[awm]

I was looking at Mike Lawrence's book on opening leads recently. He has a whole chapter on how awful it is to underlead aces against suit contracts, and another chapter about how great it is to underlead kings.

The thing I found interesting is that he gave a lot of examples like this:

---

Note that N/S should never get more than one trick in this suit. But if we underlead ace, then declarer plays low and east plays the nine, allowing declarer to score the jack and later lead up to the king.

The interesting thing is, if we switch the position of the ace and king, so opening leader is underleading a king instead of an ace, then the situation is even worse -- we have blown a trick even if partner is telepathic and knows not to play the nine, and declarer is not required to guess the suit either!

He gave many examples where underleading ace is bad as well as examples where underleading king is good, and for almost all of them if we switch the positions of ace and king the goodness (or badness) of the lead remained unchanged!

Now obviously opening leads depend a lot on the auction and to some degree on form of scoring. Here I'll assume that:

(1) You are defending a suit contract.
(2) There are no obvious clues about what to lead (partner didn't bid anything, etc).
(3) The opponents are bidding to make; it's not a sacrifice or anything like that.
(4) Scoring is IMPs.

What do people think about underleading honors?

[ArtK78]

I will sooner underlead a King than a Queen. Underleading Kings tends to produce more tricks and blow fewer tricks than underleading Queens.

I do not remember the last time I underled an Ace at trick one in a suit contract.

[mikeh]

There are a number of problems with underleading Aces, and I suspect that Lawrence has done a better job, by far, in discussing these than I can do here, even if I rambled on for pages.

The biggest ones are:

1) losing to a stiff K. Don't underestimate this: I have never lost to a stiff King because I have only underled an Ace against a suit contract twice in my life. And of course, sometimes you lose even when the K is not stiff, because with, say, Kx, declarer can win the K and then get an early pitch of the x... whereas if you had led another suit, he still gets to pitch the x but the King remains to be lost.

2) causing partner to go wrong. In addition to the examples you cite from Lawrence: consider dummy hitting with QJ10(x) and partner has the Kx(x)(x): partner cannot cover the Q because declarer may hold Ax (and a bad declarer may hold the stiff A).

Underleading a K can be costly as well. Nothing comes with any guarantees. But the underlead of the King carries with it far more chance of gaining while reducing the holdings on which it loses. The most obvious is that underleading an A loses to the stiff K while underleading a K loses nothing to a stiff A.

Furthermore, partner will never be misled: yes, sometimes the lead costs even tho partner is not misled, but he is never ducking QJ10x with Axx, while he is usually ducking that holding with Kxx.

And if you are leading from a King, even when it is 'wrong', the trick may come back. Imagine declarer with the equivalent of AQJ: your lead comes around to his Jack. But on many hands, he later has to lose to your King anyway...so the 'costly' underlead didn't cost. Now, admittedly, this has nothing to do with the underlead of the A, but it has a lot to do with a choice between a Queen and a King, and explains why many experts, in a blind auction, will lead from Kxxx rather than Qxxx (where the x's are equivalent).

[matmat]

han told me not to underlead my aces.

i am not a fan of underleading kings

[P_Marlowe]

Hi,

the main problem with underleading Aces, is partner
wont expect it.
But it can be a good idea, if you know from the bidding,
that dummy is strong and bal, and advice given by Ron
Klinger, assuming you have no better alternative or you
know, you need a miracle..

I usually dont do this, I did it once, but partner was not
in the joke, so it did not help to beat the contract.

With kind regards
Marlowe

[Cascade]

I am happy to underlead either if the conditions are right. I underlead more kings than aces.

In a recent tournament I underled two aces from memory one against 5♠ although on that hand I also underled a king and a queen and a jack. I tried low from AKQJxxx when i wanted a club ruff on the way back - the good news was that partner had the 10 the bad news was that dummy was void in the suit I led

The other one was against a more routine 2♠ contract. Here is the hand:

Scoring: IMP

---

West North East South

 -     -     -     1♦
 Pass  1♠    Pass  1♣
 Pass  2♠    Pass  Pass
 Pass

C3 CJ CQ C8
HQ HK HA H6
H5 H9 HJ H2
S6 S3 SK S5
S2 SJ SA S4
C6 C7 CA C4
HT D3 H3 H8
H7



As you can see I didn't have a good alternative lead but the small club would not have been everyone's choice.

Interestingly on the comment about underleading kings and queens my simulations, admittedly they were against 3NT at IMPs, suggested that underleading a queen was better on average than underleading a king.

One reason for this is that for a king to work you pretty much need partner to have the ace or the queen - the jack might do if dummy has Q10 and declarer misguesses. Whereas underleading a queen can work if partner has the ace, king or jack. Against suits this is probably slightly diminished since partner having the ace is less useful as the third round might never standup.

[kfay]

I've underled an ace a few times in my life. Once into KJx in dummy while declarer had xx or something like that. My partner won his queen and then.... returned something else and declarer pitched his small card later. Granted, this partner was an idiot, but it's true that your partner probably by far and away won't realize you've underled an ace until it's too late.

[ArtK78]

No trump leads and defense is different than suit leads and defense.

I don't have a strong feeling regarding the merits of underleading unsupported Kings vs. unsupported Queens against no trump. Against suit contracts, it is my experience that leading from an unsupported King works better than leading from an unsupported Queen.

[pclayton]

Underleading Aces - only in very defined circumstances and only when I don't another attractive lead.

Underleading Kings - don't mind it if I have some interior cards. Its attractive if the suit has not been bid, or has only been bid on my right (against NT).

Underleading Queens - similar prospects as Kings.

Underleading Jacks. Only when I have no other attractive alternative.

[mikegill]

I strongly dislike underleading either kings or aces. I used to underlead Ks a lot more, but I found that it very frequently costs a trick in practice. I usually don't underlead Ks now unless I'm sure an aggressive lead is required, or I feel a lead in a particular suit is required and I happen to have a K there. This is particularly true at matchpoints. I do like underleading aces when I have strong reason to believe it's right. I have done it about ~5-10 times in my life and it's been right nearly every time. Of course the last time I did it, I thought for only about 5 seconds on lead after partner had raised my overcall, and led low from Axxxx through a known to be strong dummy. It came down with KJx just like you want and declarer instakinged. Maybe I need to hold my cards back more...

[beatrix45]

Whether or not you agree with him on his each and every point, Lawrence has the best discussion of this matter in general. He also points out that on many hands, the first decision to make is what suit to lead irrespective of your high card holdings in the various suits.

During my BBO kibitzing, I have noticed that there is more underleading of aces on the go at the highest levels of bridge than Lawrence would recommend. One old-fashioned argument was that it was safer to underlead a king than a queen, because with the king, the trick sometimes came back on the third round of the suit.

[jdonn]

awm, on Nov 15 2007, 04:28 PM, said:

He gave many examples where underleading ace is bad as well as examples where underleading king is good, and for almost all of them if we switch the positions of ace and king the goodness (or badness) of the lead remained unchanged!

That may be a shortcoming of his examples, but how about mine.

---

---

---

---

Pretty much any time declarer has the king with no other honors, leading from the ace either blows a trick or breaks even, whereas if you switch the ace and king then leading the suit is probably very productive. It is only one reason, but it's a good one to focus on.

[elwood913]

Hi,

I don't have a lot of experience to tell me which honors it's better to underlead. But in reading this post it feels like sometimes underleading an ace might work well, if only partner knew... So how about having an agreement with partner that you will often underlead aces at times that seemed warranted. Then not only would partner hopefully know what was going on and you could benefit from all the times that the underlead of the ace was a winner, but also declarer would often be caught unaware. (If you had such an agreement would you have to note it on your cc?)

Just a thought,
Bill

[awm]

On the other hand, there are many positions where underleading an ace gains and king does not:

---

Underleading ace gets us our two tricks right away. While we're "entitled" to two tricks, they could easily go away on some side suit if we don't start this suit early. In addition, if partner has doubleton Qx we can get a ruff, and if declarer has doubleton Tx we probably turned one loser into two. Switch the king and ace, and we have just given declarer two tricks in a suit where he was probably only getting one.

---

Underleading ace sets us up for two tricks when partner next gets the lead. Any other lead could easily help declarer with the tempo. Switch the ace and king and our underlead has given declarer an undeserved trick.

---

Underlead the ace and declarer may well duck the trick to partner's jack, especially if there's no obvious way to dispose of the loser. This is the only way to score two tricks in the suit. Switch the ace and king and you get only the one trick you're entitled to regardless.

Anyways, to analyze some positions:

If partner has the other of the ace and king, underleading either honor is basically equivalent.

Declaring side has the other top honor only. Underleading king will work well. Underleading ace will be fine if the king is in dummy and not singleton, but frequently bad if the king is in declarer's hand.

Declaring side has the other top honor and the jack (and perhaps ten). Underleading king will often blow a trick. Underleading ace will work well if the king is in dummy (often gaining a trick) but badly if the the king is in declarer's hand.

Declaring side has the other top honor as well as the queen but not the jack. Underleading king will blow a trick if the queen is in declarer's hand. It will often blow a tempo or solve a guess if the queen is in dummy and declarer has the ace. It's fine if AQ are both in dummy. Underleading ace will blow a trick if the king and queen are split unless dummy's honor is doubleton. It may actually gain if the king and queen are together.

Declaring side has the other top honor and the queen and jack of the suit. Underleading king will blow a trick if dummy has the ace and one of the Q/J and declarer has the other. It may gain tempo when declarer has QJ and dummy has the ace. Underleading ace doesn't gain anything (may lose a tempo) but doesn't blow a trick either.

[kenrexford]

I have noticed a lot of weird under-ace leads being effective.

One classic is to lead Queen from AQx into the Kxx on dummy. The "x" spots are rather more important than suggested, though. You hope to catch Declarer with 10xx and a problem, or even 108x.

I have seen this work leading Queen from AQxx at K108x in dummy, winning the Queen, then partner winning the Jack, then winning a side entry to cash the Ace before leading the fourth for a possible uppercut.

My wife, earlier in her game, opted "fourth best" from AKJx in a side suit in a trump contract, catching Q108x in dummy. Her partner, my dad, was quite amazed to win the first trick with the 9. He fired one back, A-K-fourth round for, again, an uppercut. Declarer was equally baffled.

The classic is leading Queen from AQJ(10) at Declarer's King in hand to have him duck until you can catch his stiff with your ace (he tried to snip our communications, LOL).

Or, the surrounding play lead of Jack from AJ9...

Lots of possibilities; lots of great stories.

[cherdano]

Adam, all your examples are 4333 suits. Nobody underleads kings when both declarer and dummy are likely to be very balanced.
The reason to underleading a K is to catch partner with the queen, to set up the round 2 trick in the suit you are entitled to. The ace almost never sets up a round 2 trick.
In fact, assuming we need to be aggressive and that declarer or dummy has a doubleton, the king will do well when partner has the A or the queen. The Ace underlead will only do well if partner has the king, or the queen AND dummy has the king AND partner doesn't have the jack.

[kenrexford]

A funny story just occurred to me.

Years ago, sometime before 1990, I was playing in a sectional in Lexington, KY, in Flight A. On the very last hand of the day, I held K8 in the suit bid by dummy. The ACBL Bulletin that very month had an article suggesting that leading small from Kx in dummy's suit, in a suit contract, might work. So, I tried it. Dummy held AQJ-something and popped. It worked. I expected an average plus, because "everyone reads the Bulletin." It was a top.

[Fluffy]

I normally lead in my longest suit unless headed by the ace.

So I underlead Kings every day I play, Aces once in a while.

[helene_t]

Underleading aces works well in my experience but it is largely the surprise effect. When playing in a weak field with a weak p it pays to create confusion since defending is hopeless anyway but declarer may have a chance of getting it right if you make the expected lead.

The above issue aside, I underlead kings quite often, but rarely aces.