News:

Try the New Recent Posts Link available above Recent Topics on the Front Page

Main Menu

when do you finally split your anchor?

Started by boop, October 07, 2011, 05:04:41 PM

Previous topic - Next topic

boop

edit : Dear reader, I apologise for deleting the image for this post - it's gone forever I'm afraid  :(


I've been meaning to ask a question about this for ages because I'm often chastised by gnu for not splitting or by splitting when I shouldn't.

Here is an example that I've created ... it's probably not the best example but then I can't design a better one because I don't understand it yet!


i'm thinking here, if I split the bot must use a 4 if thrown. 4/1 5/1 6/1 would be fantastic. but at the very least, apart from if he throws 44, i get a hit.(if he throws a 4 ;-)  )
Actually I think I can work out that there could be more hits if I wait ... but it hurts my mind and in some positions I just can't work it out.

so I suppose I'm just after tips for this kind of position where you're wondering whether to split.

boop


dorbel

You need to be clear about what you are trying to achieve here. Assuming it isn't irrelevant because of the score, you are playing to save the gammon. You win about 1% from here, if you play perfectly and that isn't going to be easy, so getting off the gammon is all that you are trying to do. To this end, getting and hitting a shot is a bit of a double edged sword, as if you hit one it often runs around and hoovers up the checkers that you have spread out around the board. I would stay rooted to your anchor for this turn and just play 8 useful pips in the outfield. You're not wasting any pips in doing that and you get 10 shot numbers next turn, as well as freezing one of her best numbers, 4-4. After 6-4 and 5-4, you get 13 numbers to hit. Moreover, by staying, your anchor is still a problem for White.
If you leave the anchor, you get 14 shot numbers, but all of these will be the minimum 11 shots and usually the last that you will get. 4-4 is activated as a golden number instead of a disaster. You can now get pointed on too and you don't want to be on the bar for even one turn.
I would play 17/11, 14/12, two crossovers and give yourself a chance to make a 4 point board. Staying with both is usually right as long as you are not wasting any pips and as long as you aren't in danger of a bg. Leaving with one to force her to play the frozen number is almost always the wrong idea.

boop

#2
brilliant! good info and I love the one liners beginning with ..."it's normally good to ... " then you can try to see if it definitely is good over the board at any one time. They're like little mind anchors - somewhere to think from.

i was going to design some other positions where my home board is much stronger, to see if the possibility of a win could modify your "don't run rule" or whether you just wait until your home would be destroyed by any other move than splitting the anchor. But I think this info will allow me to observe while playing and work through my confusion.

Also I do have this example that we discussed on fibs where the bot kept his anchor untill this position hoping that I would throw a 1. (i did and he hit me twice and then beat me ... what was the french name you gave it?)
I just need to keep an eye out for when the bot bolts strangely early and see if I can work out why.


If I find one - I'll be back!

thanks Dorbel

boop

dorbel

It's called the coup classique. Note how the bot has two blots inboard in the hope of recycling to pick up the second checker. If she can close out both checkers she'll be a good favourite, typically with a winning percentage in the mid sixties.