[karlson]

pclayton, on Dec 12 2007, 12:32 PM, said:

If I read this table right, it says that NT will only take more tricks than spades about 1 out of 20 times? That seems incredible to me.

This is given that you have a spade fit (the hands that matter), so yeah, does it really seem so weird?

[MFA]

3NT.
The extra strength does it for me.

I'm confident that Karlson's analysis would have had a different result, had it been single dummy analysis.
Let them lead, for instance. They will be striving to pick a major.

[whereagles]

pclayton, on Dec 12 2007, 08:32 PM, said:

If I read this table right, it says that NT will only take more tricks than spades about 1 out of 20 times? That seems incredible to me.

yeah, but in NT you need 1 trick less

[Codo]

karlson, on Dec 13 2007, 05:13 AM, said:

That's 55.3% of hands on which spades plays at least 1 trick better than NT. A ballpark estimate of the standard error is (.5)(.5)(sqrt(1/1000)) which is about 0.8%. So this is a very significant difference if you believe the setup.

If I get your table right, you say:
If you have a spade fit you score better in 55 % of the cases but worse in 45 %.

Maybe you should make a second table without a spade fit. How often will the defending side lead significant weaker because you did not tell them about your hand and pds hand?

I would prefer some single dummy data for both cases.

[pclayton]

jdonn, on Dec 12 2007, 12:45 PM, said:

pclayton, on Dec 12 2007, 03:32 PM, said:

If I read this table right, it says that NT will only take more tricks than spades about 1 out of 20 times? That seems incredible to me.

That seems normal to me, double dummy. The main gains from notrump come from opponents doing the wrong thing. This happens much more against notrump since it can happen in many more ways, two examples being they lead a spade from Jxxx into partner's ten, or they lead low from something headed by the KQ and let a random jack score, neither of which will happen against 4♠. Still this confirms what I already believed. You simply want to play in spades when you have a spade fit opposite this hand.

Well with this much strength, we can survive a bad spade break for instance and rely on other tricks. Maybe 20:1 is normal though after I think about on a DD basis, but it certainly isn't in those cases when they lead a spade. It doesn't have to just blow a trick into pard's 10x to work, it can also blow a necessary tempo.

[pclayton]

whereagles, on Dec 13 2007, 01:14 AM, said:

pclayton, on Dec 12 2007, 08:32 PM, said:

If I read this table right, it says that NT will only take more tricks than spades about 1 out of 20 times? That seems incredible to me.

yeah, but in NT you need 1 trick less

Obv

[ArtK78]

That is not true. The scoring is match points. So you need to take the same number of tricks in notrump to succeed against those in spades.

[whereagles]

well, there's a mix-up here. don't worry. rage on

[jdonn]

Codo, on Dec 13 2007, 05:53 AM, said:

karlson, on Dec 13 2007, 05:13 AM, said:

That's 55.3% of hands on which spades plays at least 1 trick better than NT. A ballpark estimate of the standard error is (.5)(.5)(sqrt(1/1000)) which is about 0.8%. So this is a very significant difference if you believe the setup.

If I get your table right, you say:
If you have a spade fit you score better in 55 % of the cases but worse in 45 %.

You misread the table. He is saying when you have a spade fit, spades makes more tricks than notrump 55% of the time. But notrump makes more tricks than spades only 3% of the time, and they are the same 42% of the time.

[benlessard]

Im sure that is a close call. Good declarer vs bad defenses is more likely to be able to steal a trick in Nt then in 4S. So against a weaker field the temptation to go for Nt is higher. Defender will sometimes blow a trick on lead or unable to avoid some endplay. Often the hand is in S always make 5 but in nt you have to do some work to make 5.

[han]

Seems clear to use the gadget.

[Codo]

jdonn, on Dec 14 2007, 01:52 AM, said:

Codo, on Dec 13 2007, 05:53 AM, said:

karlson, on Dec 13 2007, 05:13 AM, said:

That's 55.3% of hands on which spades plays at least 1 trick better than NT. A ballpark estimate of the standard error is (.5)(.5)(sqrt(1/1000)) which is about 0.8%. So this is a very significant difference if you believe the setup.

If I get your table right, you say:
If you have a spade fit you score better in 55 % of the cases but worse in 45 %.

You misread the table. He is saying when you have a spade fit, spades makes more tricks than notrump 55% of the time. But notrump makes more tricks than spades only 3% of the time, and they are the same 42% of the time.

If I make the same number of tricks, NT scores significant better at mps.
So 55:45 is still correct.