[winkle]

If the team trials has a deal of the tournament, the following deal surely would get strong considerations:

Scoring: IMP

---

W   N   E   S
3♠ 3N 4♠ 5N
P   6♣   P 6♦
PO

This deal occurred early in the final segment of the close semifinal match between Ekeblad and Schwartz. Fred was south. Heavy preemption by the opponents caused Fred to land in the inferior contract of 6♦ on the auction shown.

On the surface it looked like Fred had to lose to the ♥A as well as the guarded ♦Q. But Fred proceeded to play the hand double dummy to make it!

He won the ♣8 lead in dummy, and played the ♦K followed by the ♦9, riding it when east followed small. When the 9 won in dummy, Fred cashed the ♠A pitching the ♣A! He then ran clubs. East ruffed the third round of clubs, but Fred overruffed, drew the last trump, and played a heart to the queen. When hearts were 3-3 he could claim.

At the other table the opponents got to 6♣, so Fred's brilliant play saved the board for team Ekeblad. If Fred had gone down, his team would have lost the match.

So I have the following question for Fred. Can you tell us the thought process you had at the table to come up with this winning play?

[Gerben42]

Really brilliant. After the singleton ♣ lead and two rounds of trumps it is decision time. West is known to have 7(8)♠ and 1♣, and his possible distributions are at this point:

7231 (8131)
7321 (8221)

If West has 7231 everyone can see the winning line: Cash the 3rd round of trump dropping the Q.

Now first you need to be a very good player to see the winning line for the 7321 distribution (the actual one). This also wins in case of 8221 as long as the ♥ doubleton is Ax or T9.

Once Fred had found both lines it was a question of deciding which line to take. Perhaps the actual possibility of the 8221 hand above that made the difference, or it was table presence, or the fact that he was on vugraph and therefore took the more spectacular line, only Fred can tell us that.

Remains to be said then: Extremely well done!

[whereagles]

Nice card reading and excellent technique. Well done.

[luke warm]

incredible hand... there was another one, played by brad, that also amazed me... wish i could remember which one, but i don't even remember the day... probably ben does, it pretty much wowed everyone there

[Fluffy]

Gerben42, on May 24 2005, 10:43 AM, said:

Really brilliant. After the singleton ♣ lead and two rounds of trumps it is decision time. West is known to have 7(8)♠ and 1♣, and his possible distributions are at this point:

7231 (8131)
7321 (8221)

If West has 7231 everyone can see the winning line: Cash the 3rd round of trump dropping the Q.

Now first you need to be a very good player to see the winning line for the 7321 distribution (the actual one). This also wins in case of 8221 as long as the ♥ doubleton is Ax or T9.

Once Fred had found both lines it was a question of deciding which line to take. Perhaps the actual possibility of the 8221 hand above that made the difference, or it was table presence, or the fact that he was on vugraph and therefore took the more spectacular line, only Fred can tell us that.

Remains to be said then: Extremely well done!

I guess 6 card openings would also jam this play, althou the bidding suggests 7-4 or 8-3

[Chamaco]

This hand should be included in the section about "Trump substitutes" of the book "Adventures in card play".... :-)

[fred]

Jlall, on May 24 2005, 01:08 PM, said:

the dude played the 8 of diamonds (even at the highest level it is naive to think that the opponents will always find a subtle falsecard such as that). Say what you want about theory but in practice when you have to guess if they're 7321 or 7231 that should tip the scales.

Justin is correct (as usual) that Berkowitz's play of the 8 of diamonds was the deciding factor for me. However, I suspect I would have got this hand right regardless only because it was a pure guess (based on what I thought was near 100% that Berkowitz had 7 spades and 1 club) and because you might as well take the "fun line" in these situations.

I spent about 5 minutes thinking about this hand before playing a card from dummy. What I was thinking about was playing a diamond to the 9 on the first round. I am pretty sure that this is the correct percentage play. Do you know why I rejected it?

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

[cherdano]

I am not sure how relevant this is but I think Fred's actual line is actually the better percentage play (after running the ♦9).
Given 7 spades and one club with West, the distributions 7321 and 7231 are exactly equally likely. However, Fred's line wins against 2/3 of the 7321 distributions, whereas cashing ♦A after ♦K and running ♦9 wins only against 50% of the 7231 distributions, as both need ♦Q onside.

Quote

However, I suspect I would have got this hand right regardless only because it was a pure guess (based on what I thought was near 100% that Berkowitz had 7 spades and 1 club) and because you might as well take the "fun line" in these situations.

In fact, when watching this hand on vuegraph, I was sure you would take the successful line when you kept the ♣A in hand, so that you could later pitch it on the ♠A

I am surprised but it seems right that low to the 9 is the best percentage play (wins against 33% of 7321 and 50% of 7231, which puts it sufficiently ahead of both lines above that 7141 etc. wont make a difference). Maybe Berkowitz would not have led his singleton with ♦Qxx, however?

[fred]

cherdano, on May 24 2005, 03:25 PM, said:

I am not sure how relevant this is but I think Fred's actual line is actually the better percentage play (after running the ♦9).
Given 7 spades and one club with West, the distributions 7321 and 7231 are exactly equally likely. However, Fred's line wins against 2/3 of the 7321 distributions, whereas cashing ♦A after ♦K and running ♦9 wins only against 50% of the 7231 distributions, as both need ♦Q onside.

Quote

However, I suspect I would have got this hand right regardless only because it was a pure guess (based on what I thought was near 100% that Berkowitz had 7 spades and 1 club) and because you might as well take the "fun line" in these situations.

In fact, when watching this hand on vuegraph, I was sure you would take the successful line when you kept the ♣A in hand, so that you could later pitch it on the ♠A

I am surprised but it seems right that low to the 9 is the best percentage play (wins against 33% of 7321 and 50% of 7231, which puts it sufficiently ahead of both lines above that 7141 etc. wont make a difference). Maybe Berkowitz would not have led his singleton with ♦Qxx, however?

Once the 9 of diamonds held the trick it really was a complete guess as to whether I should play Berkowitz for 7321 or 7231. The reasoning about "2/3 of 7321..." is flawed because at that point it was impossible for Berkowitz to have Qx of diamonds.

Jlall and Cherdano are both correct - I rejected the percentage play because I thought that Berkowitz would be less likely to lead a singleton if he had the Queen of trump.

Getting on a plane for another tournament soon. Bye for now

Fred Gitelman
Bridge Base Inc.
www.bridgebase.com

[cherdano]

fred, on May 24 2005, 04:51 PM, said:

cherdano, on May 24 2005, 03:25 PM, said:

I am not sure how relevant this is but I think Fred's actual line is actually the better percentage play (after running the ♦9).
Given 7 spades and one club with West, the distributions 7321 and 7231 are exactly equally likely. However, Fred's line wins against 2/3 of the 7321 distributions, whereas cashing ♦A after ♦K and running ♦9 wins only against 50% of the 7231 distributions, as both need ♦Q onside.

Once the 9 of diamonds held the trick it really was a complete guess as to whether I should play Berkowitz for 7321 or 7231. The reasoning about "2/3 of 7321..." is flawed because at that point it was impossible for Berkowitz to have Qx of diamonds.

I don't really feel like arguing with a USBF team trials winner, but I still think my reasoning is correct
After the ♦9 held, the holdings ♦Qx and ♦Qxx with Berkowitz are ruled out. That is, 33% of the 7321 and 50% of the 7231 distributions are ruled out.
This means that the a-priori odds of 50-50 between these two distributions (assuming for simplicity that no other distributions are possible) have now shifted to to 57% for 7321 and 43% for 7231.

(In my reasoning above I instead formulated it from the point of view of odds before touching trumps, but the numbers are the same.)

Good luck for your next tournament!
Arend

P.S.: In case anyone is interested in how to calculate this:
We have four cases:
1. 7231 is 50%; among those
1.1. 50% with 7231 with Queen on the right for a total of 25%
1.2. 50% 7231 with Queen on the left for a total of 25%
2. 7321 is 50%; among those
2.1 66.6% with Queen on the right: 33.3%
2.2 33.3% with Queen on the left: 16.7%

Now after the 9 holds, cases 1.2 and 2.2 are ruled out, and only 25% + 33.3% = 58.3% of the original cases remain. Case 2.1 thus has now a likelyhood of 33.3/58.3% = 57.1% among the remaining possible cases.

The principle of this calculation is called "Bayes' theorem", which is also the basis of the principle of restricted choice.