Hi Barb, Don't get discouraged! Yours is a common concern among those who are new to high-quality backgammon software such as GnuBg. It plays better than any human on earth, including the likes of Mochy, Falafel, and the rest of the world's top players. When it beats you, it is not because it cheats. It just plays better than you!
Russ Allerby came up with the easiest way to prove GnuBg is not cheating. Simply activate the GnuBg Tutor, and take it's advice on every move. When you activate the tutor, it will pop up to interrupt you whenever it thinks you have made a significant misplay. The GnuBg tutor won't interrupt you when you make a small mistake, so you will still not be playing as well as GnuBg does. That means you will not win exactly half the time, but you should win your fair share of games, I think well over 40%. You can turn on the tutor using these commands: Settings > Options > Tutor > Tutor Mode It sounds like you are new to GnuBg, so let me also suggest that you select World Class play in the following three places. To change these settings, you have to click the triangle for the drop-down list that contains "World Class." Settings > Analysis... > Analysis Level Settings > Players... > Player 0 > GNU Backgammon Settings > Rollouts... > Play settings > First Play Both Another setting new uses should check is Eval Threads. This lets you turn on multithreading. If your computer has an Intel duo-core chip, for instance, you will have two cores, each of which can run two threads. The total number of threads in this example is 4. If you have a quad-core chip, then the total number of threads is 8. Once you figure out how many threads your computer can run, you should enter that number here: Settings > Other > Eval Threads This is an important setting because the default is 1. When you increase it, GnuBg will run much faster for many tasks. Hope this helps. Regards, Mike Mannon P.S. For several decades now, neural-net-based backgammon programs such as GnuBg have been beating the pants off human opponents. That has generated a whole a slew of assertions that the programs must be cheating, but, of course, they are not. Most developers have gotten tired of the question, so they are sometimes impatient when it arises. Yours was a reasonable inquiry from a new user, so it has been well-received here. In the past, however, we have had some trolls who won't drop the issue. You will find that they get short shrift on this list. _______________________________________________ Bug-gnubg mailing list Bug-gnubg@gnu.org https://lists.gnu.org/mailman/listinfo/bug-gnubg