the dorbel daily

Wednesday, 27 April 2011

The Big Quiz part 2.

Well, two plays really stand out here. We can discard 13/7, 13/11. 28 shots (without including the numbers that hit in White's home board) and the opportunity for White to hit twice is a bit too big. 13/7, 8/6 is a little better, but still 28 shots. 8/6, 8/2* is a lot less shots, only 18 but strips the 8pt of its last builder and stacks the 6pt even higher. This brings us down to 13/5 which I played in the match, 24 shots but slots a very valuable point and duplicates fours for White to hit on the other side of the board. It's a good play, but slightly inferior to the dynamic 13/11, 8/2*, 20 return shots for White but crucially taking away half of White's next roll, limiting her choices and giving her some awkward numbers. This hitting to keep your opponent off balance is a key weapon in the opening, even when it involves hitting deep in the board. I don't do it often enough.
So, two points for 13/11, 8/2* and one point only for 13/5. Nothing for anything else, as all the other choices are blunders.
Then the play went like this. Blue 6-2, 13/5
White 5-1, 8/2*
Blue 4-2, bar/23*, 5/1*
White 4-1, bar/24*, bar/21
Blue 3-1, bar/22, 23/22
White 6-4, 21/11 (probably correct)
Blue 6-6,........................

There is an interesting thread on fibsboard about doublets and the difficulty in playing them. This is obviously a great roll for Blue, how should he play it? By all means post your answer as a comment, no need to be shy!
I'll tell you the answer tomorrow. Until then, enjoy the game!


RickrInSF said...

22/10(2) looks the best to me.

kamikaze said...

Hi Dorbel, i'm a bit of a n00b but I'd probably go for 22/16(2) 13/7(2). Hopefully better than some weird hitting play but probably wrong as it seems a bit too obvious :)

I still think it's ok though because it

a) threatens whites blot on the 14 point with a direct 2 shot, coercing him to run or cover it next roll
b) advances the back checkers out in a safe pairing to the opponents outer board
c1) makes the bar point
c2) creates a block of 3 consecutive points, the beginning of an effective prime, 3-1 priming, 6-4 hitting and more are potentially good for us next roll
d) creates good communication / connectivity of checkers
e) outfield is controlled - if whites blots on the 1 and 2 points try to run they will most likely get hit and closed out (restricting his options).

All of these plus points and probably still not the optimal move :)

ah_clem said...

Escaping the runners with 22/10(2) is tempting - it puts blue way ahead in the race with both checkers escaped. The problem is, it leaves a bunch of candlesticks, doesn't make any homeboard points, and doesn't unstack any of the four towers. Nope.

22/16(2) 13/7(2) is a nice balanced play that makes a nice blocking point, unstacks the 13, and keeps the checkers in communication. This is probably not too far wrong (or correct) but it still doesn't address the fact that blue is outboarded. Plus it abandons the anchor, which might be a useful fallback if things go south.

13/1*(2) is often correct, but I don't think it is here. It doesn't address the stacks on the 8 and 6, and it burries a pair of checkers.

13/7(2) 8/2*(2) is far better than 13/1*(2) - the two point is better than the one, and owning the bar point with him on the roof has to be pretty good.

Finally, we should look at 13/1* 8/2*(2) - anytime you can send to to the roof it's worth serious consideration. Hitting twice takes away the entirety of your opponent's roll, so it's almost like getting a roll for free. Is it sound here? Blue has an advanced anchor, so it can't be too risky. But I don't like playing three checkers to low points this early - maybe this move is best, but I'm going to chcken out - we're ahead in the race and he has the better board so I'm going to eschew the double hit.

13/7(2) 8/2*(2) for me. It puts a man on the roof, unstacks the 8 and 13 points, makes a homeboard point, and makes a nice blocking point. That's a lot for one move.

kamikaze said...

With a bit of a read of the gnubg faq i think i just completed my first rollout analysis of a position for the top 4 moves w00t! i won't divulge the outcome as i think dorbel will post the best move soon. let's just say i was close but no cigar favouring slightly more chances of a gammon if i've read the data right - not much in it though :)

here's the gnubg id for the gnu'ers who want to quickly set up the board - it should work according to the faq.

Position ID: mGfkASjg8+ABDA