|
BGonline.org Forums
Real Backgammon
Posted By: Timothy Chow In Response To: Real Backgammon (Chase)
Date: Monday, 26 April 2010, at 4:48 a.m.
Chase wrote:
Finally openings are one of the few areas of the game that can be definitively understood, as each generation of bots comes closer to "solving" them.
I'm sorry, but I don't believe this for a second. Certain endgame positions can be definitively understood. Certain other decisions are so "obvious" that we can be confident we know the right answer even without a bot, even if we can't mathematically prove anything about them. Beyond that, I think we are very far from definitively understanding anything about backgammon. For most positions it seems likely to me that equities, strictly speaking, don't even exist. Even if you ignore that, we are far from solving backgammon opening positions. For example, since the equities of opening positions depend on the equities of all possible backgame positions that might arise from them through "optimal" play, we can't possibly have a definitive evaluation of the openings until we have definitive evaluations of all those backgames.
Can we acquire valuable, useful, stable information about the openings through bot rollouts? Yes. But definitive? Not even close.
It's the same in most other strategy games. Despite all those books on chess opening theory, we have no definitive knowledge of any openings. Is the opening position a draw with best play on both sides? Almost certainly yes, but it's hopeless to expect to be able to prove this. On the other hand, we do have definitive information about endgames with up to six pieces.
If you really enjoy mastering something definitive about backgammon, then I think studying the endgame is the way to go. Endgame positions come up almost as regularly as opening positions and misplays are often more costly since the cube value tends to be higher by the time the endgame rolls around.
|
BGonline.org Forums is maintained by Stick with WebBBS 5.12.