[Flame]

Is it better to play 2M in 5-2 fit or play 1nt ? (or maybe when is each one better?)
There are some bidding sequences that lead to this question.
One is for those who always or almost always support with 3 cards in the sequence 1m-1M, so for them a 1NT rebid is always or almost always with 2 cards in the major, should partner having 5 card bid 2M now ? another example is
1m-(p)-1H-(1S)-1NT playing support double this deny 3 hearts.
Im not talking on caseswhere there is still a chance for a 5-3 fit when partner is opening 1nt.
If someone has dealmasterpro or any other problem which could deal and play double dummy, i would be happy to hear what the program result on this matter.

[whereagles]

If pard's 1m-1M-1NT rebid denies 3-card support, the 5-2 fit is still good if the hand either very weak or lacks side entries to the suit (e.g. KQJxx and out). When the hand has scattered honors and the hcp strenght is enough for 1NT, it is ok to let it rest in 1NT.

If pard's 1m-1M-1NT does not deny 3 cards, then a 2M rebid is usually in order. If there's a 5-3 fit, 2M is almost always better than 1NT and some 5-2 fits play well anyway.

[FrancesHinden]

The 5-2 fit will generally play better:

- the 5-card suit is in the weak hand and there is no other source of tricks (with KJ109x xx Qxx xxx I would pull to 2S after 1C - 1S - 1NT, but with 109xxx Qxx J10 Kxx I passed a strong 1NT on Saturday)
- when you have a side-suit un- or not very well stopped (a bit of an obvious one)
- when you will play on elopement lines (taking a lot of low ruffs in the long hand)
- when you can take a ruff in the doubleton hand (another obvious one)
- when they have an obvious opening lead against 1NT
- when the choice is playing 2M by the strong hand or 1NT by the weaker hand

The 5-2 fit will play worse:

- when you have a source of tricks elsewhere
- when you have exactly 7 tricks in either contract
- when you have slow side-suit cards that they can ruff (e.g. QJxx opposite 109x is two tricks eventually in 1NT, but may be none in a suit contract).
- at matchpoints, when you have the same tricks in either contract

A couple of general observations about another common 5-2 M suit fit:
Playing semi-forcing 1NT (or forcing 1NT) you play in a 5-2 fit as opener much more often than playing a more SA or Acol-based system when you simply pass after 1M-1NT. Having played both methods, my feeling is that it's very common for both contracts to take 8 or 9 tricks, but that 1NT goes more off when it goes off. That makes 2M more attractive at IMPs than at MPs.

The auction 1m (P) 1H (1S) 1NT is slightly different, because opener volunteered the 1NT bid over the 1S overcall, so should have a positive desire to play in 1NT rather than just be making the 'systemic' call - they could have passed over 1S. So I would remove much more rarely on this auction.

[Walddk]

FrancesHinden, on Feb 1 2006, 01:59 PM, said:

The 5-2 fit will generally play better:

- the 5-card suit is in the weak hand and there is no other source of tricks (with KJ109x xx Qxx xxx I would pull to 2S after 1C - 1S - 1NT

I agree, as long as partner never rebids 1NT with 1444 or 1345. Some do (I don't like it), but if you have an agreement that 1NT always shows a balanced hand, or at least semi-balanced (2245), it's usually best to pull to 2 of the major.

Roland

[inquiry]

I didn't feel like doing a simulation, so I used BridgeBrowser. I set the following requirements for my search.

1) Opener had 11 to 14 hcp
2) Opener had exactly 2 hearts
3) Opener started biddign with 1C or 1D
4) Opener could not have a signleton
5) Responder had exactly 5 hearts
6) Responder had between 7 and 9 hcp
7) The bidding started (with silent opponents) 1m-1H-1NT,
8) The final contract was either 1NT or 2H (both from the openers side, yes there were some 2H contracts from the other side, but I used bridgebrowser to ignore those...btw, it turns out NOT to be a good idea to play in 2H when they are split 5-2 against you and they have a weak NT versus a responder with 5hearts and 7-9 points).

In the last quarter (from october to end of december) on BBO tournaments and team games (I ignored main room). This was the results....

1m-1H-1NT all pass occurred 607 times (with the hcp and suit legnths specified).

The result were, 403 imp hands for an average +0.54 imps, and 117 matchpoint hands for an average of 57.83% matchpoints.

There were more auctions that ended up in 2♥, but I did not limit the auctions to 1m-1H-1NT-2H (which I could have). There were 453 of these (with the limits on hcp and distribution). Of these, 340 were imps, resulting in a net -0.17 imp per hand, and 113 where matchpoints, resulting in 52.71%.

So from this analysis it would seem that, in general, playing in 1NT is better than playing in 2H. But there was no division for "weak" versus "strong" responders.

Once you find the hands with BridgeBrowser, you can then quickly see how the hands do with various "combined" hcp ranges. For instance, if the partnership held a combined 18 hcp, do they do better in 5-2 fit or 1NT? There were 6 such imp hands in the data I looked at, and 1NT average +1.19 imps, and 3 such MP hands and playing 1NT with less than half the deck averaged 96.12%. Comparing this to playing 2H with 18 hcp. There were only two such hands, one at imps and one at mp. (1.39 , and 17.39%).

Similar informaton is availabe for each of the other combined hcp range, or you can look by heart suit quality, specific shape, number of "controls", etc.

If you haven't looked at Bridgebrowser lately (or never looked at it), the new plot feature, where after you find the bids, you can plot the distributions for each bid, the hcp distribution, the suit quality, the final contract, make finding info very easy and intuitive. For instance, on the auctions that began 1m-1H-1NT listed above (were responder was also limited to exactly 5H and 7 to 9 hcp), the 1NT rebids hcp was
11 hcp - 62 times (averaged 0.71 imps, and 60.88% Mp).
12 hcp - 337 times (average imps 0.1 and 52.49%MP score
13 hcp - 400 times (average imps 0.27, and 56.96%)
14 hcp - 371 times (average imps 0.11, and 56.69%)

The most comon responder shape was 4432 (525) followed by 5332 (493) then 5422 and 6322.

Of course this type of analysis doesn't show rather people like Frances is correct when they say, 5-2 plays better in hearts...if. There are a number of ways you can figure that out with Bridgebrower. One way is to click on the final contract to call up all the hands, then sort by score (imps or matchpoints), then look at the ones that did well, and the ones that did poorly.

Anyway, hope that helps. If you are interested in BridgeBrowser, you might want to check out their webpage, or the new homebase club, which will be offering HUGELY discounted online bridgebrowser access (forget about the pricing listed on the BRBR page if you join homebase, it will be extremely affordable if you are interested in it).

For more information about BridgeBrowser, see: http://www.microtopi...BRBRonline.html
For more information about HomeBase Club, see:http://forums.homebaseclub.com/ or coming soon... http://www.homebaseclub.com/

[luis]

Great post Ben, thanks for the information!

[barmar]

Hey Ben, how much is Stephen paying you to be his shill?

[FrancesHinden]

Ben, why 7-9 HCP in responder's hand?

If I responded 1H with xx KQ109x xxxx xx I'd be bidding 2H over a 1NT rebid and confidently expect that to be the winning action. The weaker responder is, the more likely (I think) it is to be right to pull 1NT.

[kenrexford]

Obviously, these statistics are a guide and useful. However, I am concerned about some tactics assumptions as well as some theory-to-player linkage.

As to the latter first, I suspect that 1NT scores better that 2M in some part because better declarers tend to prefer 1NT to 2M. Not because of bidding theory, but because declarers with skills love to declare 1NT. Hence, the level of declarer ability might not be identical between the two sample groups.

More importantly, though, is that the various stopping points have different likelihood of ending the auction. A stop at 1NT is corrected by the opponents rarely. However, a stop at 2H, and even 2S, often yields competition, placing our side on defense. When both 2M and 1NT score above average at times, competition is confirmed.

The importance of this suggests that competing over 1NT may be under-appreciated. At a minimum, you transfer the opposition to 2M, apparently, for a better score if the contract is the problem. You may even buy the contract, perhaps above 2M, which seems to be superior statistically.

Thus, the statistical analysis suggests some additional principles. First, 1NT may be "superior" because most people defend 1NT poorly. Second, 1NT is also superior, perhaps, because most people do not compete over 1NT despite statistical advantage to doing so. Third, 2M may be inferior because it invokes competition.

Applying all of this, one might argue that 2S is superior to 2H, because the opposition bidding must be a level higher, that doubling might be under-used and need re-thinking, and that the decision between 2M and 1NT might be governed by defensive strength, suggesting that 2M is best when holding defensive values.

[MickyB]

When you are in a 5-2 fit, I don't think you should mind the opponents competing.

[inquiry]

barmar, on Feb 1 2006, 10:31 AM, said:

Hey Ben, how much is Stephen paying you to be his shill?

I really like BridgeBrowser, I have been using it for years. If you search my name and Bridgebrowser on this forum, you will find I have mentioned it a lot, because I like it. While Dealmaster Pro and other methods (jack, dealer with GIB, etc) can simulate hands, I have always preferred looking at real world hands to answer questions like the ones posed here. BridgeBrowse allows me to do that. So when I can try to contribute to a thread with such information, I do so. I always have.

But a disclaimer since you asked. Stephen and I are both associated with the new private homebase club exclusively on BBO, and this browser will be available to homebase club members at least 80% off list and in some case even more to members/teachers. While most people here probably have only a vague clue what BridgeBrowser is, the club will start doing things soon, so I took the opportunity to mention the some of the features and advantages of BridgeBrowser. There are many more, but no need to go overboard. Most people who might join the homebase private club could not care less about bridgebrowser, and those people who will want nearly free or free access to the bridgebrowser online will contact us or join the club eventually with or without comments made in this thread. So the hawking is not necessary, I simply mention it in passing for those interested... (As an aside... i can't help myself, after looking at 30,995 1NT direct seat overcalls on 15-18 hcp and balanced distribution, the average result was -0.10 imps and 50.14% matchpoints.. note this is just for the hand overcalling 1NT without analyzing what the final contract was... did that for the raptor thead, but didn't include it, as the numbers seem obvious, especially with such a large number of bids).

[han]

Very interesting Ben. I agree with Frances that rebidding 2M seems most interesting when responder is weak, say 5-7 HCP. Even better, let responder be weak with at least some number of HCP in hearts, like the hand that she suggests xx KQ109x xxxx xx. I bet that if you set up the search to include only those hands with which experts would bid 2M then the 5-2 will be a winner.