[hanp]

Phil's "MPs with 8-card fit" reminded me of a hand I had a couple of weeks ago:

J10x
xx
KQJ9x
AKJ

1NT - 2D
2H - 2NT
??

Nonvulnerable at IMPs, 1NT was 14-16. (Alert!)

I thought this was an interesting decision and I did some double dummy simulations afterwards. I will share the results later, I thought they were somewhat surprising.

What would you do? How do the negatives (small doubleton heart, poor placement of club honors?) weigh against the positives (great 5-card suit)?

Is this another WTP? Just like Phil, the decision I made at the table was the losing one, and just like Phil's, my partner thought my bid was clearly wrong.

[Apollo81]

I would just bid 3NT.

If my range was 15-17 then I could be talked into a NF 3♦ bid.

[whereagles]

It's a 3NT wtp for me.

If pard has a singleton spade, then it's just that the methods are inferior

[Phil]

3N looks kind of obvious. Maybe even more obvious than my thread.

Curious to see the conclusion.

[ArtK78]

From the point of view of the NT hand, this is a clear 3NT call. There is no way that the NT bidder could possibly conclude that any strain other than NT is right, and it sure seems like a 3NT bid.

However, given the title of this thread, it seems that 4♥ is the right spot. That is something that the other hand might be able to figure out.

It is not unusual for the 5-2 fit to play better than 3NT. The factors which the responder may be able to use to figure this out are the solidity of the suit, the lack of outside entries or a point of vulnerability in the responder's hand. Looking at the NT hand, I would guess that responder is short in spades and that is why hearts is likely to play better than NT.

Matt Granovetter commented a number of times in his long-lost magazine that the 5-2 major suit fit often played better than 3NT.

[jdonn]

I think you are reading too much into the title.

I also think this is a wtp 3NT.

[hanp]

Yes, playing in 4H did not occur to me.

[gwnn]

3NT, yay to 14-16, I think it's highly likely the range pwned the field this time. (I know that some other times we miss games and that it's possible pard has a 10 count)

[neilkaz]

gwnn, on Mar 19 2009, 04:53 PM, said:

3NT, yay to 14-16, I think it's highly likely the range pwned the field this time. (I know that some other times we miss games and that it's possible pard has a 10 count)

This is an awfully good 15 HCP here, and playing 15-17 I also accept the invite to be honest. I love my ♦ I like my JTx♠ and on a lucky day I get a ♣ lead.

Huge WTP 3NT playing 14-16, IMHO.

[hanp]

Again I dealt 1000 hands, partner holding 9 points, 1-3 spades, 5 hearts, 1-4 diamonds and 1-4 clubs.

3NT made 33% of the time.

Again, this is a double dummy statistic and may not be accurate for real world results. Also, if you expect partner to hold this shape and 10 HCP then the chance of making 3NT should obviously be higher.

Quote

3N looks kind of obvious. Maybe even more obvious than my thread.

I think that the double dummy results suggest the opposite (33% vs 43%). I am not surprised.

[hanp]

This may be an interesting comparison:

On 1000 hands where partner haa the same constraints (9 HCP, 5 hearts, etc.) and we have 15 HCP, exactly 2 hearts and 2-5 cards in the other suits, we make 3NT on 36% of the hands. That's a bit more than on the given hand.

Does that mean that the given hand is slightly worse than an average 15-count when partner invites with 5 hearts? Or does it mean we shouldn't trust double dummy results here?

Roger makes a good case for the latter, I won't try to put words in his mouth.

[Fluffy]

3% in 1000 hands doesn't look like definitive to me, what I read is that our hand is average (when I though it was actually very good 15 HCP as others). Not bad one.

I think heart holding makes all the difference here.

I also think 3NT is a contract that plays double dummy better for the defenders.

[gnasher]

hanp, on Mar 20 2009, 04:23 AM, said:

Again I dealt 1000 hands, partner holding 9 points, 1-3 spades, 5 hearts, 1-4 diamonds and 1-4 clubs.

3NT made 33% of the time.

On how many of those hands would 2NT have made?

[JLOL]

Downgrading for xx in partners suit in this auction should be normal. A lot of 15s with xx should pass I think. Also downgrading for a poor spade holding because partner has at most 3 spades and they are likely to lead the suit effectively should happen more often than it does.

That said KQJ9x is a very nice source of tricks and should clearly be upgraded for, and as gnasher implies it's not like 2N always makes when 3N is going down so we should really be looking for reasons to bid 3N not for reasons not to.

[hanp]

I think that's all exactly right JLOL. Gnasher's question is of course good even without an answer, but I will come back to it when I have the time.

[jdonn]

It's a very difficult question to answer because even if 2NT makes, you might go down more than 1 in an effort to make 3NT (say taking a finesse for the 9th trick with a suit unprotected.) It can actually work the other way too (the best line for 3NT making and the best line for 2NT going down) but that seems a lot less likely.

[hanp]

gnasher, on Mar 20 2009, 08:44 AM, said:

hanp, on Mar 20 2009, 04:23 AM, said:

Again I dealt 1000 hands, partner holding 9 points, 1-3 spades, 5 hearts, 1-4 diamonds and 1-4 clubs.

3NT made 33% of the time.

On how many of those hands would 2NT have made?

I redid the simulation with the same specifications. This time 3NT made 345 times and 2NT made 688 times.

Ignoring all other factors the IMP expectation is:

345*(+6) + 343*(-5) + 312*(-2) = -0.269 IMPs.

Again, the double dummy scores might be worse than table results, although this auction is very different from 1NT-3NT. Also, as Josh pointed out, at the table you might take more or fewer tricks in 3NT than in 2NT. Finally, 3NT might get duobled on a bad day. And if you expect partner to often hold a 10-count instead then you should clearly accept (1.910 IMP expectation).