Replies to 31P

Following up my earlier post on opening replies.  I want to take an isolated look at replies to each opening roll. 

I start with 31P.  All replies are from the XG opening book, available at this link (http://www.extremegammon.com/ob/) 

Replies are presented in an experimental format.  I list the best reply, and also alternative plays.  Which alternate plays I show is subjective - sometimes I list them and somtimes I don't.  I list them usually if I think the plays are sufficiently close in equity, or if the alternate play is a mistake I could see myself making.

Note the color coding.  I made-up definitions for the size of an error as follows:

.005 or smaller I consider anything this small to be tied
.005 to .02  Small error
.02 to .06 Medium error
.06 or greater Blunder

As with the match equity table, I performed the original work in Excel.  Everything is formatted and sized great within Excel.  Not sure how this look in a blog post.  Hopefully everything translates OK.  Feedback welcome.

One last note.  As I go through the replies to individual opening rolls, it is quite possible my answers will disagree with what I wrote in the Opening Reply Project.  One reason is age - those replies were generated last year on my previous blog.  Also, in some cases, several plays are essentially tied.  I tended to choose plays which I thought were thematically similar.  Even if not necessarily the technically best play, I figured I would benefit from having a similar strategy to tackle related positions.  For example, I like to bring a checker down from the midpoint whenever I hit loose in the inner board, even if hit & split is sometimes the better play.

Replies to 31P

Error Key

Tie

Small

Medium

Blunder

Per XG opening Book

< .005

< .02

< .06

> .06

General Notes

31P best opening roll.  Unless you double or make a point, opponent wins already high 50's & gammons high teens

Opponent has started nice prime.  Split is mandatory and priming is wrong game plan.  All Two Down and Slots are way wrong.

32Z is a bit better than 32S b/c you start a less desirable point for your opponent (ie, not the next point of his prime)

63S is .015 better than 63R while 64R is .015 better than 64S.  The 1 pip for racing appears to be the difference

64 Surprises.  64S and 64P wrong but that that bad.

22 and 33 Surprises.  22N wrong conceptually but not too bad.  Same with 33 plays 33@ and 33E (wrong but that terrible)

11e and and 32Z are most logical and correct, but the alternate plays 32S and 11N are essentally tied

Reply Roll

BEST

2nd

3rd

4th

Point Making Rolls

31

P

8/5, 6/5

42

P

8/4, 6/4

53

P

8/3, 6/3

61

P

13/7, 8/7

Split or Slot

21

S

24/23, 1311

$

13/11, 6/5

0.0572

41

S

24/23, 13/9

51

S

24/23, 13/8

Split or Two Down

54

D

13/9, 13/8

52

S

24/22, 13/8

43

Z

24/21, 13/9

D

13/10, 13/9

0.0224

prime is wrong concept b/c opp has better prime already

32

Z

24/22, 13/10

S

24/21, 13/11

0.0073

split mandatory, but better to go for point he wants to make less

6-X Rolls

65

R

24/13

64

R

24/14

S

24/18, 13/9

0.0142

P

8/2, 6/2

0.0189

Run is .015 better than Split

63

S

24/18, 13/10

R

24/15

0.0152

Split is .015 better than Run

62

S

24/18, 13/11

R

24/16

0.0415

Doubles

11

e

24/22, 6/5(2)

N

8/7(2), 6/5(2)

0.0029

22

E

24/22(2), 6/4(2)

N

13/11(2), 6/4(2)

0.0132

33

B

24/21(2), 13/10(2)

@

24/18(2)

0.0175

E

24/21(2), 6/3(2)

0.0196

44

P

13/5(2)

55

P

13/3(2)

66

B

24/18(2), 13/7(2)