Open Side Menu Go to the Top
Register
Problems with grabbing opponent's 5 too early Problems with grabbing opponent's 5 too early

09-02-2012 , 06:38 AM
I used to believe that grabbing the opponents 5 given the opportunity is standard. It gives your pieces a good anchor and attacks your opponents outer board. However, I see that I am getting stuck and not knowing where to go from there. Often it turns into a high anchor game and I feel like that is not so ideal. Over the last 500 games or so I am running into a lot of problems rushing to snag the 5 over using those rolls to play offensively.

Also, leaving the guys back for a while makes it more difficult for my opponent to build up his inner board. Granted I run the risk of being primed in.

I am assuming that I am playing it wrong after I grab the 5. I often get it on the 2nd 3rd or 4th role and I often split one of the back checkers to the opponent five instead of moving it to my 9. Am I going for it too early? And what is the right way to play after grabbing the opponents 5.
Problems with grabbing opponent's 5 too early Quote
09-02-2012 , 08:09 AM
So, what are you thinking when you encounter positions like these two?

White - Pips 164

Black - Pips 162
41S-21S-43 Black to Play 4-3

White - Pips 158

Black - Pips 162
41S-63S-43 Black to Play 4-3
Problems with grabbing opponent's 5 too early Quote
09-02-2012 , 08:48 AM
So in the first position I play 13/9 13/10. Because he split his back checkers I want to cover my 9 even more. Also, he only has one extra potential builder in his outer board, so there is not much to worry yet. However if his back checkers were not split and my 9 checker was on the 8 instead I would grab his 5 over 13/9 13/10.

Second position: I am facing a direct hit on the 9 but there is really not much to worry about if he hits me with a 2 or a 7. Grabbing the 5 gives me a solid anchor for the rest of the game. If he doesnt hit I have a chance to hit or cover the 9 5 or 7.

He has a checker on the 15 as opposed to the 14 in problem one. This makes a difference as it is more likely to be used to fill up whites inner board.

I could also consider 13/9 13/10. His inner board is not built up and I do not need to panic right now. If he doesnt hit I have a good chance at grabbing my bar (thats what she said).
Honestly, without looking at this for as long as I did my gut reaction would have been 13/9 23/20, but after looking at it for longer I think that 13/9 13/10 is the better play.


If white had my 18 or 16 in these positions I would most likely grab his 5 point. Double 5s or double 6s is usually enough to seal the game. 7s are the most commonly rolled number giving me a chance to run one of the checkers out when I feel the other one left behind is in low danger.
Problems with grabbing opponent's 5 too early Quote
09-02-2012 , 01:10 PM
Just by curiosity, treeunit, did you at least consider 9/5 8/5 (making the offensive 5-pt) in both scenarios?
Problems with grabbing opponent's 5 too early Quote
09-02-2012 , 02:20 PM
Quote:
Originally Posted by treeunit
So in the first position I play 13/9 13/10. Because he split his back checkers I want to cover my 9 even more. Also, he only has one extra potential builder in his outer board, so there is not much to worry yet. However if his back checkers were not split and my 9 checker was on the 8 instead I would grab his 5 over 13/9 13/10.

Second position: I am facing a direct hit on the 9 but there is really not much to worry about if he hits me with a 2 or a 7. Grabbing the 5 gives me a solid anchor for the rest of the game. If he doesnt hit I have a chance to hit or cover the 9 5 or 7.

He has a checker on the 15 as opposed to the 14 in problem one. This makes a difference as it is more likely to be used to fill up whites inner board.

I could also consider 13/9 13/10. His inner board is not built up and I do not need to panic right now. If he doesnt hit I have a good chance at grabbing my bar (thats what she said).
Honestly, without looking at this for as long as I did my gut reaction would have been 13/9 23/20, but after looking at it for longer I think that 13/9 13/10 is the better play.


If white had my 18 or 16 in these positions I would most likely grab his 5 point. Double 5s or double 6s is usually enough to seal the game. 7s are the most commonly rolled number giving me a chance to run one of the checkers out when I feel the other one left behind is in low danger.
the vast majority of the time in the early game it is correct to make your five point given the chance. in both of the problems posted by mike, making your own five point seems clear. making the defensive anchor, or the outfield point is the wrong idea. unless the opponent has some major threat on his side of the board, wait to make the advanced anchor in the early game. also, plays like 13/9 13/10 leave a stack on the 6 point and strip the midpoint leaving poor distribution of checkers. making the five gets the checkers where they should be, and maintains some meat on the midpoint.
Problems with grabbing opponent's 5 too early Quote
09-03-2012 , 04:07 AM
I hope you enjoyed the positions I posted. They were selected to let you begin to identify some of the features of a position that might suggest not making the advanced anchor. Be careful, though. If you play the quiz factor (QF), you may think that I have chosen only positions where making the anchor is wrong. Perhaps, that is true, but you should (and I think you are) make the play that you would make over the board.

Here is a hint: The top two plays in both positions involve making one 5pt or the other. With this in mind, take a look again at the two positions above, and give a reason for preferring one 5pt over the other.

Below are two more for you to practice on.

.

White - Pips 161

Black - Pips 162
41S-51S-43 . Black to Play 4-3

.

White - Pips 160

Black - Pips 162
41S-61P-43 . Black to Play 4-3

.
Problems with grabbing opponent's 5 too early Quote
09-03-2012 , 05:59 AM
Position 1: 9/5 8/5
Position 2: 24/20 23/20 or 23/20 13/9.

In position 2 white is already building up a prime and I would either play 23/20 13/9 with the hope of closing the 20 within the next two rolls, or 24/20 23/20 hoping to close my 9 or 7 in the next few rolls. It really depends on what mood I'm in at the time. Position 2 is exactly what I was talking about in my post. I usually grab it instead of making my 5. Late game white seems to magically get passed my offense (weakened probably by me not grabbing the 5 early) and swing his back checkers around while I am usually stuck with mine on the 20 late game. I am not sure what usually happens.

On a side note in the first positions posted at the beginning. I did not even see 9/5 8/5 as options. I am 100% certain over the board I would have made that play for both positions, but I was so hyperfocused on the subject of the post that my mind wouldnt even let me see into my inner board. I am quite shocked by this and really need to pay more attention and not, like you said above, assume the answer must be dealing with the topic at hand.

Last edited by treeunit; 09-03-2012 at 06:06 AM.
Problems with grabbing opponent's 5 too early Quote
09-04-2012 , 12:40 AM
Without revealing just yet what plays are best in the 4 positions discussed in this thread, I wonder if TreeUnit (and any other readers) would help me with some counting. I have two questions for each position:
  1. If you make the anchor, @ = 24/20, 23/20, how many shots will your opponent have at the blot on the 9pt?
  2. If you make your 5pt, how many ways will your opponent have to make his 5pt on the next throw?
You can use Nactation strings to identify the 4 positions:
  • 41S-21S-43
  • 41S-63S-43
  • 41S-51S-43
  • 41S-61P-43
(In Nactation, S means “split,” P means “point,” and @ means “anchor”)
Problems with grabbing opponent's 5 too early Quote
09-04-2012 , 05:23 PM
Quote:
Originally Posted by Taper_Mike
Without revealing just yet what plays are best in the 4 positions discussed in this thread, I wonder if TreeUnit (and any other readers) would help me with some counting. I have two questions for each position:
  1. If you make the anchor, @ = 24/20, 23/20, how many shots will your opponent have at the blot on the 9pt?
  2. If you make your 5pt, how many ways will your opponent have to make his 5pt on the next throw?
You can use Nactation strings to identify the 4 positions:
  • 41S-21S-43
  • 41S-63S-43
  • 41S-51S-43
  • 41S-61P-43
(In Nactation, S means “split,” P means “point,” and @ means “anchor”)
Mike --

I realize that Nactation saves space, but we have a lot of space available -- virtually an infinite amount, I would say.

In the future, how about describing plays in Magriel's notation system, so everyone here will easily know what's what. (i.e., 4-1: 13/9 24/23).
Problems with grabbing opponent's 5 too early Quote
09-04-2012 , 05:44 PM
Quote:
Originally Posted by Robertie
Mike --

I realize that Nactation saves space, but we have a lot of space available -- virtually an infinite amount, I would say.

In the future, how about describing plays in Magriel's notation system, so everyone here will easily know what's what. (i.e., 4-1: 13/9 24/23).
I already did better than that. I gave a diagram for each!

In this case, there has already been some confusion as to which position is being referred to. TreeUnit's Position 1 and Position 2 actually refer to the 3rd and 4th positions. By introducing Nactation, I was trying to provide an unambiguous way to refer to the 4 positions without having to redisplay them (just yet).

I think you already know that Nactation is much more than a way to "save space." It is a way to visualize the moves. For instance, 43@ = 43-anchor, is much easier to comprehend than 24/20, 23/20.

At the end of the day, this is your forum. So, if prefer to keep it a Nactation-free zone, that is your choice. I understand your desire to keep it slanted towards beginners and intermediates, and your concommitant bias against the posting of the very tools that you probably rely upon yourself, i.e, rollouts and the Nactation that is used to summarize them.

Last edited by Taper_Mike; 09-04-2012 at 05:51 PM.
Problems with grabbing opponent's 5 too early Quote
09-04-2012 , 06:11 PM
The thing is, I don't think you can use Nactation throughout a whole game. Correct me if I'm wrong, but can Nactation be used past the early game phase? Magriel's notation is still more standard for most recreational/intermediate players and you can use it throughout a whole game.
Problems with grabbing opponent's 5 too early Quote
09-04-2012 , 07:03 PM
You are right.

Nactation is primarily a tool for recording moves in the early part of a game. That is why it is appropriate for the 3rd-roll positions under discussion in this thread. It is also the best way to name a rollout file for one of these early game positions.

Many players, and not just beginners and intermediates, learn enough to do that, and no more. I know many advanced players who use Nactation only for the first few moves of a game. For that purpose, it is far superior to traditional notation.

Nactation can, however, be used to record an entire game. This is very useful in live settings, when you are trying to write down the moves of a match in realtime. In a post at BgOnline, I explain how to nactate every move in a game that was played at this year’s Osaka Open. It makes for an easy read, primarily because there is a board diagram for every roll.
Problems with grabbing opponent's 5 too early Quote
09-05-2012 , 09:03 AM
Quote:
Originally Posted by Taper_Mike
I already did better than that. I gave a diagram for each!

In this case, there has already been some confusion as to which position is being referred to. TreeUnit's Position 1 and Position 2 actually refer to the 3rd and 4th positions. By introducing Nactation, I was trying to provide an unambiguous way to refer to the 4 positions without having to redisplay them (just yet).

I think you already know that Nactation is much more than a way to "save space." It is a way to visualize the moves. For instance, 43@ = 43-anchor, is much easier to comprehend than 24/20, 23/20.

At the end of the day, this is your forum. So, if prefer to keep it a Nactation-free zone, that is your choice. I understand your desire to keep it slanted towards beginners and intermediates, and your concommitant bias against the posting of the very tools that you probably rely upon yourself, i.e, rollouts and the Nactation that is used to summarize them.
Just for the record, I never use Nactation. I think it's inherently unintuitive, and very off-putting to beginners and intermediates. In my view, Magriel's system is vastly better. We also have an immense body of literature all using the Magriel system, and a new system would have to be dramatically better and clearer to warrant replacing it.

When I'm recording a match, I use a modified Magriel system. Refer to moves by the roll and the endpoints (i.e. 4-1: 13/9 24/23 becomes 41 9 23), use * for a hit, -- for a borne off checker, and a superscript for putting two or more checkers on a point (6/4(2) becomes 4^2). It's a system that's very easy to learn and remains unambiguous throughout the game, whereas Nactation becomes hopelessly unwieldy. (Try Nactating 1-1 in a long race.)
Problems with grabbing opponent's 5 too early Quote
09-08-2012 , 05:52 PM
I like the idea of keeping this site Nactation free. I think describing the solution in terms of nactation is confusing whereas the Magriel system just seems logical and easily grasped by all levels of players.
Problems with grabbing opponent's 5 too early Quote
09-09-2012 , 06:50 AM
I think nactation for the opening few moves is best and after that, magriel
Problems with grabbing opponent's 5 too early Quote
09-09-2012 , 09:25 PM
Quote:
Originally Posted by Taper_Mike

I think you already know that Nactation is much more than a way to "save space." It is a way to visualize the moves. For instance, 43@ = 43-anchor, is much easier to comprehend than 24/20, 23/20.
I strongly agree with Bill here. Using a dozen different methods of describing the moves does not make it easier.
If you want to say something more intutive than 'make the opponents 5pt' would be it. Everyone understands 24/20, 23/20. 43@ is as clear as mud.
Problems with grabbing opponent's 5 too early Quote
09-10-2012 , 04:04 AM
This stuff makes descriptive chess notation look convenient.
Problems with grabbing opponent's 5 too early Quote
10-24-2012 , 08:21 PM
Mike, I always enjoy your posts and find your positions and thought useful and insightful.
I always appreciate rollouts (hopefully XG2) because they are proof and I can see the win and gammon % (you need these to truely understand a position). I am happy with either notation - please just keep on making your fine contributions.

PS if you ever come to New Zealand contact me.
Problems with grabbing opponent's 5 too early Quote
10-27-2012 , 04:30 AM
Where is real nack.
Problems with grabbing opponent's 5 too early Quote
10-27-2012 , 09:22 PM
Thanks, Grant, for the kind words. If I ever make it to New Zealand, I look forward to meeting you.

As I enjoy writing almost as much as I love backgammon, I am more verbose than most of the posters here. But I have no illusion that quantity makes quality. There are a score of really good players who post here, some frequently, others less often, who bring better insights that I. Grant Hoffman is one of them.

Grant is the winner of the 2007 New Zealand National Championship, and is also the author (with Clyde Wolpe) of the upcoming book “Backgammon Secrets.” Meanwhile, after a 20-year layoff, I have been playing only since November 2009. I’m getting better, but I am not in the same league as Grant.

More interesting than that apology, however, are the four positions in this thread. You can see their diagrams above. In each, the correct play is to make one of the 5pts. Before this thread went dormant, I posed two questions to TreeUnit designed to elicit the correct moves:
  1. If you make the anchor, @ = 24/20, 23/20, how many shots will your opponent have at the blot on the 9pt?
  2. If you make your 5pt, how many ways will your opponent have to make his 5pt on the next throw?
Any counters still out there?
Problems with grabbing opponent's 5 too early Quote
10-28-2012 , 06:10 AM
Okay, i don't know why i do this.
1a. 12 1b. 6 2a. 2 2b. 2
Problems with grabbing opponent's 5 too early Quote
10-30-2012 , 07:19 AM
Quote:
Originally Posted by yogiman
Okay, i don't know why i do this.
1a. 12 1b. 6 2a. 2 2b. 2
Your efforts are appreciated.

I am putting together some rollouts and comments that I hope to post tomorrow.
Problems with grabbing opponent's 5 too early Quote
10-30-2012 , 09:17 AM
You should not appreciate my efforts, cause i was pulling your leg.
2a. double 1,3,4 and 3/1 = 5
2b. double 1,2,3,4 and 3/1, 2/1 = 8
Problems with grabbing opponent's 5 too early Quote
11-02-2012 , 08:19 AM
This thread began when TreeUnit asked for tips about making the opponent’s 5pt early in a game. “I used to believe that grabbing the opponent’s 5pt, given the opportunity, is standard,” he wrote. “Am I going for it too early?”

Although it is hard to answer these questions without reference to a specific position, I can offer one guideline that I follow almost dogmatically. If your opponent has not make much progress offensively, and you are presented with an opportunity to make either a strong offensive play or an equally strong defensive one (such as grabbing your opponent’s 5pt), it is often correct to go with offense. If your opponent has made a key offensive point, however, or is threatening to do so, then you are often better off choosing defense. In short, offense before defense.

Four third-roll positions were posted to help demonstrate the point. All have Black opening with 41-split (24/23 13/9), and then rolling 43 on the third roll. In between, the opponent White rolls a variety of numbers on the second turn. The first of these is shown below. White played 21-split (24/23 13/11) on his turn.

.

Position 1: Black 41-split, White 21-split, Black to play 43

Black can make either one of the 5pts, his opponent’s or his own, but not both. If Black does not make the anchor on White’s 5pt, then White will have 8 rolls that make it himself: 63 61 31 33 11. With 44, White could make the 5pt, but should pass it up in favor of a stronger play: making his 2pt and 4pt while hitting! White will also be pointing on Black’s head when he rolls 55 or 66. Depending how you count it, something like 25% of the time Black will pay a steep price for passing up the anchor.

That’s enough to tip the balance in favor of defense. The 3-ply XG rollout below has making the anchor on top. Making the 5pt trails by 0.023 points of equity.

Black’s blot on the 9pt will have to fend for itself. It can be hit by 10 numbers: 61 52 43 62 53. White’s 22 and 44 are better played by not hitting.

.

Position 2: Black 41-split, White 63-split, Black to play 43

In the second position, White played 63-split (24/18, 13/10). Now White’s offensive threat is somewhat less. He makes his 5pt with 7 rolls (53 51 31 11). With 33, he should not make his 5pt. He should anchor, and make his 3pt, instead. White still makes a point on Black’s head with 55, but his 66 is better used to make an anchor.

If Black makes the anchor, White will have a direct shot at the 9pt. He hits with 12 numbers: any 2, plus the joker 44. That’s 5.5% more than in Position 1. With 53, White should not hit. Making his 5pt is better.

At the rollout below confirms, Black should play offense here. 43-point (9/5 8/5) beats 43-anchor (24/20 23/20) by 0.042.

.

Position 3: Black 41-split, White 51-split, Black to play 43

In Position 3, White played 51-split (24/23 13/8). Offensively, he is not threatening very much. He can make the 5pt with 4 rolls: 33 31 11. The joker 44 plays well for White. He should skip making the 5pt, and make two points inside, while hitting.

If Black leaves his blot on the 9pt, White can hit it with the same 10 rolls he has in Position 1.

Once again, this is a case of offense beating defense. The 3-ply rollout that follows has 43-point (9/5 8/5) beating 43-anchor (24/20 23/20) by 0.016.

.

Position 4: Black 41-split, White 61-point, Black to play 43

In Position 4, White made his bar point, playing 61-point (13/7 8/7). Perhaps this is the easiest of the four positions. White has only 5 numbers that hit on the 9pt: 62 53 44. Meanwhile, White has blocked Black’s direct exit from the 24pt. Black will have to advance his back checkers before they can jump into the outfield. This is the perfect time to anchor.

The rollout favors anchoring by 0.083.
Problems with grabbing opponent's 5 too early Quote
11-02-2012 , 08:20 AM
.

Position 1: Black 41-split, White 21-split, Black to play 43

Code:
XGID=-aa---E-CA--dDa--c-e---AA-:0:0:1:34:0:0:3:0:10

X:Player 1   O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver

Pip count  X: 162  O: 164 X-O: 0-0
Cube: 1
X to play 34

    1. Rollout¹    24/20 23/20                  eq:+0.2459
      Player:   55.50% (G:14.05% B:0.50%)
      Opponent: 44.50% (G:8.13% B:0.30%)
      Confidence: ±0.0052 (+0.2407..+0.2510) - [100.0%]
      Duration: 1 hour 36 minutes

    2. Rollout¹    9/5 8/5                      eq:+0.2232 (-0.0227)
      Player:   54.77% (G:17.71% B:0.76%)
      Opponent: 45.23% (G:11.84% B:0.50%)
      Confidence: ±0.0055 (+0.2177..+0.2287) - [0.0%]
      Duration: 1 hour 21 minutes


¹  10368 Games rolled with Variance Reduction.
   Dice Seed: 22858044
   Moves: 3-ply, cube decisions: XG Roller


Rollout by Taper_Mike
2012-Aug-31
eXtreme Gammon Version: 2.03
41S-21S-43-tm.xgp
41S-21S-43  [@ P23]  "<=10
41S-21S-43  [@ P15]  "&e
Problems with grabbing opponent's 5 too early Quote

      
m