49.32%) of games.

Thus, weak bot won 49.18% - 48.57% = 0.61% less points and 49.32% - 48.93% = 0.39% fewer games in the cubeful session than in the cubeless session but the difference is so small that I consider the weak bot still performed better than expected in the cubeful session.

Nobody had made any predictions about the outcome beforehand but you all feel free to make any after the fact comments and interpretations.

For more details, visit my page at https://montanaonline.net/backgammon/py3.php

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.