the dorbel daily

Monday, 28 November 2011

The Sharp End

These are some problems I encountered recently, all coming at the sharp end of the game where there is still a lot of contact and truck loads of equity are at stake on every roll. The first two are from a match I played. Black is on roll, trailing 6-7 to 9. What's your play?

Position One

The second best play is a double blunder, so be careful.

Position Two

And this is the next roll, another opportunity to make a double blunder. What's your play?

The very next day I had a match to analyse for a student and these two positions reminded me of the two we have just seen.

Position 3

It's double match point, so what is Black's play. You can't make a big mistake here, but it's good to get it right and know what you are trying to do, because it's dmp and the wrong play can cost you two wins in a hundred.

Position 4

Same game, next turn. Again, you can't make a big mistake but it is dmp, so even a small mistake can lose you one game in a hundred. What and why please.

Lastly, the bastard child of Positions 1 and 2. I got down to this, trailing 3-away, 2-away. Black on roll. Is it a cube? If it is, is it a take?

Position 5

Fun for all the family! If you get all these right, my hat is off to you. If you know why you are doing it, then you are either Stick Rice or God, certainly a better man than me in either case!

Until we come back to these, enjoy the game!


poppp said...

boop says

1) oh dear … 8/6 8/3 is screaming at me but

keeping 3 points in the outfield stops black from breaking his anchor with 61 62 or 66 (gaining me pips while I'm way behind) and with any other 6 white must run with 1 counter leaving a hit with 1s 2s and 6s and I'm looking good to double. Any non 6 combination or 11 22 and white moves further into wastage and the above applies again next move.

9/2 leaves 1s, 2s & 6s to hit if white rolls a 6 and 3s, 4s and combinations to cover the 5 point. looks good enough for double and pass

2) still behind and following a similar idea i move 9/3 6/2

3) (did I not learn that at DMP you should double immediately as it can only be a very minor mistake?) anyway .. I'm 20 points behind after this move so i'm moving the blots in my home (likewise even if i was ahead in the race) and i need to close the high numbers if possible so i don't move the 5. 4/1 and 31 21 leaves 2 open blots. 6/5 4/2 only leaves 1 but I'm not really sure if I shouldn't safety the 5!!

but just for fun and cos it looks a bit braver 6/5 4/2!

4) i can't think of anything else other than 14/8 14/10 . . . it minimises hits down to 11 and keeps maximum builders for the 5 point. The only other thing that I can see is begging white to hit me with 13/7 14/10 which seems mad to me or 14/4 which has less builders . . . i must be delirious after all this concentration - 14/8 14/10

5) 20 behind. ive not really managed to grasp match equities yet but if i double and lose it's over and if i don't double and lose I'm down 3 away 1 away crawford which means i must win the next 2 or a backgammon :-( -- but if i double and win it's a total turn around with me up at crawford. DOUBLE DOUBLE DOUBLE!!

there are a lot of good numbers

54 52 51 42 41 21 11 22 44 = 15 hit and close the 2 point
5 7 or 9 hits and leaves a blot but white is not favourite to return hit.
so more than half the numbers are in my favour plus i must be braver i think due to the match equity thingy so it is a double.

and with whites 6 and 4 points open I think it may even be a pass.

phew! good fun, good night :-)

Steve said...

1. Black's trailing by a mile so shouldn't run, which argues for 9-4. Then 4-2 maximises flexibility & opportunity to improve the home board and, if WHite hits she probably leaves us with returns.
9-4, 4-2.

2. Black's still trailing by a street, which again argues for not running, so 9-3 has to be one of them. My instinct (which usually gets me into trouble) would then be to follow the theme with 6-2.
9-3, 6-2.

3. Black needs to preserve a 4-pt board, 6-5, 4-2 does that without leaving blots at home.
6-5, 4-2.

4. Minimise shots. 14-8, 14-10.

5. I count 15 rolls for Black that hit and cover and no non-hitting ones that leave direct shots, so it looks like a double. White has 3 crossovers to make before she can bear off but is in front and has a 5-pt board, so I guess can take. Double/Take.

ah_clem said...

1) There's a rule of thumb to never volunteer a shot in a holding game. I'm sure that there are exceptions, but I'm not seeing how this is one of them. I'll clear the 8 point here - this moves two checkers into the homeboard which is good for the race and makes 61 play poorly for white. Now all 6's except 62 play poorly and after next roll 5s should play poorly too.

2) Making the 5 is certainly tempting, but again I don't see the utility of volunteering a shot. White wants nothing more than to race by and if she can do so with tempo even better. I'll just play the awkward looking 9/3 6/2.

3) Ok. I'm still not volunteering a shot. White's board is even stronger here. I'll go for 6/5 6/4 but I'm not at all confident that it's better than other plays that shuffle checkers in the homeboard.

4) I'll clear the 14 and minimize shots. 14/8 14/10. This is also best for the race.

ah_clem said...

5) I think White has to take - there are a lot of escaping numbers even if white gets hit and white has a nice racing lead.

I count 15 shakes that hit and cover - and I think any sequence that involves a dance is market loser. Damn near everything at least allows a loose hit (63 66 33 don't) and black will hit loose. Again, a dance seems lite a cash. So, given the huge number of market losers and the looming Crawford I'd double here. D/T.

Timothy Chow said...

1. As a general rule, I'd recommend avoiding comments such as "The second best play is a double blunder" because that is often an enormous hint. It means that if there are two candidate plays that are obviously very similar to each other then neither one can possibly be correct. Anyway, we're behind in the race here so our hope is to squeeze White off her anchor. I'd play 8/6 8/3. The alternative is 9/2 but that seems to leave our board in poor shape to contain a White checker even if we hit a shot.

2. The main question of course is whether to play 9/5 9/3. I think I would. If we play safe with 9/3 6/2 then that cuts deeply into our winning chances. Even if we hit a shot we probably won't be able to contain it. At this score, it's possible that if White is forced to leave a shot after 9/5 9/3, the cube action may be D/P. Not sure.

3. 4/1 looks natural. I'll be interested to learn why it's wrong if it's wrong.

4. 13/4 keeps the back point and maximizes contact, but leaves more shots and moves another checker past our open 5pt, versus 14/10 14/8. Not sure at all about this one but I'll try 13/4.

5. White is not likely to get gammoned here. If she drops she'll have 50% MWC; if she takes and wins she'll have 100%, and if she takes and loses (assuming no gammon) she'll have 32%. So her take point is 18/68 or slightly more than 25%. Black's raw pip count is 70 and White's raw pip count is 53; however, White obviously has more wastage. If it turns into a race then I think it will be fairly even. Black points on White with 15 rolls and misses entirely with 66 63 33. With the remaining rolls it's an interesting question whether Black should hit loose or play safe. I think Black should probably hit loose, so let's ballpark White's chances by pretending that White loses if pointed on and wins 1/3 of the other games. That gives White 7 wins out of 36. D/P is my assessment.