I've been studying, practicing and playing and my game has improved a lot over the past year or so. My cube play still lags behind though, XG rates it as intermediate.

Does anyone have a recommendation on how to study and improve cube action?

Books? Practice methods? Reference position study?

Working with XG has improved my cube handling from casual player to intermediate level but I want it to get better. Any advice is much appreciated.

I think, there is no clear cut rule. First, and that is like in every game, or profession, work out and execute in excellence the basics. Mastering checkerplay is basic and can not overestimated. From knowing how to play a move, you will develop a sense and feel, what a position, you will reach with your play, is worth. Get a set of simple positions, a race or holding games for example, and work out, what it means to double. Use Danny Kleinmans picture of comparing the struggle for accumulate equity in backgammon with the struggle of gaining yards in football. Break the muddy water down in pieces, that can be recognized (despite the bots will deliver sometimes other results, Advanced Backgammon or 501 will help you to set up your mind). Study Woolsey's law, study Woolsey's five point match. Be aware, that cash game is completely different from late stage match. Study volatility. Study, what is better, to double in or to double out. Work out rules of thumb for double and take decisions. Study, study, study. Despite you're not a math nerd or a new Stick or MCG it will take 10 years or more, depending on your work, you will devote to this.

I started studying Kit Woosley's Backgammon Encyclopedia recently and it has really improved my cube action for money. If you don't have a copy I highly recommend picking one up. It breaks down groups of reference positions in a very well structured way. I ordered it through gammon press.

Hi,

I dont know how much you already know, but this is how i do it:

In anchor situations i use the PRAT rule, Position, race, threat. Two plusses out of three is a double. For the take i use Roberties criteria: anchor, five prime, contact on the other side. If yes, no, yes: it is probably a take.

In no anchor situations I use the a blitz standard benchmark position, with a 3 point board against a 1 point board, with one checker on the bar and one blot in the board, and 2 attackers on the blot, this is double take, more is double pass, less is no double.

From there on i made up a few rules for myself like:
3 extra attackers is almost an extra boardpoint
4 extra checkers in the zone (in the second quadrant but not attacking the blot) is almost an extra boardpoint
a 4-anchor is 1,4 boardpoint
a 5-anchor is 1,5 boardpoint
both backcheckers escaped is 2 boardpoints
this way i can calculate a lot of positions over the board towards my benchmark.

When there is no checker on the bar, but one in the board, and one blot somewhere else then a direct shot on this blot calculates as a barblot, but with one exception: these threat doubles are never Double pass.

Thats how far i came so far with studying a lot of 5-move games against gnu.

greetings k.

Another good method is to post positions on here where you made some sort of cube error and get some feedback on how other players would have looked at the same position. Understanding their thought processes will be very helpful in organizing your own thinking.

Here's an important tip: when learning about the cube, practice by playing money games (NOT tournament matches) against XG. You first want to learn how to handle the cube in standard money situations, and only later blend in match score considerations. Trying to learn by playing 5-point matches will be very confusing, since you'll be mixing the underlying merits of the positions with the adjustments you have to make at a given score. Cash play comes first, match play only later.

Hi Bill,

I dont know if you refer to my reaction, but I did not play 5 point matches but 5 move games,

greetings k.

I wasn't referring to your comment. A lot of beginners practice by playing matches against the bots, but then the cube action gets confusing because it's based on both the match score and the merits of the position. I always recommend starting with a firm basis in money games, and only moving to match play later.

One of the techniques that is helping my cube play is to check the cube % in Gnub cash games quite often, and watch how the % changes and by what margin as the game develops and defines itself. Generally I find this helps me get a good idea of how large the margins are and how they change, and also helps with the over all feel for quite when you should be making specific choices. It certainly very quickly removed many areas where I was making serious errors, and is helping me a lot in more marginal spots.