Thursday, April 5, 2012

Using double to find out about fit

Lately, I've been looking for a simple, unifying principle for the system of two-way (a.k.a. cooperative or DSIP) doubles which I've been promoting in these blogs.  If you're a regular reader, you'll know about my set of triggers which switch the meaning of double from takeout to cooperative and from cooperative to penalty.  A hand came up the other day which is a perfect example of using double to find out whether we have a good fit.

Playing with my GNT teammate Jay in an ACBL speedball tournament online (IMPs), I picked up this hand: ♠ –  JT93  T8743 ♣ KQJT.  Everyone was vulnerable and partner opened 1♣.  This was overcalled 1♠ on my right and I made a slightly pushy negative double.  LHO now jumped to 4♠ and this was passed back to me and I had some thinking to do.  First of all it seemed that we had at least half the deck, high-card-point-wise and possibly even the balance of power.  We play a strong no trump so it was possible partner had a weakish balanced hand.  On the other hand, playing any sort of notrump, he could have a reasonably shapely hand with clubs and somewhat more than a minimum.  He doesn't know that my hand has a good club fit, after all.

I decided to get very pushy, and a bit undisciplined given the void in spades, and doubled.  This was no doubt due, to some extent, to having enjoyed a beer earlier in the evening.  Partner's hand was ♠ AJ3  AQ84  Q96 ♣ 974.  He was happy to pass and we set 4♠ by one trick for a gain of 7 imps.  Had he been dealt a hand like ♠ J4  AQ8  Q96 ♣ A9743, then he would pull to 5♣ and either they would bid on to a making 5♠ or we would be down one in 5♣X.

At least, that's what should have happened.

In reality, not having had a chance to discuss this sort of auction, I was worried that my double might be interpreted as penalty.  Of course, with partner's actual hand, everything would work out just fine.  But what if he had a small doubleton spade and five clubs to the ace?  We could easily suffer a double-game-swing.  Therefore, I bid 5♣ instead.  Partner pulled this to 5 and played it there.  The defense was imperfect and we ended up not only undoubled but down only two.  This was good for a 1.25 IMP loss on the board, almost a push.  Other, less fortunate pairs played 5, going down three doubled and losing 11 IMPs.  That's a total swing of 18 IMPs!!

Interestingly, of the 83 times the board was played, the scores ranged from -1100 to +500 with only eleven pairs actually making a contract (nine in 3♠, two in 3).  25 pairs played doubled contracts one way or the other, all failing.  Despite the void in my hand and 18 total trumps, there are only 17 total tricks available on this deal.

What follows is a long theoretical discussion.  Feel free to skip it :)

The basic issue is fit.  If you are considering outbidding your opponents in a competitive auction, by which I mean where each side has approximately half of the high cards, you are going to feel much more confident when your side has established a "good" fit.  While this may sound somewhat arbitrary, I'm going to define a good fit as nine cards or better (or two eight-card fits).  In the type of auction I'm talking about the "law" of total tricks is more relevant than in other situations.  And since most of these competitive decisions will involve bidding at the three or four level, a nine-card or better fit is going to give you a fighting chance of either making your contract, or at least not being penalized more than 100.

According to a pet theory of mine (actually a corollary of "the law"): if each side has exactly half the high cards, and if the deal is distributed in a reasonably "pure" and appropriately shapely manner, then each side will be able to take the number of tricks equal to their trumps.  On average!  Since this is a corollary of the LOTT it is certainly no more accurate and in practice it will be even less accurate for any particular hand.  A queen more or less will, again on average, make a difference of one trick either way. 

That's why a nine-card fit is so useful for these competitive situations.  A reasonable alternative to a nine-card fit, by the way, is a double-fit of two eight-card suits.  We're coming to the point soon, I promise.

In a perfect world, a competitive auction is won by the side which has the majority of the following assets/attributes:
  1. the higher-ranking suit (asset);
  2. the balance of power (attribute);
  3. non-vulnerability (ditto);
  4. an established good fit (ditto).
 So, let's take as an example, an auction in which the opponents have bid 3.  You're not sure if they have a "good fit" but that's not your concern.  You're considering your options: do we have the higher-ranking suit (in this case spades)?  This is the only "asset" on the list: it cannot apply to both sides as the others can. Do we have the balance of power?  It's hard to know this for sure but it should be possible to make a statement such as "with high probability we have at least 20 hcp."  Are we non-vulnerable?  This is as simple as looking at the board.  Do we have an established good fit?

If all of the answers are yes, then I think we can safely bid on.  In fact, we might want to consider bidding 4♠ any time we have 1, 2 and 4.  If we have 1, 3 and 4 then we can probably bid 3♠ with reasonable safety.  What if we have 2, 3 and 4.  This will involve bidding a part-score at the four-level which is not one of the choicest contracts in bridge.  We might want to consider a penalty double if we have the right holding, assuming that we really do have the balance of power.

Now we come to the crux of the matter.  What if 1, 2 and 3 pertain?  We have the spades, we think we have the balance of power and we are not vulnerable.  But we don't know if we have a good fit.  We might have a fit, but we don't know how good it is.  The vulnerability plays a big part in the decision.  If they are vulnerable, we are somewhat more predisposed to defending doubled and if we are vulnerable and they are not, especially if we're not sure we have the balance of power, then we are more prone to pass.  How can we tell if we have a good fit?

We double, cooperatively, showing shortness in the enemy suit, typically a doubleton at lower levels and a singleton at higher levels.  This shows a moderately shapely hand with support for any unbid suit and tolerance for partner's suit (if any) that we have not raised. If partner has wastage in the enemy suit and/or a very balanced hand suitable for defending, and no undisclosed length in our suit(s), then he will convert the double into penalties.  If it turns out we have a double fit or extra length in any known fit (if there is one), partner will pull the double.

Let's now consider the situation where we do have the balance of power, but not the rank advantage.  Let's say, too, that both sides are vulnerable.  If we have already established a good fit, then there is no need to use double cooperatively.  We can now use double for penalties! See the DSIP Rule Summary.  For other blogs on this topic see, for example, The Dead Auction Rule, Inferences from Two-way Doubles, Doubling Intervention After We've Found a Fit.


  1. What are your thoughts about supporting clubs at your first turn? You know by that time that the opponents have at least nine spades and so you can anticipate some spade raises. Since spades outranks either clubs or hearts, is there something to be said for raising clubs right away so as to avoid having that unfulfilled feeling later, when you have not yet had the opportunity to show your club fit?

  2. I'm generally a fan of supporting immediately if I have a minimum sort of hand, so your idea had occurred to me. Nevertheless, I think it might be considered a little eccentric when we have a perfectly biddable (via double) four-card major.