Saturday, October 12, 2013

When is a sequence not a sequence?

I'll admit it up front. I'm looking for sympathy.

First, your hand and the auction: ♠Q643 K862 K4 ♣832 with all vulnerable. RHO is the dealer and he passes as do you. LHO opens 1NT (15-17) which is passed out. Partner leads the ♠T ("standard leads") and this is the dummy that we see: ♠72 QT3 6532 ♣AT76. Partner appears to have around 10 points yet is apparently making a passive lead. That's a little odd but presumably there's a good reason. So, which spades in partner's hand can we definitely rule out? The knave for sure. That's for damn sure, as Jack Reacher would say. And of course the queen (we have it). It looks like declarer must have something like AKJ(x). Is there any possibility that partner has the king?

Let's take a short diversion now and consider the "standard" leads in bridge versus notrump contracts. Clearly, the highest of any sequence of honors (the ten or above) is a candidate. AKQ(x), KQJ(x), QJT(x) are all obvious and automatic leads (unless we have a very strange hand). But what if we have only two honors in sequence? We normally like to have a third honor (or high card) which is only one card separated, as a "kicker". Such sequences are called, rather euphemistically it seems, interior sequences. So, for example, AQJ(x), KJT(x), QT9(x) are "standard" leads, as are AKJ(x), KQT(x), and QJ9(x) – note that (so I believe) in these latter cases the "standard" card to lead is the lower of the touching honors [does anybody really do that?]. Of course many pairs don't play these time-honored standard leads – there's Journalist (or Vinje), Rusinow, jack-denies/ten-implies, and agreements asking for attitude or count.

So, how safe are these honor leads, anyway? And why does having a kicker increase the safety or effectiveness? Well, of course if the missing card is in partner's hand, there's no problem (unless it ends up blocking the suit). If it's sitting under you in declarer's hand, there's a decent chance that partner will get in and finesse against the missing honor. If it's sitting over your with sufficient guards, then too bad, unless partner can squish it with his hoped-for honor. Then again, there's always some chance that the missing honor will be unguarded in one of the opponents' hands. The longer your suit, then obviously the more likely this will happen. It would be tragic to lead low from you long suit headed by KQT only to see dummy win with the singleton jack!

How about the situation when there are two missing honors? Suppose that you have AJT83 opposite 52? You stand an excellent chance of running this suit for four tricks if lead the jack and your partner can get in later. Several bad things can happen – declarer holding up, the KQ being split (or both in dummy) – but in general, you have done your best and you've made it easy for partner. So this seems like a worth exception to the idea of an interior sequence having a gap of just one card.

But what about KT98(x)? Is this a sequence for the purposes of opening leads? According to Eddie Kantar (500 Defensive Tips) it is. Yet is there any good reason for making this lead? I've done a thorough study and I have not been able to find one layout where there is any advantage in leading the ten rather than the deuce. And if the deuce is fine, so is the eight. Is there any time when we don't want partner to play his ace unnecessarily? What if dummy has QJx, partner Axx and declarer xx? But we can never win five tricks with this layout and partner will do best to hop up with his ace and lead back low, keeping communications open as appropriate.

So, I'm not a believer in this kind of honor sequence lead and therefore I tend to discount it as a possibility. Is there any reason to put up the queen? Could it do any harm? Yes, it could do harm if partner's spades are T98 and declarer has AKJ7. We will have just given him the seven as his fourth spade trick.

That was my thinking in any case. Yet, twice now (including quite recently) I've been broadsided by (different) partners leading T from KT98(x). Declarer made two tricks out of AJ tight! Aargh!

So here's a request to all my partners: please don't treat KT9(8) as an interior sequence. I don't think there's a good enough reason for it. And, even when you think there is, remember that your partner is just too dimwitted to figure it out at the table.


  1. I prefer to lead the T because the 8 looks like top of nothing.

    1. Good point. It depends a lot on your agreements about leading from bad suits. Many experts like to play second highest from bad suits and yes, the 8 would look like that. But I'm assuming "standard" if there is such a thing where the top card is led from a bad suit. We would hope that the layout would come to light after the first trick if the 8 is indeed fourth best. My problem here was a trick one decision - which I got wrong :(

  2. Interestingly, Stephen Rzewski has an interesting article on when not to lead Q from QJ983:

  3. Robin, did you say that the standard lead from KQTx and QJ98 is the second high honor?

    I do not believe that is the case; I think standard is to lead the high honor, just as if the third card were completing a three-card sequence instead of being one card away from completing a three-card sequence.

    As far as what to lead from KT98, given that standard is to lead the T, if my partner led the 8 I would be unlikely to correctly fathom his actual holding. Meanwhile, if I hold the Q and he has led the T from T98, isn't my ten likely dead anyway? If so, then why not play the Q just in case the T lead is from KT98? Stated otherwise, what holdings were you hoping to gain from by withholding your Q?

  4. meant to say "isn't my queen likely dead anyway"

  5. My excuse being that I read your post remotely, thus making it harder for me to see it all, I now see that you have tried to answer the question about when withholding the queen might help.

    Unfortunately, you know that situation does not apply in the instant case. Declarer can't have AKJ7, because the seven is in dummy. And declarer can't the equivalent of AKJ6, because the six is in your hand. Thus, if your partner has led from T98 tight, then declarer has AKJ5. There's no chance of the 5 setting up as a trick, given your keeping length parity with declarer, right?

    Sorry, no sympathy this time.

  6. Ha, you're right about the 7, Jeff. It seems that a more complete analysis of this particular suit would have commanded the queen to step up.

    But my comment about sympathy was more just to get your attention. What about the KT9xx sequence in general? Does it really help partner in any material way to lead the T?

  7. It seems to me that "Winning Notrump Leads" was written to answer this kind of question. However it does make intuitive sense to me that the T is the correct lead and that Q is the correct play. The goal is to run the suit and that won't happen until the other honors are forced out. I don't see the point of withholding the Q because declarer is marked with the J. If declarer has the K also your Q is dead anyway (no harm done). If not then you can trap the J by leading the suit back when you get in. In more general terms, stick with third hand high unless you are sure.

    1. Sorry, kind of missed this comment earlier. I couldn't lay my hand on my copy of Winning Notrump Leads but it's not concerned at all with the realm of agreements and how partner will interpret the lead. It is only about efficacy (double-dummy).

      Clearly, there is no difference in efficacy between the T and the 8 (or 9) -- so that won't factor in at all. There is a difference between leading the K, T98 and/or 2 but the differences between the T98 and 2 are very small and tend to balance out.
      See Jeff's last comment for the real reason that my play of a low card was misguided.

  8. You asked whether, (presumably as a practical matter), if leading the T from KT98 holding could materially help partner. The answer is "yes". If Qxx showed in dummy and third hand held AJx, third hand would duck if dummy ducked on the ten lead. But if the eight were led instead of the ten, third hand might be tempted to win the ace and switch to a more promising suit.

  9. You are certainly right that in this particular layout, knowing that partner has the T9(8) in sequence (together with the K) would be useful. But don't hold your breath. When partner has led from the critical holding against 1 NT passed out (pretty rare in and of itself), you will hold AJx approximately once every 130 occurrences. If they had opened a 12-14 notrump, the frequency would be slightly higher. Perhaps it might happen once or twice in a lifetime? If they are in 3NT, instead, the frequency drops to about 1 in 100.

    Perhaps the most significant issue is the one that Barry raised: what do you lead from a bad suit. If you're unlikely to be leading the T then it makes leading the T from KT98(x) [a good suit] more attractive because there is less possibility of ambiguity.