[diana_eva]

Because I am such a nice helpful gal, I was rewarded recently with a few free lessons on BBO from a friend who is also a bridge teacher. I was thrilled to give it a try, and went there confident like a sun flower. Unfortunately, at the end of the session I was feeling more like a rotten carrot, given that I had failed approx 75% of the declarer play problems.

Here's one where I didn't see the right play until the teacher walked me through it:

IMPs. Just make the contract, nothing tricky. Lead is ♠Q

---

[Cyberyeti]

Edit: garbage deleted, serves me right for failing to count properly, I thought I had 8 off the top not 7,

[PhantomSac]

Probably not what your teacher wants but I would always duck the first spade irl.

Spoiler

I am guessing that they want you to win the first spade and try a heart finesse, if that wins then you can safety play diamonds and play K A and another so you don't lose to Qx on your right, if it doesn't you need to play for 4 diamond tricks so you can hook. In real life I don't view a club shift at trick 2 as likely at all if I duck trick 1 so I think it is a lot better

[billw55]

I like problems like these. Everyday, realistic hands add up to far more IMPs than newspaper deals.

Spoiler

I have 7 cashing tricks, so need two more. There seem to be four chances for a trick: 4th spade, heart hook, diamond hook, 4th diamond. So I need 50% success, and probably correct order. The spade seems less likely based on the lead, but that is not certain; ♠QJT or even QJx is a reasonable lead on this auction.

I also must consider that a club switch at any time is very dangerous. So I cannot duck trick one - I'll take it in hand. Now I try diamonds first; if I am lucky I can get both extra tricks there. So ♦K and low to ♦J. If this loses and a club comes back, I may still make if diamonds break and the heart hook works.

If the ♦J holds, now I can afford to duck a spade and try all three remaining chances for the 9th trick, heart hook last.

[PhantomSac]

Yeah billw's line must be right, I forgot spades might be 3-3 lol. That has to be more likely than Qx of diamonds offside obviously. It also preserves lots of squeeze/endplay chances if they mess up.

[Winstonm]

Like Bill showed, there are 7 top winners. The deal it would seem would be to try to develop two more tricks and keep as many options open as possible.
It looks like a club switch should be discouraged, so I agree that winning the spade in hand is right. The best source of extra tricks is diamonds, and that is where I start with an immediate diamond finesse. If that wins, I now have three chances for a ninth trick - 3/3 diamonds or spades or the heart finesse. As I want to have as many options as possible, I will test the diamonds. If they are not 3/3, I can still duck a spade, giving myself an extra chance for the spades to be 3/3 or the heart finesse to be on.

If the diamond finesse loses, I am pretty much committed to 3/3 diamonds and a winning heart finesse, unless the opps continue spades which I could duck and then win the 3rd round.

[diana_eva]

The lesson was exactly what Jlall posted first time. Before playing on diamonds find out how many you need. Heart finesse worked, so I had to take the safety play in diams for 3 tricks, and Qx fell offside. I don't think the exact hand matters, it was a prepared deal, but here it is:

---

[PhantomSac]

Here is a better example of this theme since I think billw's line is better even though I got the "solution."

---

You get to 6N and they lead a heart. How do you play?

Spoiler

If the diamond hook is on, you can lose a club and thus can afford a safety play of ace and low towards the K9 picking up the suit all the time. However, if the diamond hook is off, you need to take 4 club tricks and should start with a club to the jack. So play a diamond to the jack early. This is definitely a good theme to know about.

[PhilKing]

I think teacher was definitely right a few years ago. Nowadays, three card major leads are becoming more popular, so it's pretty close.

The hand is still instructional, and to my way of thinking is more so because of that. Bridge is not a textbook game, even though knowledge is important, and set hands should reflect that a bit more often.

[diana_eva]

Good point re spade lead. I actually asked him how likely it would be for the lead to be from a 3 carder based on the bidding, and he said assume these opps wouldn't lead that. He just wanted me to see the safety play theme.

I didn't include this hint in the OP because I was curious how it should be played if the lead could be anything.

[BillPatch]

 PhantomSac, on 2014-August-07, 10:03, said:

Probably not what your teacher wants but I would always duck the first spade irl.

Spoiler

I am guessing that they want you to win the first spade and try a heart finesse, if that wins then you can safety play diamonds and play K A and another so you don't lose to Qx on your right, if it doesn't you need to play for 4 diamond tricks so you can hook. In real life I don't view a club shift at trick 2 as likely at all if I duck trick 1 so I think it is a lot better

The reason that this post received recs was because while the danger of a club shift is severe, if the shift is not found, both finesses, 3-3 splits and the safety play in diamonds may be attempted, if W fails to find the 2nd round shift into opener's declared strength and the first finesse by declarer in hearts works. My calculations are that billw's pure line wins 58%, beating the pure safety play line's 56%. The first round duck gambit loses 15% points when W finds shift, and gains 14% points otherwise.

[whereagles]

Definitely not a straightforward hand.

[Zelandakh]

As an aside - Justin, if you were sitting West and ♠Q held with partner discouraging, do you think you would find the shift on this hand?

[BillPatch]

 Zelandakh, on 2014-August-08, 02:37, said:

As an aside - Justin, if you were sitting West and ♠Q held with partner discouraging, do you think you would find the shift on this hand?

AS posted by Diana_Eva, not even Zia would be able to find the switch to Q7 in ♣ rather than continue from the solid ♠ sequence, without a good look at declarer's hand.

[BillPatch]

I overestimated the superiority of the mixed strategy ducking line. Declarer can benefit from the safety play after the duck only if West continues ♠, and the ♥ finesse works. If W switches to a ♥ declarer should postpone the heart finesse and follow the order of the billw's line, also in the very rare chance W shifts to a ♦. If W persists in ♠, declarer takes the ♠ in hand and tests the ♥ finesse; if it succeeds he takes the ♦safety play A♦, low to K♦, low towards the J ♦. If the ♥ finesse fails: take the regular finesse in ♦, if that succeeds test both ♦ and ♠ for 3-3 breaks.

I analyzed that the teacher's line works 56% and billw's line 58%. When West continued ♠ the mixed strategy gained 7% on billw's line, a ♥ or ♦ switch used billw's line, a ♣ switch lost 15% on billw's line.

I simulated the hand on Jack 6 25 times. Jack continued ♠ 17 times, switched to a low heart 6 times, and switched to a club twice. Based on this, the mixed strategy beats billw's line 4.76% and loses to it 1.2% for a net gain of 3.56%. Thus the winning percentage for this strategy equals 61.6%.

[yunling]

 BillPatch, on 2014-August-11, 17:18, said:

I simulated the hand on Jack 6 25 times. Jack continued ♠ 17 times, switched to a low heart 6 times, and switched to a club twice. Based on this, the mixed strategy beats billw's line 4.76% and loses to it 1.2% for a net gain of 3.56%. Thus the total net gain for this strategy equals 61.6%.

I suspect your simulation is not suitable for the calculation here. When west holds 4 or 5 spades, the chance for him to continue a spade goes up. When west only holds 3 of them, in which case you really want a spade continue, he is more likely to switch to a club.

Also, have you considered the card followed by East? Maybe you should use different strategy when he follows ♠2 or ♠8.

[BillPatch]

I found an improvement in the mixed strategy. After the duck, when W switches to a heart, take the free finesse immediately. If it wins, take the safety play in diamonds. If the finesse loses, win in hand and test spades for 3-3 split. If split works, take the safety play in diamonds. If the split fails, take the diamond finesse and if it wins try the diamond 3-3 split. The chance of this line working is 61.5%. (50% * 81%) + (50% * 36% * 81%) + (50% * 72& * 50% * .36%)

Since Jack shifted to a heart 24% and this is a 3.5% improvement on that portion the mixed strategy gains .84% over the previous line, New winning percentage 62.4%.
Edited new winning percentage to actually show an increase.

[BillPatch]

 yunling, on 2014-August-12, 03:29, said:

I suspect your simulation is not suitable for the calculation here. When west holds 4 or 5 spades, the chance for him to continue a spade goes up. When west only holds 3 of them, in which case you really want a spade continue, he is more likely to switch to a club.

Also, have you considered the card followed by East? Maybe you should use different strategy when he follows ♠2 or ♠8.

Observations from performing the simulation. As PhantomSac noted in his first post, it is very difficult for West to find the shift to clubs at trick 2. The critical decision at the table is Will West find the right shift often enough to outweigh the gains by being able to test spades when West doesn't find the right switch. According to PhantomSac's judgment it was so. To put the null argument into figures the critical calculation was Is X(the % of club switch)* 15 >= Y(% of Spade continuation) * 7 + Z(% of heart shifts) * 3.5 ?

From the start of the simulation it was easy to see that regardless of his other 11 cards(Q of spades led J of spades assumed)
the west Jack was unlikely to find the club shift as forecast. Jack's signal from East at trick 1 was uninformative. The obvious shift suit was hearts, so East's main question was which major he wanted. When Jack East had three or four cards in spades, often with the 10, I only saw encouraging signals.

When W did shift to hearts he did not have the ♥Q, but playing for the drop in ♥ is very anti-percentage and also never worked. When W shift to clubs he always had two small in that suit.

Regardless of West's spades or East's signal, West continued spades over the 15 to 7 critical ratio so according to the critical calculation the modified mixed strategy appears to be optimal.

Despite knowing that the critical calculation had been satisfied, and PhantomSac's judgment thus proven again I continued with other calculations for the exercise. Since they were redundant to the task at hand. I see no purpose in extending the simulation to defend them.

[1eyedjack]

I am curious to know what factors were fixed and specified and what factors were left as random variables in the sim that you ran.

[BillPatch]

 1eyedjack, on 2014-August-26, 19:42, said:

I am curious to know what factors were fixed and specified and what factors were left as random variables in the sim that you ran.

Declarer's hand and dummy specified as given. Q and J of ♠ were also specified as being with W based on opening lead. The aucton as specified was given on Jack6.