I'm trying to bring a better understanding of how these influence my doubling decisions. For now I have to use XG's helpful dice distribution chart, I find that it takes me ages to work out which dice are good and which dice are bad. The obvious ones are, well, obvious, but spotting how many are poor is another story altogether.

Anyway, on the chart, it shows you the percentage of market losers that you can throw. What sort of a percentage am I looking at as being a tipping point in my decision to double? Obviously the same applies to anti-jokers, or do you discount those for some reason that I haven't worked out?

Any help appreciated.

There's a principle called O'Hagan's Law that's very useful in early game and middle game situations:

Double if you have at least 9 market losers, and your other numbers are all at least OK.

Hum.... That sounds like "too good to double" or not?

I don't understand your point here. Do you think that 9 market losers makes you too good to double? Or do you mean something else?

Mm, 9 market losers means 9 rolls out of 36 (25%) that crush my opponent (that change the position from D/T to D/P) + the rest "OK" i cannot fear anything, and unless its a cash game i play on for the gammon

Stick put it this way (BGO Msg. No. 160835):

The basic idea of O'Hagan's Law is that you usually have an initial double:

1. If you have 9 market losers
2. Are doing alright on the other 27
3. If you're within 6% of the opponent's true take point.

In a non-volatile position, you can wait to double until you reach the high end of the doubling window. When a position is volatile, however, and there are many market losers, you may want to double sooner. O'Hagan's Law gives you a rough quantitative guide to help decide just how much sooner.

BTW, O'Hagan's Law is more a guideline than a law. Depending on the position, for instance, you can lay off your bad rolls against some of your good ones. If you have 12 big market losers, but also have 3 rolls that swing big the other way, you can subtract the two to reach the 9 needed for a double.

Caveat emptor! This is a guideline, not a rigid rule.

Mike

Caveat emptor indeed! Should be printed on every backgammon board the world over!

see my brief writeup on the doubling process here


http://bigmoneybill.blogspot.com/201...g-process.html

Hey, that's really handy. I've printed it out and am going to glue it next to my screen.

Thanks.

The point of O'Hagan's law is that it is 9 NET Market losing SEQUENCES not rolls. A sequence is your roll and your opponent's roll.

I also matters how much you lose your market by, hence highly volatile positions can be doubles without 9 sequences due to the high volatility.

Also your typical holding game may have only 4 market losers (11.1% or rolls) but you have those 4 market losers roll, after roll after roll so that is why you can double holding games. To lose your market you need to have 4 or less checkers on your (typically) 13 point, as with 5 you cannot lose your market. Sometimes called the rule of 4. Note that having only 2 checkers on your 13 point is much better as 22 plays a lot better - you get those 2 checkers to the 9 point instead of 4 checkers to 11 point.

Doing some research and reading a few articles on 'Losing Your Market'

1. There seems to be some contradiction on if you are considering your next roll or
the sequence of your next roll / opponants next roll. My understanding the tipping point is 9 net, that is say for example 12 market losers less 3 bad rolls gets you to the 9.

2. How do you count this over the board ? I have tried to go through my possible rolls against a computer (doubles, then other combinations counting down from 65),
though finding it hard in live play. If you are conting sequences then even harder.

Comments on both 1 and 2 appreciated.

I thought I would resurrect this discussion:

What is the easiest way to count Market Losers over the board ?

Why don't you publish 2-3 positions and we can talk about the thought process. Doubling is more a matter of learning how to think about positions than trying to count market losers.