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

BGonline.org Forums

Cubeless bakgammon

Posted By: Timothy Chow
Date: Wednesday, 21 April 2010, at 7:39 p.m.

In Response To: Cubeless bakgammon (Gregg Cattanach)

Gregg Cattanach wrote:

Cubeful backgammon adds a skill that doesn't exist in cubeless backgammon: namely the need to evaluate how often (a quantitative analysis) the game is won/lost and gammons won/lost from the current position. In cubeless backgammon, the only skill called for is to make the checker play that has the best equity (a qualititative analysis: which play is best, you don't really care by how much).

I hear this kind of thing asserted frequently but I remain unconvinced.

If you want to play checkers at the highest skill level, then you need to compare how many extra games/gammons Play A will win/lose compared to Play B, and trade them off against each other quantitatively to decide which one is the better play.

Conversely, to make a cube decision, you just care whether your double or take is correct and you don't care by how much it is right.

Perhaps people's judgment in this issue has been clouded by overexposure to bots. Bots often get checker-play decisions right even if their absolute equity estimates are wrong. On the other hand, today's bots typically use absolute equity estimates in an essential way to make cube decisions, and if their absolute equity estimates are off then their cube decisions will be off. But how bots do things isn't necessarily an indication of how the decisions must be made, or of how humans make them.

Or, perhaps the confusion comes from the fact that for most checker plays you can get by with intuition and no equity calculations, whereas it's easy to come up with cube decisions where someone who doesn't have a feel for quantitative analysis will be totally at sea.

Intrinsically, though, accurate checker play is no less quantitative than accurate cube handling. Cube skill is certainly distinct from checker-play skill, but it is not correct to say that it's because checker play does not require quantitative estimates of wins/losses/gammons.

For a concrete example, how about Stick's wild checker play? Figuring out the right checker play here at gammon save in a cubeless checker match surely requires quantitative analysis.

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.