DeepSqueeze

The ABC’s of Encapsulations

You may have noticed a lot of cryptic strings floating around like 'WLc, w < a, b'. Probably you aren't interested and should just ignore them.

But in case you are curious, this is a brief introduction to an important element of how a lot of this work was accomplished.

We all have an intuitive concept that some positions are "the same" in all important respects. It doesn't matter if our last winner is an Ace or a seven that is currently high. It doesn't matter if all our spades are actually switched with all our clubs. Often an "idle" card could be in several locations without any real impact.

The use of 'x' in a diagram ("I opened 1♠ with AKxxx") is an example of abstraction: diagrams with x's are asserting that the exact spots are irrelevant. You could shuffle some around and the essentials would stay the same. Of course there are risks: sometimes those x's play important roles that are obscured. You wouldn't want to use AKx/Qxx in the context of an entry-shifting squeeze.

Here, we attempt to uncover the essential elements of a squeeze, and use this underlying structure for various goals. Starting with being able to determine whether 2 squeezes are meaningfully distinct, or not. To do this we attempt to summarize the layout of an entire suit with a string of letters, and thus the layout of an entire deal with 4 such strings. First a few examples...

NorthA J
WestK Q
Eastany
South2
NorthQ 7
WestJ 9
East3 2
South4

This is a pretty classic "two-card" menace over West. In an important sense, the first is identical to the second.

We represent this with "Wa":

W is for "winner". W vs w distinguishes whether an entry is available. Here South's small card provides a link so we use W. An otherwise identical layout with South being void would be "wa" instead.

'a' is the most powerful standard length-based threat: it is isolated against one opponent. Read above, or grade-A. We also have b for an isolated threat below its stopper, and c for a threat stopped by both opponents. (A guard menace or clash menace requires a bit more notation than we're describing here.)

From one suit to a whole deal

Positional squeezeWLa, a > Lw
Open in Explorer ↗
Double squeezeWLc, a > Lw, a
Open in Explorer ↗

The first deal is a positional squeeze with some minor untangling required: ♦A, ♠K, ♦K and West is squeezed in the majors.

We represent this as WLa, a > Lw

This introduces a bit more detail. First, we have a '>' which divides the 'encaps' into 2 different primary hands (North then South) and also indicates the lead (by pointing to South). The notion of "primary hand" is important, too: this will be the longer holding. In basic notation one letter corresponds to one card in the longer hand. So WLa tells us we have a winner (see above), an L (see below), and an a-threat (see above again) in North.

The convention is also that a capital letter indicates a paired card in the South hand (such as W). JRL coined an extremely useful term, TicKet, for the pair of cards that provide an entry. Here, the ♠3 ('Tic') links to the ♠A ('Ket') to provide a TicKet, a frequently essential element of a squeeze. And quite often you can predict with reasonable certainty that they will be played on the same trick, though of course anything is possible. They are captured together by 'W'.

We've already seen this with W vs w. But now we add L to indicate an entry to the shorter hand. Here the ♠2-♠K are both captured by L. WWa would be AKx/xx in contrast.

Hearts (a) is also primary-North with just a single threat, also grade-a. Meanwhile, diamonds is primary-South, but is a blocked position (Lw). The ♦A is an entry to the opposite hand (link card or tic ♦3), while the ♦K is a winner without a link.

What about clubs!? On the first layout, the ♣2 is idle and need not be specified. Inspection will tell you it could be in any suit without materially impacting the play. If we wanted, we could (and in some contexts will) add an 'i' to capture an idle card. Sometimes we want an encapsulation to focus on the essentials, and other times we want to fully represent a particular deal. For the latter we would add 'i' in the appropriate place. (You may also see some m, o, u describing idle cards in other hands.)

It does look like yet another a-threat, though, which brings us to the second deal. This is a double-squeeze, where ♦A, ♠K, ♦K squeezes West in the majors and East in the blacks (simultaneously, if you care). Only East's hand has changed, beefing up spades so East also stops the suit.

This position is WLc, a > Lw, a. Very nearly the same as the last one, but one a-threat (in spades) has been downgraded to a c-threat, while one idle club has stepped in to add a threat against East, which is ultimately required.

The hope is that by discerning the underlying structure of these positions we can accomplish various goals like

  1. Putting up a good defense even in an ultimately losing position.
  2. Explaining a solution concisely instead of 37 if-this-then-that statements.
  3. Recognizing similarities between positions. Or even essential sameness.
  4. "Cooking" positions by taking a working squeeze and testing various slight weakenings.
  5. Finding untapped latent possibilities in a position
  6. Discovering novel squeezes