Advanced-Level Backgammon: An Outsider's Guide
Problem #1
White rolled 61 in the opening and played the standard 13/7 8/7, making their bar point. How, then, should Black play the 41 they just rolled?
Candidate Moves
Let's look at some candidate moves:
- 13/8, passively redistributing a checker from the midpoint to the 8-point.
- 24/20 6/5, slotting both the Golden Point and the 5-point.
- 13/9 6/5, slotting the 5-point and readying a builder to help cover it (or make another point).
- 24/23 24/20, slotting the Golden Point and putting pressure on the 18-point and 17-point.
- 24/23 13/9, splitting the back men against White's 3-prime and bringing down a builder from the midpoint.
13/8
13/8 can and should be rejected here. It may avoid leaving any blots and help to unstack the midpoint, sure, but it's much too passive a play in the current position, considering White's lead in development. Playing passively now would let White get an even bigger lead, one from which Black would be unlikely to recover. Besides, the danger of being hit is less than it seems, as White only has a single point closed in their inner board and Black would be an overwhelming favorite to re-enter immediately if hit—and it's far less of a threat than letting White take the initiative even further in any case.
24/20 6/5
24/20 6/5 is much too risky in this position, but why? There are a couple of terms that come up in discussions about splitting your back men: Specifically, the "hit and split" versus the "slot and split". The hit and split is generally a solid tactic, allowing you to split safely in the face of opposition while your opponent is off-balance from being hit (since being hit deprives them of half their next roll).
The slot and split, however, is generally inadvisable, as leaving multiple blots across the board without hitting your opponent results in a weakness of position that your opponent will absolutely seize upon: Even if they miss your slotted man on the next roll, they could launch an attack on your split back men and leave you unable to cover the slot in the ensuing melee before they manage to hit it. Indeed, with this play, White will be able to hit at least one of Black's blots no matter what they roll next turn!
13/9 6/5
13/9 6/5 is more interesting. In older theory, this is the recommended move, and it's not without merit, as it slots the ever-important 5-point and gives Black a builder on the 9-point (the most useful spot for a builder, especially during the opening), putting Black in a good position to improve their position quickly should they not be hit.
Of course, according to the old theory, this move is preferable for this specific type of position less because of those factors and more because it avoids splitting the back men: White's 3-prime is stiff and will be difficult to extend in the next few turns due to the poor checker distribution (i.e. all of White's points are either stripped or heavily stacked), but White having 9 checkers in the zone puts them in a decent attacking position, thus compelling them to attack Black's split back men.
The older theory also describes how splitting the back men in the opening makes it harder for the split player to develop, since bringing down builders now leaves 3+ blots scattered across the board; and for that matter, aggressive slotting becomes too risky after an early split because it effectively puts you into a "slot and split" position. There's some truth to all of this reasoning, but we would do well to avoid adhering to it blindly, especially since newer theory (backed up by rigorous computer analysis) has successfully demonstrated that the back checkers are just as powerful of assets as any of the others.
Consider: The blots left by splitting the back men are cheap blots, but the opponent's slots and builders they'd be striking against will all be expensive blots. Thus, in any exchange of hits in the opening on the opponent's side of the board after splitting, you're statistically more likely to end up ahead. Moreover, an early split isn't quite so dangerous from an attacking perspective because, by virtue of the starting position, your opponent will only have 8 checkers in the zone early on, which is too small a force to conduct a successful attacking game—not to mention how, thanks to an early split, it becomes much harder for your opponent to bring in more checkers safely for an attack!
Always slotting aggressively in the opening like this is something of a hallmark of the older theory that has fallen out of vogue, with positional plays involving the back men consistently being shown to be a more sound strategy overall. That being said, 13/9 6/5 is by no means a bad move (especially if Black were trailing in a match and thus needing a gammon win), but let's examine the other options.
24/23 24/20
24/23 24/20, a somewhat curious-looking move, accomplishes quite a bit: It slots the Golden Point, giving Black a good chance of anchoring in the next few turns, and it puts pressure on White's outer board, making it difficult for them to bring down builders safely and thus hampering their development. Additionally, White's checkers on the stripped 18- and 17-points become inactive builders, as Black would have plenty of shots should White break either of them, even if White pointed on Black in the process. Therefore, although this play does open Black up to attack by White (a non-trivial threat, given the 9 checkers at White's disposal for attacking, along with White's desire to hit Black off of the Golden Point), it does have its merits, and is by no means a bad move overall.
Note as well that this play avoids leaving any expensive blots, which, combined with its positional strength and the potential of anchoring quickly, would make it an even stronger choice in a DMP situation, or if Black were well ahead in a match and very much wanted to avoid being gammoned.
24/23 13/9
24/23 13/9 is the most promising and well-rounded candidate move overall. In addition to getting the back men moving, which significantly improves Black's odds of anchoring and locks down White's men on the stripped 17- and 18-points, it creates a powerful builder on Black's 9-point by unstacking their heavy midpoint. While it does leave Black vulnerable to an attack by White in White's home board, White would have to hit Black on their ace-point or 2-point, neither of which are points White especially cares to make (as doing so would eliminate any chance White had of building a good prime, as well as taking the checkers used to make the points out of play)—not to mention how most attacks using the inactive builders would leave White vulnerable to some painful return shots. Outside of certain match play scenarios, this is Black's best move.
Answer
- 24/23 13/9 is the best move overall (for money games, early match games, etc.), though 24/23 24/20 is also acceptable.
- 24/23 24/20 is the best move if Black is well ahead in the match, though 24/23 13/9 is also acceptable.
- Both 24/23 13/9 and 13/9 6/5 are strong if White is well ahead in the match.
- 24/23 13/9, 24/23 24/20, and 13/9 6/5 are all strong at DMP.[1]
Key Points
- Don't play passively during the opening.
- Splitting is stronger against stripped points.
- The most valuable builder is a 9-point builder.
- 9+ checkers in the zone is a strong attacking force.
- Don't slot a point and split your back men with the same roll.
Rollouts
(click for larger images)
Footnotes
[1] Interestingly, 24/20 6/5 is also considered a strong move at DMP according to the rollouts, albeit ever so slightly inferior to the other moves. It does make some sense: The increased risk of a gammon loss incurred by the "slot and split" play is no longer a concern thanks to the match score, so the move's potential to quickly make the 5-point or the Golden Point (or both!) becomes more attractive.
<-- Problem #001 -->