Quote:
Originally Posted by Skinny Bunner
I have been using some backgammon training software to try to improve. Playing single games against the expert computer, it will normally prompt me to offer the cube when I'm about +.60 or so in equity, and take the opponent's cube if I'm no worse than -.80 or so provided my position is still workable.
Coming from a poker background, it makes more sense to me to try to just "get the money in ahead" and take much smaller edges than that, and drop when I'm behind.
If you only had 49% equity, would you fold to a small bet? Hopefully not. You've got more to consider than just your equity. You have to think about the size of the bet and the size of the pot. Folding when you have pot odds to call and chase is a mistake. So... that, but backgammon words.
Quote:
So why wait til you have such a significant lead in the match, and conversely, take when you're so far behind in equity? Is it because a single roll can swing the equity so much, and the decision is more position-based rather than pure equity? Or is it because of the value inherent in the fact that you still might win the game? In the process of typing this I've just realized - that's probably it. Haha.
It has nothing to do with equity swings, just equity. You might want to read up on the doubling window. It basically tells you when it's mathematically correct to double, and then take/drop, depending on the % of wins/losses/gammon wins/gammon losses that there are (and also match score). I'm sure someone else here can point you to a good article.
Quote:
Also, why would one perfect computer ever take the cube offered by another perfect computer?
Absolutely, yes.
-- Edit: Ummmm... it helps if I don't skip words when reading. WHY a computer would do it is because dropping has a lower value than taking.
Edit 2: If you drop a double, you always lose 1 point. but if you take, you might lose only .9 points (on average). So why give up on that tenth of a point?
Last edited by Aaron W.; 09-09-2014 at 02:07 AM.