Notes from Kantar’s Modern Bridge Defense

by Matthew Kidd

When I started to get serious about bridge I worked my way through several of Eddie Kantar’s books on play and defense, including “Big Red’, Defensive Bridge Play Complete (1974), the bible for intermediates and players seeking to be such. From a design perspective, “Big Red” isn’t much to look at, seemingly no more than 500 typed pages cheaply glued together. But those pages are packed with tons of good information presented in a very readable style.

By time Kantar repackaged much of “Big Red” in the much more visually appealing two volume set Eddie Kantar Teaches Modern Bridge Defense (1999) and Eddie Kantar Teaches Advanced Bridge Defense (1999), I had moved on. But recently I ran across these two books and decided to read them quickly, looking for tidbits I had never absorbed, methods that might be standard in 1999 that weren’t so standard in 1974, and also for better ways to teach new players. There are of course many places to graze for defense tidbits but Kantar has a real knack for focussing on situations that are both important and common, indeed situations that will come up every session. I figured I would find some good stuff and I wasn’t disappointed.

Here I list a few things that caught my eye. I’m not summarizing the entire books. If the excerpts here don’t make much sense to you, then go buy the two books. They are available used for about $12 each and will improve your game greatly. But if the material does make sense, I suggest sitting down with your favorite partner and seeing whether you are on the same wavelength in the situations discussed.

Unfortunately, very little of the convention card is dedicated to defense and most of that section is for opening lead agreements. Moreover, very few defensive layouts are named in short memorable ways as is the case for bidding, e.g. Stayman, Jacoby Transfer, negative double. Two decent players with a strong understanding of the logic of defense may end up doing the right thing—“Oh, this has to be a suit preference situation”—most of the time, but I strongly suspect even decent partnerships could boost their matchpoint session average at least 2% or 3% just by tightening up their defensive agreements for common situations.

Four preliminaries

Kantar’s examples all assume standard carding because his material is targeted toward intermediates. He mentions upside-down carding so players will know what it is and doesn’t say anything negative about it, perhaps even viewing it as marginally superior. If you prefer upside-down carding, you can easily modify the carding in the examples below.

In both books, ace is led from ace-king, the majority position today though possibly not in 1999 when the books were published. This is a change from “Big Red” where king is led from both ace-king and king-queen which was entirely standard at that time.

Top of nothing with three small is standard against notrump. The smallest from three small is standard against suit contracts in a suit partner hasn’t bid. Nonetheless, Kantar recommends “top of nothing” as well against suit contracts, while admitting that experts usually stick with the standard small lead. There is no good solution to this old problem (don’t say MUD—it has issues too) but personally, I like “top of nothing” because I think partner can often distinguish from a doubleton based on the rest of hand; for example whether the opening leader turns up with a trump control. But my reasoning is not foolproof either since some doubleton leads are intended more as a safe lead, often through a suit bid by dummy, than as an effort to score a ruff.

Kantar also comes out in favor of “Jack denies”, at least for beginners and intermediates. With this agreement, the jack is never led from AJT or KJT as it would be normal using standard leads and therefore the lead of a jack always shows a sequence like JT9 or JT8, a singleton, or a doubleton, but in all cases denies a higher honor. Experts tend not to like this lead agreement because it gives too much information to an expert declarer. Even if “jack denies” does not help declarer on the lead, it can be enormously beneficial in placing the other defensive high cards. Since I am targeting advanced intermediates, I will ignore examples that involve “jack denies”.

Equal honors

Partner leads the ace, dummy has small cards in the suit, and you hold J732. Should you encourage with the seven? Or similarly, when partner leads the king and you hold T732? No and no. Only encourage when you hold honors that are equal to partner’s honors, i.e. which complete a sequence with the honors that partner has promised. In the first hand partner has promised the king, so encourage with the queen. In the second case, partner has promised the queen so encourage with the ace or jack. If you encourage with any old honor, partner may end up leading into a tenace or lose a trick playing low, expecting to be able to put you in.

There is nothing sophisticated about this and yet I feel many books on defense don’t make this point very clear when introducing attitude signals. Giving a good name to an idea that otherwise takes a sentence or two to explain, is often half the battle. “Equal honors” is typical of Kantar’s direct and clear style, a term on par with Audrey Grant’s “golden fit” to refer to eight card or longer major suit fit. Teachers should adopt the term “equal honors” immediately if they do not use it already.

Attitude has priority over count on the opening lead but give count if all the missing honors pop up in dummy. For example, if partner leads the queen, and dummy has AKT2, give count.

Encourage with highest of equals

When partner leads the ♠Q and dummy plays low, encourage with the ♠9 which denies the ♠T. Partner will know not to continue the suit.

I think most players get this right, after all if you going to encourage it might as well be as encouraging as possible. Just don’t get lazy and play the ♠8 because partner might think you have T 8 x and are showing an equal honor.

Highest card plays from perfect sequences

Standard defense for decades is to lead the top of a sequence and for the third hand to play the bottom of a sequence. Kantar hews the line here. But he parts way with standard dogma in second seat, recommending playing the highest in a sequence of three or more cards (a “perfect sequence”), i.e. the card you would have led had you been on lead. He claims this is easier for partner to read and is the favored approach of experts.

When dummy leads a small spade, play the ♠Q. If declarer wins, partner can unblock if that is useful.

Exception: Holding KQJ(x), don’t play the king because partner may think you have AK. Play the jack because partner will usually be able to tell the difference between KQJ and JT9.

Be careful about splitting from QJTx and JT9x if declarer is known to have four cards in the suit. Setting up a trick for declarer is far worse than clarifying the suit position for partner.

When you have an honor sequence, or the equivalent of an honor sequence, and declarer wins the trick, play the highest honor.

When East switches to the ♠T, and declarer plays the ♠K, West plays the ♠J, denying the queen but confirming the low equals.

Kantar also recommends discarding the top of a three card or longer sequence (provided of course that you can afford it). Here is an example of how useful this can be.

South preempts 5♣, silencing everyone. West begins with two high spades before exiting passively with a trump. What will West discard on the run of clubs? If West discards the Q early, West will know it is safe to bare the K (or even pitch all his diamonds) and the defense can not be denied the setting heart trick.

Equals in second seat

Kantar’s rules for the play of equals in second seat are:

  1. If you know you are going to take the trick, take the trick with your lowest of equals.
  2. If you can’t be sure you are going to take the trick, play you higher or highest equal (the one you would have led), but with KQJ, split with the jack (see above) and with KQ, split with the queen.

If you work through the cases, with AK you know you will win the trick, so play the king and rule #2 says to play the queen with KQ. These are the old-fashioned splitting rules. It is only with weaker equals like QJ and T9 where you play high to clarify the position.

Showing doubleton honor sequences

It is standard to reverse the normal lead from a doubleton honor sequence, e.g. leading the king from AK tight or the queen from KQ tight. When you show up with the other card or must have it given declarer’s line of player, partner knows you are a doubleton. This can be extended to other situations, for example playing the queen in third hand and then showing the king, indicates KQ tight.

Ace from ace-king

Ace from ace-king only applies to the opening lead. Lead the king from ace-king on later tricks (many players go wrong here). For later tricks, starting with the king from ace-king is more sensible because one needs to cash out an ace late in the defense or start a suit with the ace when not holding the king. Similarly, lead the king from ace-king at trick one if partner has supported the suit. This is to distinguish between holding ace-king and merely leading the ace in the suit your side is strongest in, often a good lead.

Finally, lead the king from ace-king against five or six level contracts. If dummy has the queen, third hand must give count so that the opening lead knows if the ace is cashing.

Queen led against a no-trump contract

It is useful to adopt the agreement that a queen is led against a notrump contract, requests partner to unblock the jack, catering to a lead from KQT9. If partner does not have the jack, he signals encouragement holding the ace and discouragement otherwise. This rates to be a safe unblock because it is the only time a queen is led against notrump without the jack (the other standard queen leads are from sequence led by QJTx, QJ9, and AQJ).

Supported versus unsupported suit leads against notrump

If you are leading partner’s suit but have not supported it:

With a doubleton, lead high.
With three cards, lead low, e.g. lead the 2 from either 862 or Q62.

If you are leading partner’s suit but have supported it:

With three or four small cards, lead high.
With three or four cards headed by an honor, lead low.

In the first case, you are communicating length; in the second attitude because partner already knows something about your length, i.e. 3+ cards.

Attitude or count dilemma

When a low card is led against notrump, signal attitude if dummy wins the trick with the king or ace, given count if dummy will win the trick with queen or low. This makes sense because in the later scenario, your attitude is known—you were not able to cover!

Most common suit preference scenarios

The most common suit preference scenarios are:

  1. When giving partner a ruff
  2. When partner leads an obvious singleton
  3. When partner leads an ace in a supported suit and sets up a number of winners in dummy
  4. When you have promised six or more card in a suit
  5. When dummy has a singleton

The first scenario is very standard and the third is commonsense. See the next section for more on #4. The second scenario hinges on the word “obvious.” A typical situation is when partner leads a suit you are long in and which declarer also bid but which did not end up becoming the trump suit. It is more obvious if partner leads small, especially if you know it is his smallest card, ruling out a doubleton lead.

It is worth discussing your partnership’s treatment of the final case when dummy has a singleton. I’ve always played it as suit preference for the higher or lower of the two remaining non-trump suits. But the problem with this approach is that sometimes continuing the suit is important, typically either to promote a trump holding or remove a trump entry prematurely, often with the consequence that declarer cannot draw trump finishing in dummy. Kantar’s treatment is to play this more like an attitude situation, where a high card requests continuing the suit and a low card requests the “obvious” shift. There is that word “obvious” again. Nonetheless, thinking back to cases where dummy has had a singleton, I agree with Kantar that partner has an obvious shift 90% of the time. Still what if you need the non-obvious shift? You request that by dropping the high card you can afford. Since dummy has a stiff, you are likely to have enough cards the suit led to cater to all three scenarios.

Suit preference with known 6+ card suits

Normally suit preference signals convey a choice of two suits, the suits remaining after the suit in which the preference is given and trump suit are excluded. For example, if hearts are trump and the ♣2 is a suit preference signal, it is suit preference for diamonds, the lower ranking suit of diamonds and spades, the two remaining suits.

But when the defender giving preference is known to have a six card or longer suit, e.g. after a preempt or a suit bid and rebid, he will have enough cards to reliably distinguish between three suits. Here are two examples.

East preempts 3 and South winds up in 6♠. West leads the A and East plays the K to request a shift to diamonds. The 3 would request a club continuation. A middle heart would request a heart continuation. Discuss this with partner. Many partners will interpret the K as an attitude signal and continue hearts. But I think Kantar is right to introduce the three suited preference treatment to intermediates. Yes, it is one more thing to remember but it addresses the common situation of communicating well when partners leads the suit you preempted in or rebid.

East preempts 2 and South winds up in 3NT. West leads a heart, and declarer ducks the J and K. East can now drive out declarer’s ace with any heart. The Q indicates a potential spade entry, the 5 indicates a potential club entry, and the middle card denies a likely entry in either suit. When East plays a middle heart, West can work out that the Q is the only hope of East ever getting in again, and make the spectacular play of discarding the A on the third heart.

Signals when discarding twice

When you make two discards in the same suit, the first discard is attitude and the second is present count. This is easy to remember because it is analogous to what you do in third seat for the suit led (provided you don’t need to cover or aren’t immediately giving count at trick one because your attitude is clearly negative).

Giving count when discarding in dummy’s strong suit

When dummy has a powerful suit missing the ace or king, the defender who does not have the missing honor can make a count discard in that suit. This allows his partner to work out declarer’s length in the suit and hold up an appropriate number of rounds.

The Left Jab

Against 3NT, you lead the ♠K, partner playing low denying the jack or ace. Declarer plays low. Next you continue with the ♠J, your lowest remaining equal, requesting an unblock of the ♠T. Partner obliges, and declarer wins the third spade, discarding a club from dummy.

What do you play when declarer leads a low diamond toward the entryless dummy? The Q! Declarer can’t let West in and is therefore held to two diamonds for down one or even down two if he finesses the 9 later. If you play small at trick four, partner wins and returns a heart. But declarer hops up with the A and runs diamonds to make an overtrick. If partner doesn’t cover the 9 at trick four, declarer cashes out winners to make the contract on the nose.

Eric Rodwell named this type of play, i.e. the Q here, the “Left Jab” in his advanced card play book The Rodwell Files. But Kantar in his unassuming style pointed it out to intermediates fifteen years earlier in a short section titled “Second Hand High?”

This is the most common left jab position, dummy having no certain entries. If West plays his honor on the first trick, declarer is held to one trick and will get none if he tries to finesse twice. If West plays small, declarer will get four tricks if East takes his honor and two tricks if he ducks.

Jack led from dummy, ace in your hand

In second seat, the general rule is to play the ace at notrump, but low against a suit contract.

When the jack is led, you know declarer isn’t finessing because you have the ♠T. Play the ♠A to promote partner’s ♠8. However, if you are sure declarer has no side-suit entry to hand, win the second spade to cut declarer off from his spades.

At a suit contract when you are not looking at the ten, it is usually right to duck the jack. Duck here, hoping that declarer misguesses giving the defense two tricks.

The general difference between the notrump and suit contract scenarios is whether you are trying to permit the eventual establishment of tricks or take as many immediate tricks knowing declarer can ruff away losers.

Don’t be queen foolish

Do you cover when declarer leads the jack and dummy has the ace? No! And not with Qx, Qxx, Qxxx, or Qxxxx. For lots of reasons. If declarer has KJT9 he may overtake the jack with the ace and take a losing finesse into you. If declarer has JTx, covering the jack allows declarer win in dummy and finesse the ten. Ducking holds declarer to one trick. Think through the other scenarios.

On lead with AQxxx or AQxx against notrump

Should you lead into declarer’s like king to get AQxxx or AQxx started? Kantar says the fifth card usually makes it worth it from AQxxx. But stay away from AQxx.