01-20-2011
, 04:22 PM
Please tell me you didn't think it was some weird takeout thing.
01-20-2011
, 05:26 PM
The 2H balance is not terribly strong — probably averages 12 or 13 HCP, and can be lighter. So we aren't stunned that partner doubles. But what does the double show?
- First, the double isn't random: partner has very good shape for the double, extra strength, or both.
- He rarely has six spades, as most players will rebid a six-bagger in this spot. But we're playing five-card majors so play him for exactly five.
- The auction (including RHO's heart raise) tells us opps have nine hearts most of the time — 8 is possible if RHO has extra strength but he might have chosen redouble or 2S in that case, and 10 is also possible but less likely statistically and also suggested against given the not-terribly-aggressive opp bidding so far. So let's say pard has two hearts.
- He doesn't have a second five card suit, because most players would bid it here and because the double suggests tolerance for anything we do.
- This means pard's most likely shape is 5=2=3=3, with a chance of 5=2=4=2 (perhaps planning on bidding 3D over our 3C, not that that would be a great idea); 5=1=4=3 or (less likely because of our hand) 5=1=3=4 are also possible, but these unbalanced hands are less likely. Even 6=1=3=3 could happen if the spades are bad — but notice that for the spades to be bad requires a particularly unusual layout.
- But wait — if partner's most likely shapes are balanced, what does that tell us? A weak balanced hand shouldn't have doubled (nothing extra), and a 15–17 count balanced hand would have opened 1NT (assuming we're playing strong NT and assuming partner isn't an idiot); that means partner's most likely hand is something like a 5=2=3=3 18 count (without a heart card, or he'd have rebid 2NT), with a 5=1=4=3 15–17 still in the running. Interesting.
Notice that the law of total tricks also says don't bid: they usually have nine trumps but we usually have 8. Bidding 3-over-3 usually requires 18 total trumps.
What if we were 3=1=4=5? Now things are different: Partner is now more likely to have the big balanced hand so we have more stuff, our hand is contributing a likely ruffing trick, and that ruffing trick will also provide a hand entry. That's why we can consider bidding on. I'd do it w/w because it's really wrong only when we're down two and they double, and doubles are less likely when we're not vulnerable. Note if we're setting them we'd have to beat them three, or double them and beat them two, to beat the value of the 2S we probably could have made with the cards lying well, and those things aren't happening.
r/w, it's probably still not worth the risk. If opps are red, I know that all we need is to beat them 2 to get a very good board, and that's possible if pard has a strong hand. I still might do it if w/r (I said earlier that I wouldn't, but that may have been wrong), just because it's not likely to cost very much, either because I'm not doubled or because I go down only one.
So what about on the extended auction, when opps bid game and it swings to us?
Well, pard didn't double the final contract. Since we're providing a likely 0 tricks against a heart contract, we can safely assume that opps are making their game most of the time (but not always, and that's important too). But why did they bid it? They probably have either extra strength or extra shape, each making an unbalanced, and on average weaker, hand for partner more likely (weaker because the unbalanced 15 counts are now more likely, while the balanced hands were all too strong to open 1NT.)
If it's w/r, we can lose six tricks and still profit in 4S, provided we think the game will be bid at most tables. (We can't win if nobody else is bidding it.) Assuming LHO bid confidently, we can consider the sacrifice when we're 3=1=4=5, because we probably have a nice secondary fit in one of the minors which is both tricks and a late entry; we won't be surprised to lose a spade, a heart, and four minor suit tricks, for down three and a decent board. If we're red it needs to be a trick better than that, and I don't see it being more likely. If they're white they're slightly less likely to have stretched for the game but again we need to do a trick better, and it's less likely the field will bid the game. If it's r/w, forget about it.
What if we're 3=2=3=5? It's really bad if pard also has two dead hearts, because again that's two guaranteed losers and no board entries... but it does make it more likely that pard is the one with the stiff spade. Is that enough to bid on at good colors? I don't think so, because with a 5=1=4=3 good enough to hold our non-heart losers to five, partner might have found a third double, and he passed. That's why I'm passing at all colors with the flat hand. We'll probably get an average; bidding on rates to be a bottom a lot of the time and a top only when it's right and everyone else bid the heart game.
How does all this change at IMPs? Not a ton. Re the partscore decision, it makes bidding on when vulnerable a little more scary, but it was pretty bad at matchpoints too because of the risk of –200. But when not vulnerable, a double of 3S is extremely unlikely (it rarely pays much, and risks a game bonus), so I'd be slightly more inclined to do it w/r than at matchpoints (with the 3=1 hand only). At w/w I think it's a wash.
The game bidding is interesting. If they're vulnerable at IMPs it's more likely than it would be at matchpoints that they're stretching for the game, and sacrificing against a failing game is a huge swing, so I would bid on only if my opps' bidding was very confident, with the 3=1 hand. At all other colors, and with the balanced hand, no — even when the sac is right it's only two IMPs, and when it's wrong it's more than that, sometimes considerably more.
Last edited by atakdog; 01-20-2011 at 05:41 PM.