Bray’s Learning Curve: Third Move Thoughts

Money Play. How should Red play 62?

2018 - Beginners 4

XGID=-a—-E-D—fD—c-e–A-A-:0:0:1:62:0:0:3:0:10

There are 15 possible opening rolls. Many of these can be played in different ways. In response each of the possible 21 numbers can also be played in a number of different ways. There are approximately 600 opening rolls and responses. Most players will get the majority of these plays correct over the board.

The same is not true of the third move because by now there are too many possibilities for us to remember and we must rely on general principles to guide us most of the time. Those guiding principles are:

  • Make new points
  • Unstack the heavy points
  • Hit an opposing checker
  • Start the rear checkers moving

Over and above these principles I also like to try to maintain flexibility so that I have the ability to adopt different game plans as the game evolves.

This week’s position occurred in a teaching session with two of my students. Red opened with 52 played 24/22, 13/8 and White responded with 65, played 24/13. How should Red now play this 62?As usual we must consider the candidate plays and then make a choice:

  • 24/18, 13/11 too many double hits.
  • 22/14 minimises shots but it will be difficult to improve the position next time.
  • 24/16 leaves the rear checkers connected, but gives White many more hits, some of them double hits.
  • 24/22, 13/7 gives Red a solid anchor and slots a good point and would be the choice of many.
  • 22/16, 13/11 unstacks the mid-point while reducing White’s hitting numbers and keeps some connectivity between the rear checkers.

My students both selected 24/22, 13/7 and I wouldn’t argue with anybody who wanted to make that move in a chouette. However, settling for the 22-pt anchor so early is a little negative and the blot on the bar-point is exposed to 17 shots.

Long experience had me leaning towards the more flexible 22/16, 13/11 but I wasn’t certain by any means. A long rollout proved me to be correct. It turns out that 24/22, 13/7 is too committal so early in the game and technically it is an error. Remember that flexibility is a key element of backgammon and 22/16, 13/11 provides Red with exactly that.

The key here is to remember the solution to this problem and then apply the technique in similar future positions of your own.

Rollout Data from Extreme Gammon

Beginners 4 Rollout

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

 

Bray’s Learning Curve: Deep Thought

Money Play. How should Red play 42?

2018 - Experts 3

XGID=–bBbbCBCB-Ba—–abbab–A:0:0:1:42:0:0:3:0:10

Before the advent of the bots this roll would have been played bar/21*, 8/6 and nobody would have known that they were making a big error. The early bots such as TD-Gammon and JellyFish were not trusted to play back games correctly and so their recommendations would have largely been ignored in positions such as this.

However, by the time that gnubg and Snowie arrived players had begun to accept that their ideas were fundamentally sound, although there was still some doubt about complex positions. By the time Extreme Gammon took to the stage the bots were viewed as very strong players and everybody paid heed to their suggestions.

I misplayed this position over the board, making the pre-bot era move. As soon as I analysed the position with XG I realised I had forgotten to take its teachings into account. The correct move, by a long way, is bar/21*, 3/1, a move that would have been laughed at in the 1970s. Anyone wanting to make that move in a chouette of that era would have had their mental stability questioned.

However, when you look at the move and analyse it, it makes perfect sense. White has a very well-timed back game and his checkers are all working. In contrast thirteen of Red’s checkers are reasonably well positioned but the two checkers on his 2-pt are out of play. In military terms those two soldiers are off the field of battle and need to re-join their colleagues if possible.

The best way to correct Red’s timing is to force White to hit those checkers so that they can be recirculated. This is the key piece of logic behind bar/21*, 3/1. Having studied bot play in some depth I knew all about this particular recirculation tactic and I was annoyed with myself for missing the opportunity in live play.

Bar/21*, 8/6 is very nearly a blunder and any other move is a blunder or worse. So, please remember the idea and make use of it in your own games, the opportunity occurs more frequently than you might expect.

Rollout Information from Extreme Gammon

Experts Rollout 3

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

 

Bray’s Learning Curve – Volatility

Money Play. Should Red double? If doubled, should White take?

2018 - Intermediate 3

XGID=-B-BBaCCB-A———-bbcg-:0:0:1:00:0:0:3:0:10

This is a type of position that I often see misjudged. What normally happens is that Red rolls something like 63 which is played 10/4, 8/5*, White fans and then Red wins a point by doubling White out.

Red has committed the cardinal sin of losing his market because he feared the sequence where he hits loose and then White return hits from the bar and then Red fans. Beware, considering unlikely nightmare scenarios is not the way to make decisions in backgammon.

Red’s thinking should go along these lines:

  • If I point on White’s blot (15 rolls), unless he rolls 26 from the bar I will have lost my market.
  • I will pick and pass with 14, 24, 34, 54 (8 rolls) and if White fans, again I will have lost my market.
  • With all remaining numbers except 64, 44 and 66 I will hit loose and if White doesn’t hit back yet again I will have lost my market. I note that 44 and 66 are both very good for the race.
  • Yes, I will lose some games where I do get hit and stay on the bar but overall, I will gain many more points by doubling now than I will lose in the odd case where thing go wrong.
  • The position is incredibly volatile and so I must double now or I risk losing my market.
  • Therefore, I double.

On the other side of the coin White has an easy take based upon his hitting and racing chances. Like many middle game positions this is both a correct double and a correct take.

Note that if White had a five-point board or even a closed board the doubling decisions for both players remain the same.

Success in backgammon is largely determined by one’s cube handling and here not doubling would be a blunder. If you didn’t double because you were afraid of being hit from the bar then hopefully after reading this article you will have learnt to think more positively.

Rollout Information from Extreme Gammon

Intermediates 3 Rollout

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

 

 

 

Bray’s Learning Curve: War Games

Money Play. How should Red play 65?

2018 - Experts 2

XGID=–BB-bBCA—bB—c-bcbCa–:0:0:1:65:0:0:3:0:10

Never forget that backgammon is a representation of war, as indeed are most of the older board games. Thinking in terms of war can often help you to arrive at the right decision with difficult plays.

Faced with this 65 to play over the board, Red elected to move 22/16, 8/3. This seems logical, escaping one of the rear checkers and moving the blot on his 8-pt to safety. The problem is that Red now has one soldier away from the field of battle (the third checker on his 3-pt) and another (the blot on his 16-pt) that might struggle to reach safety next turn.

He could have elected to play 13/8, 13/7, playing completely safe but that play cedes a huge part of the field of battle (the outer boards) to his opponent and he will really struggle to escape his three rear checkers. As is often the way in war cowardice is not rewarded.

He could try 22/16, 13/8. This makes the important 8-pt but leaves two checkers exposed to White’s attack and his hitting numbers are not duplicated with the numbers he needs to cover his blot on the 2-pt. Once again, Red will struggle to safety the blots next turn even if they are missed.

By now you will have found the correct move, 22/11. This keeps all his soldiers (checkers) focused on a key area of the battlefield. The move duplicates White’s three and sixes and now, if the blots are missed, they can be usefully employed to make new points or more easily made safe. If White does give up his anchor to hit (in war terms giving up a key strategic point) then Red’s checkers (soldiers) are well-positioned to counter-attack.

You cannot implement a battle plan that do you not even consider. Here many generals (players) would not even see the possibility of 22/11 and even if they did, they might consider the risk to life and limb too high a price to pay.

However, much of war and all of backgammon is about exercising judgement, creating a balance between risk and reward. Here, the risk is worth the reward. Leaving a double direct shot is relatively rare but this position demands precisely that. In backgammon terms, any move other than 22/11 is a blunder.

Rollout Information from Extreme Gammon

Experts 2 rollout

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

 

 

Bray’s Learning Curve – Match Play Doubling

Match Play. Red trails 8-9 to 11. Should Red double? If doubled, should White take?

2018 - Intermediate 2

XGID=-a-B-aDAB—cD–ad-e—-B-:0:0:1:00:8:9:0:11:10

For years match play doubling was little understood. Witness this quote from Georges Mabardi in 1930, four years after doubling was introduced into backgammon: “If two absolutely perfect players engaged in a match, there would never be an accepted double.” Match play would be dull indeed if we followed that piece of advice!

Until work began on match equity tables (Woolsey and Kleinman amongst others) everyone doubled as if playing for money. Slowly things changed, and the advent of the bots gave us a huge leap forward. Nowadays most players know at least some of the theory of match play doubling. To be successful in match playyou must know the correct doubling strategies for all the scores near the end of the match.

3-Away vs. 2-Away recurs constantly. The correct strategies for the two players are clear. The 3-Away player should double earlier than usual because the 2-Away player needs 28% winning chances to accept the cube (this figure can easily be derived from match equity tables).

If there is any threat of a gammon for the trailer, then the double should come even earlier. Note that for the leader the cube has no value (he can never win with a redouble) and winning a gammon is meaningless as it gives him more points than he needs.

I had this position against Eric McAlpine in the final of the Scope Charity Tournament earlier this year. I was Red and trailing 8-9 to 11. I studied the position for quite a while and then decided it met the criteria outlined above. The race is even but I am ahead in position and threat. Crucially, White does not have an anchor and hence I could sniff a hint of a gammon in the air.

Much to the surprise of the spectators I doubled. Eric correctly accepted. My next roll was 61 played 13/7, 6/5*. Eric didn’t anchor, and after I soon rolled a set of double fours he was in real trouble. I won a gammon (and the match) with a roll to spare.

The key lesson here is not to be afraid to put the match on the line if the odds favour you. This concept is alien to some players but until you learn that lesson and put it into practice you will never be a good match player. As you can see from the rollout data not doubling here is a blunder. Most experts would get this position right, but intermediates and beginners struggle with problems like this one.

Rollout Information from Extreme Gammon

Screenshot 2018-09-01 16.53.51

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

 

Bray’s Learning Curve: One Back versus Two

Money Play. How should Red play 32?

2018 - Beginners 2

XGID=-aa–BD-B—dF–ab-d-b–A-:0:0:1:32:0:0:3:0:10

This is a type of position that I often see misplayed and not only by beginners! Red has done well to escape a rear checker and has also made his own 5-pt. Meanwhile White has made his 4-pt and split his rear checkers. What is Red’s game plan?

Ideally, he would like to escape his other rear checker, failing that he will try to make some new points, or if rolls something like 44 he could switch to blitz. However, he has rolled a poor 32 and must decide how to play it. Realistically there are four choices:

  1. 6/1*
  2. 13/11, 13/10
  3. 24/22, 13/10
  4. 13/8

Let us now discuss these in detail:

Move a) is the weakest of the four. White is not threatening anything on his side of the board and Red should not risk getting a second checker sent back having just escaped one.

Move b) looks like a natural developing move but it does leave nine shots. More importantly if White doesn’t hit one of the two new blots he may switch to attacking Red’s blot in his home board as a diversionary tactic. If Red were to stay on the bar at any point, then those two blots represent a large (and unnecessary) liability.

Move c) combines trying to escape with creating a new builder, a sort of halfway-house. This move gives White two targets as he will certainly hit on his3-pt if he can (assuming he doesn’t roll a number that hits Red’s outfield blot). 24/22 puts Red ’under the gun’ unnecessarily.

Move d) is the correct move. It maintains Red’s slight racing lead, doesn’t give White any new targets to shoot at and brings a new builder for Red’s home board. If you read Magriel’s Safe Play vs. Bold Play criteria in his seminal work “Backgammon” you will find that this position meets the criteria for a Safe Play. 13/8 looks somewhat dull and prosaic but it is the right play for this type of position.

When you have only one rear checker it is rarely right to advance it to a point where it can be attacked, unless it is imminent danger of being primed.

Rollout Data from Extreme Gammon

Screenshot 2018-08-22 11.17.09

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

 

Bray’s Learning Curve: Match Play Complexity

(a) Money Play. How should Red play 32?

(b) Match Play. Red leads 7-2 to 11. How should Red play 32?

Bray’s Learning Curve Match Play Complexity 8-21-18

XGID=–BaB-BBCa–dC—–b-bcbA-:0:0:1:32:7:2:0:11:10

The most common error that I see in backgammon is players not adjusting their play to the score and consistently making money-game decisions, both in checker play and cube handling, in match play situations. At an even score a lot of decisions are the same as their money counterparts but once the score becomes skewed then beware, a lot of things are going to change.

To demonstrate that point let us analyse this week’s position. In a money game it should be clear that the ‘do nothing’ play of 13/8 is not the right idea. Giving White his full roll cannot be the right idea. With two White blots to attack Red may well win with a blitz.

The correct play for money is 8/3*, starting a blitz. The strength of this play is that when White fans, which he will do 25% of the time, Red will win the game with a well-timed double. Red wins just too high a percentage of gammons for White to be able to accept the cube while of course the Jacoby Rule protects White from losing a gammon, i.e. Red will not play on.

Now let us consider the match play scenario. At this match score Red will be doubling much later than usual and ideally in a situation where there are very few gammons for either player. That is not the case at the moment.

Now after 8/3* followed by a fan Red cannot even double! White will accept with alacrity and be looking to redouble to 4 at the earliest opportunity. A gammon with the cube on 4 will get White to 10-7 (Crawford) with his opponent on an odd number which is perfect for him. (Always beware of giving your opponent the opportunity to get to either Crawford or to win the match with perfect efficiency)

Equally well, 13/8 is far too passive. Red needs to improve his position if he can. Is that possible? Yes, it is. It is close, but he should play 8/5, 7/5, making the all-important 5-pt. This leaves one fewer shot than 8/3* but crucially it strengthens Red’s home board so that in any ensuing exchange of hits Red’s stronger board may have a powerful influence and that in turn will reduce his gammon losses. With the score as it is Red wants to reduce the inherent volatility and making the 5-pt does just that.

8/5, 7/5 is difficult to find, particularly over the board where one’s natural instinct is to hit. The key is to make sure you consider the play. As has often been pointed out: if you don’t see a move, you can’t play it.

Match play is far more difficult than money play and although doubling has been around for 92 years we are still only now getting to grips with understanding it, and in match play our knowledge, for the vast majority of players, is still not where it needs to be. – Chris Bray

Rollout Information from Extreme Gammon

Money

Bray’s Learning Curve Match Play Complexity 8-21-18 XG Money

Match

Bray’s Learning Curve Match Play Complexity 8-21-18 XG Match

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

Bray’s Learning Curve: An Anchor Question

Money Play. How should Red play 65?

8-13-18 Intermediate Position

XGID=-B-BaCB-B—bB-c-c-d-Bb—:0:0:1:65:0:0:3:0:10

This sort of position often confuses even good players. Red was hoping to attack the White blot on his 4-pt but instead rolled this useless 65. How should he play it?

The instinctive reaction of the majority of my students when shown this position was to elect to hold the status quo with 8/3, 8/2. A few brave souls decided to risk a 3-shot and cleared the mid-point with 13/8, 13/7. This play has the merit of keeping those two checkers in play. About 10% chose the running play 21/10, exposing two blots. So which group was correct?

One of the things I stress to my students is that irrespective of how they conduct themselves in their day-to-day lives, when they play backgammon they have to be bold and fearless. That is easier said than done and it takes a long time to learn to adopt the correct persona over the backgammon board.

One of the best books for backgammon players was written 2,500 years ago! It is “The Art of War” by Sun Tzu, a Chinese army general. Backgammon is fundamentally a representation of war and many of Sun Tzu’s maxims apply as well today as when they were first written and browsing some of the quotes from the book will help anybody’s backgammon game.

Back to the position. 8/3, 8/2 buries two checkers (soldiers away from the field of battle are of no use) and does not improve Red’s position. 13/8, 13/7 is better but now the rear checkers are really stranded (a divided army fights inefficiently).

Red must make the bold play 21/10. This play follows the principles of war by engaging in a skirmish while you have superior forces (a stronger home board) and also follows the basic tenet of backgammon, “when ahead in the race, race”.

White’s sixes are duplicated and unless he manages to point on the Red blot in White’s home board, Red will have lots of counter play, including remaking the 4-pt anchor in some variations. The play forces White into a fight when he is not quite ready.

The XG rollout shows just how weak the two ‘do nothing’ plays are. In XG terms they are both triple blunders, precisely the type of mistake none of us can afford to make.

21/10 looks risky and we are taught to only give up anchors after long thought but here it should be very clear it is the right play following the right game plan. If White rolls 66 so be it, you will just know it is not your day!

Rollout Information from Extreme Gammon

8-13-18 Intermediate Position-XG Rollout

Bray’s Learning Curve — A Great Member Benefit
Bray’s Learning Curve is a USBGF online series by author Chris Bray. Each week Chris lends his sharp insight and easy to understand analysis to help you improve your game. Visit the USBGF Facebook page every Monday to view an interesting backgammon position and join in the lively discussion, return on Tuesday to view the answer. In addition, as a USBGF member, you get access to this companion blog article that includes an expanded explanation.  More about Chris Bray

 

Bray’s Learning Curve: Early Game Dilemma

Money Play. How should Red play 31? 

XGID=-a-a-BD-B—dE—d-e—-B-:0:0:1:31:0:0:3:0:10
CB problembd 1

For this first column I am going start at the beginning of the game. This position is reached after Red opens with 31: 8/5, 6/5 and White responds with 32 which he has correctly played 24/22, 13/10. If your opponent opens with 31 it is very often correct to split the rear checkers as soon as possible. Red now has to play another 31.

Newcomers to the game will quite likely play 13/10, 6/5, unstacking the two heavy points and getting a good distribution of checkers. Unfortunately, this move and the more aggressive 24/21, 6/5 are both bad blunders. In backgammon there are errors (a mistake but not a huge one) and blunders, which are very bad mistakes and to be avoided if at all possible.

In the opening we try to do three things: make new points, unstack the heavy points (the mid-point and the 6-pt) and get the rear checkers moving. There is a fourth possibility when replying to the opening move and subsequently and that is to hit an opposing checker and send it to the bar. If your opponent opens with 32, played 24/21, 13/11, and you respond with 64 I am sure you will hit 24/14*.

What beginners often do not realise, because they have yet to learn the tactic, is that a hit in the home board can often be the correct play. It may seem counter-intuitive to give your opponent a direct shot at a home-board blot because if it is hit you lose a lot of ground in the race. However, that is negative thinking which is to be avoided whenever possible in backgammon.

The key to hitting in the opening phase of the game is that it takes away half (or sometimes all) of your opponent’s next roll as he must first enter his hit checker from the bar. He might return hit one of your checkers but unless he rolls a double he won’t be making any new points in his home or outer boards (he might make an anchor in your home board).

In the opening taking tempo (time) from your opponent is very important. If he is busy dealing with your threats he won’t have time to develop his own threats.

The correct answer to this week’s problem is 24/23, 6/3*. This move activates the rear checkers, unstacks a heavy point and puts your opponent on the bar – not bad for one simple play! On four of your opponent’s rolls (66, 55, 65 and 56) you will be delighted with the outcome. He will return hit you with sixteen rolls but that’s no disaster as he has no home board. With double 1s he will play bar/23, 6/5(2). He misses with the other fifteen rolls after which you will have a good advantage.

An expert will play 24/23, 6/3* here without even thinking but as a beginner you have to understand the reasoning and then make the play. Once you have learnt the technique it will become second nature.

If you extend this concept you will understand why after your opponent opens with 52, played 24/22, 13/8, and you respond with 64, the correct play (although only just) is 13/3*. In summary, the opening is largely about tempo.

–Chris Bray

Rollout Information from Extreme Gammon

CB problem 1