| |
BGonline.org Forums
Non-doublet Replies: Rules for beginners
Posted By: Axel Reichert In Response To: Non-doublet Replies: Rules for beginners (nack ballard)
Date: Friday, 28 April 2017, at 8:21 p.m.
Hi Nack,
Many thanks for your detailed explanation of how the data I assumed to
be rollouts is really a can of worms ... I think I will be pragmatic
for now and just take the data as given. The target audience could not
care less about rollouts, variance reduction and standard deviations.
Many thanks also for your detailed explanations of the intricacies of
Nactation. I was not aware that D for a partial move could be used for
8/7*. Notation is the source of all misunderstanding ...
Which reminds me that once the Reply Rules are Rough and Ready for
Rookies ("The Perfect Wagnerite"), they definitely should be
reformulated in terms the target audience can understand. For
beginners, Nactation may look like noise on the line. (-:
My 2nd-die-after-hit convention was simply a Nactation laymen's
shortening of Jeremy's article. Of course, the inventor of Nactation
has interpreted it in the precise meaning defined by himself, whereas
I have interpreted it in the precise meaning defined by Jeremy. My
understanding was: "I cannot play the 1 down (from mid), so I have to
split 24/23". Shame on me. (-: So what I meant to nactate with rule 3
was:
After hits from mid: Play from rear, else from mid.
After hits from non-mid: Play from mid, else from rear.
As a software engineer, I did not like "else" in my original
formulation, since I wanted to avoid ambiguities to which "if" the
"else" relates. I hope this formulation is clear now. And it elegantly
avoids the "virtual" blunders for 32S-21, 43Z-21, 32S-41, and 43Z-41,
which were only caused by misunderstanding. (-:
But you are also right in that for 7* (be it from 13 or from 8),
playing the second die from 24 is usually better than from 13 (summed
over all cases, it loses 0.28 equity less), so your new suggestion for
the 2nd die of a hitting reply is great and better than the convention
used by Jeremy and better than CRMU. I have reformulated it slightly,
see below.
I also like the piggybacking of $-64R.
Thanks for your polite and quiet remark about the uber-whopper for
32Z-65 and 52S-65 being played 13/7 8/3*. As usual, i found it during
"beta-testing" my rules 30 minutes after the post was out ...
But since I did not want to move 65R (along with its exception
6xS-65K) upward, I explored the hitting rule "Hit > 3" instead of "Hit
> 2". This indeed results in some better and some worse replies:
32Z-21 [S H-0.0043]
32Z-41 [U H-0.0045 Z-0.0106]
32Z-51 [S R-0.0014 X-0.0209]
32Z-32 [K X-0.0040 H-0.0110 S-0.0619]
32Z-52 [S K-0.0016 X-0.0077 H-0.0176]
32Z-43 [X H-0.0106 S-0.0930 Z-0.0990]
32Z-64 [H R-0.0258 S-0.0396 P-0.0546]
52S-21 [S H-0.0176]
52S-41 [D U-0.0062 H-0.0432 S-0.0564]
52S-51 [S R-0.0272 X-0.0582]
52S-32 [H S-0.0079 K-0.0190]
52S-52 [D S-0.0061 H-0.0400 K-0.0438]
52S-43 [H X-0.0028 U-0.0460 S-0.0630 Z-0.0920]
52S-63 [R S-0.0189 H-0.0291]
52S-54 [S R-0.0299 H-0.0434]
Overall, "Hit > 3" loses less equity (0.15) than "Hit > 2" (0.30),
provided an exception "Hit 3 with 3" is added, to avoid blunders >
0.06. I think this is a quite elegant solution and easy to memorize
because of the "numerical alliteration". What do you think?
Regarding "most similar legal alternative": Once a beginner knows the
standard 27 plus the (popular with beginners) 62R, he can be advised
to take the "closest" one from this "library": For the "D or S"
openings, the "R or S" openings, and the "$ or S" openings there is no
alternative move in the "library", if the split is blocked. In
contrast, for 42P-32, your rhyme recommends D instead of the illegal
S. Z would be "more similar" and 0.0266 better. More importantly, what
does your rhyme say about 64P-21, 64P-41, and 64P-51? You probably
need to add a "$" at the end.
To summarize, this is my next iteration, now formulated for beginners,
except for the "exceptions" (which I prefer at the end, to have a
"modular" list which can be appended over time for intermediates):Openings/replies (for playing without blunders > 0.06)
Non-doublets
1. Hitting
a) With pointing roll: Hit > makeable point.
b) With non-pointing roll: Hit > 3. Hit 3 with 3.
2. After hitting
a) After hits in the two outer boards: Play from rear, else from mid.
b) After hits in the two inner boards: Play from mid, else from rear.
3. Not hitting
a) Make a point, run with 65, split reverse with 43, else split.
b) If move is blocked, play most similar legal alternative.
4. Exceptions (omit all for playing without "bad" moves > 0.08)
a) D-x1$ ("Slot after Down")
b) D-64R ("Run past Down")
c) $-64R ("Run after hit Slot")
d) 6xS-65K ("Kill after 6-Split")
I tested this wording already in our club with good success (and
print-outs for the rookies).
By now I think we have pretty much converged on the non-doublets, so
it's high time for the doublets:
Doublets
11. 8/7(2) 6/5(2)
a) Non-6x split: 24/22 6/5(2)
b) 6/4*(2)
22. 13/11(2) 6/4(2)
a) Hit 20 along with 6/4(2)
33. 24/21(2) 13/10(2)
a) Hit 5, 3: 8/5(2) 6/3(2)
b) 13/7*(2)
44. 24/20(2) 13/9(2)
a) Hit > 3, hit again > 1, cover < 6, play on other side
55. 13/3(2)
a) Hit 3, 1 with 8/3(2) 6/1(2)
66. 24/18(2) 13/7(2)
With this, we get equity losses > 0.04 only for:
32D-33: -0.0532
63R-33: -0.0508
64S-22: -0.0450
I tried hard to come up with an easy rule for 44, but it is not
easy. Any brillant insights?
Best regards
Axel
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.