Playing Doublets Part I

Perhaps many beginner and intermediate players aren't aware of the following considerations regarding the playing of doublets (duplicated dice). Since doublets offer many more choices for you to play (and err), they're an excellent benchmark to assess your own checker-play skill progress. Once you find yourself consistently matching your doublet play with the top choice indicated by your analysis program (set at an expert or world-class level of course), you can be certain your whole checker play has reached a very strong level.

Let's review the steps in assessing a doublet play:

A: Identifying your top priority goal in the current situation.

B: Identifying the first one (or two or three) out of the four checkers to move, that best fit your top goal.
Here's when we have to identify all reasonable plays available, trying not to overlook any (obvious as it sounds...) Always keep in mind that doublets don't need to be played in pairs, consider the moves one-at-a-time rather than paired.

So far, similar task as on a regular roll.

C: Once settled your top priority moves, you will assess the most efficient use of the remaining moves.

You use a similar approach, eventually considering a second-priority goal. This step is obviously specific to doublet rolls.

Like in regular rolls, the biggest mistakes in playing doublets are due to failing A Erring here is like playing against your own benefit.

Second biggest mistakes are due to failing B Erring here means that you quite understood the requirements of the position but failed to choose the most effective play towards them.

Mistakes at C aren't usually so costly but still can make a difference since they represent additional bonus being wasted.

I'm an expert player, yet I misplay doublets more often than I should. It's good practice to train yourself to optimally play doublets, because it will then be a lot easier for you to optimally play regular rolls. So, when reviewing your games, do pay a bit of a special attention to how you played doublets.

Now we're going to illustrate the steps described above with a couple of not very complex examples. They were taken from rgb (, the usenet backgammon forum). Comments are based and elaborated upon those discussions.

In these examples, Black checkers move counter-clockwise (from 24-point towards 1-point). For the sake of simplicity, match scores and doubling cubes were removed, making the games 1-pointers. Coincidentally, in both examples Black has to play 44.

Position 1

A: Identifying your top priority goal in the current situation

First of all, let's find out where we're standing in order to determine what kind of game we should be striving for.
Let's assess the race:
Black = 100 - 16 = 84, White = 87.

After rolling 44, Black catches up and gets a bit ahead. Still the race gets close, so he may not be justified to go for a straightforward play if other options exist. He should rather keep pressure upon White's runners, thus trying to delay him in the race.

B: Identifying the first checker(s) to move, that best fit your top goal.

In order to "keep pressure" you will have to choose between 10/6(2), 9/5(2) and 8/4(2).

  • 8/4(2) doesn't seem to add any significance, so discard it as a first candidate.

Both remaining plays exert pressure in different ways:

  • Static - 10/6(2) holds the physical blockade on the 9-point, thus blocking most sixes.

This looks logical but, how long will this blockade last? It's doomed to be dismantled in one or two turns. Meanwhile you're left with an awkward hole on your 5-point and a weakened board that will greatly diminish the effectiveness of a hit when White runs with one checker.

  • Dynamic - 9/5(2) shifts from the 9- to the 5-point, thus strengthening the board.

This play enables (forces) White to run with one checker on a six, but this won't certainly be welcomed by the one left behind, under the gun. Okay, this creates the 66 joker, but White will get that chance soon anyway. Additionally, 9/5(2) fills in a gap for a smoother and safer bear-off. So 9/5(2) is clearly the play that best counters White's racing chances.

C: most efficient use of the remaining moves.

Now you consider:
a) 10/6(2) straightforward
b) 8/4(2) more pressure on White's runners
c) 10/6 8/4 better flexibility to achieve a safer bear-in

(c) is quickly dismissed -- not worth leaving four shots (61 and 43) for just a little more flexibility.

(a) doesn't improve towards the main goal -- pressure on White's runners . . .

. . . but (b) does! -- it adds aces as an additional attacking number.

So the strongest play is: 9/5(2) 8/4(2)!

Position 2

This may appear like a complex position, but on a closer look you will see that the real options aren't that many.

A: Identifying your top priority goal in the current situation.

What's your top priority goal here -- escaping? priming? attacking? NO -- just surviving! (four blots spread all over the board...)

Then, what would your second priority be (your winning plan)?

Let's assess the race:
Black = 151 - 16 = 135, O = 111.

After rolling 44, you’re still trailing by a significant amount. This calls for a blocking (priming) game or, if this wasn't possible, attacking -- Anything but racing.

B: Identifying the first checker(s) to move, that best fit your top goal.

Whichever the game plan you could then adopt, it's pretty apparent that you have to achieve some degree of safety so you at least can think about your options in relative calmness.

You rolled this mediocre number. Hoping to make the bar point, instead you've now got to manage to put together Humpty-Dumpty. To worsen things, White's board is strong, and stronger than yours. All this screams for 20/16 as the very first and vital four to play.

C: most efficient use of the remaining moves.

In this example, the strategic part rests on the remaining moves, which should adhere to the second objective stated in A.

Since you're unable to prime White's straggler this turn, you've got to decide whether to play on for the prime or to attack. That is, to choose between 13/9(3) and 13/9 5/1*(2) or 13/1*.

  • 5/1*(2) is plain awful! This trades an excellent blocking point for the worst one. Actually you're easing White's life, pushing him to enter on a much better point than the one he stands on right now. Discard.
  • 13/1* at least follows to a reasonable idea -- trying to protect the slotted bar-point. Nonetheless, it creates another liability on the ace point.

The problem with any attacking play lies in that White's board is much stronger than your own, turning any return hit into a potentially winning play. As a general principle, you're eager to attack when you have the stronger board, or when you have no option. Not the case here -- which leaves us with the last and only good prospect:

  • 13/9(3), the cold-blooded play. It keeps hoping that the prime is the true winning method. Should White miss the bar-point, you've got good chances to block the White straggler real hard. Even if getting hit the game isn’t over, with some luck you could still enter making an anchor and resist from there.

So the best play is: 20/16 13/9/(3).

*  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *

Well, I hope you found this article helpful somehow, but wouldn't like to finish without the usual word of discouragement to the student-player.

Many (most?) times, the situation won't be as relatively-clear in its conclusions as it turned out to be in these examples. You will often be faced with conflicting factors, which overlapping each other could turn a "normal" advice around. Monsters like match score, cube value and position, gammon threat, can twist the advice towards a more bold or a more conservative decision than the pure basic concepts would lead to. I removed those features to keep these examples easy to understand.

The depicted method is just a logical skeleton. At least, it will help you guess why you blundered so badly on that doublet...

Good games