Today's suit combination

[fred]

A9876

 

Q32

 

You need 4 tricks.

 

This suit combination arose in the Blue Ribbon Pairs in the recent ACBL Nationals in Orlando. I didn't know the answer off the top of my head so I had to try to figure it out (well I didn't have too - I could have just gone with my instincts).

 

Question 1: What do your instincts tell you? No thinking allowed!

 

Question 2: What is the right line of play?

 

I actually figured out the right line of play, took it, and got close to a zero since the line of play that most people (including me) think is instinctive happened to be necessary on the actual lie of the cards (it was a 1NT-3NT auction, a normal opening lead, and there was nothing to the play of the hand except the play of this suit, so my poor score was definitely a result of making the mistake of thinking!).

 

Fred Gitelman

Bridge Base Inc.

www.bridgebase.com

[Free]

1) no thinking: 5 tricks are impossible

 

2) I think cashing the Ace and then low to the Q for 4 tricks

[fred]

Sorry typo - I meant 4 tricks. I will edit the post.

 

Fred Gitelman

Bridge Base Inc.

www.bridgebase.com

[Yzerman]

A9876

 

Q32

 

You need 5 tricks.

 

1 - 5 tricks, Not possible

 

2 - 4 tricks, I think the normal (and percentage) play is small to ace and small to queen. This will win whenever single K in LHO and any Kx or Kxx in RHO (assuming no entry problems).

 

Some friends and myself recently play in regional team game, and we had a similar combination (if not identical) that cause a swing. The result of the swing, was that this combination consume ~ 1 hour of conversation during dinner break. The "common sense" or "logical" thinkers agreed that low to ace and low to queen was correct, however I argued for a different play (however all my f2f partners will tell you I am the master of esoterica). My argument was to play small from dummy, and keep all your options open. You "might" catch RHO sleeping and will play K from Kx, you can still pickup Kxx and Kx in RHO (by playing queen) and you can now pick up Jx (intrafinesse), 10x (intrafinesse), J, and 10 in RHO. I think we decided that small from dummy is technically best cause it preserves many more options (that does NOT mean you WILL get correct though - but at least keep options open).

 

HENCE - If given this problem, i would cross and play low toward Queen, if no King appear I would play low on 1st round then play intrafinesse (pinning honor-x) - STUPID ESOTERIC ME!

 

MAL

[luke warm]

all i can say is, i'm glad no thinking is allowed :P ... i'd play like most people here, ace and small toward queen

[luis]

First Instinct:

Play the Ace and if we only get low cards low to the Q but if the ace picks a T or J finesse the 6 next (Restricted choice)

Without looking I think a very solid alternative is to play low to the queen first if RHO plays a J or T finesse next (restricted choice)

 

The intrafinesse is very tempting as well :-) Ride the 6 and then play the Queen to pin the J or T doubleton in RHO's hand. If Chagas approaches the table and looks interested I'd play the intra-finesse :-)

 

KJT Missing => Luis playing an intrafinesse for down 1 with everybody plus....

 

 

This one was good Fred :-)

 

Luis

[Guest Jlall]

1. leading low to queen is very instinctive.

 

2. The right play is actually the ace first, applying restricted choice.

[inquiry]

I guess my instincts are ACE from dummy and then low up to the queen if I needed four tricks. I think this might be the right play as well unless East plays two low cards on first two rounds of the suit (that is under the ace and then when a low is lead from dummy), then I would duck the second in this suit.

 

The reason for the second round duck, is that KJ or KT doubleton is twice as likely as specific JT doubleton. Of course. West with Jx might or Tx might get cute and drop a high card on first round but then the dummy card wins, and if WEST has KJT i am not winning four tricks anyway. Hope this makes sense.

Ben

[Cascade]

West with Jx might or Tx might get cute and drop a high card on first round but then the dummy card wins

If West has Jx or Tx then East has KTx or KJx and the T or J will pop on the second trick. In this case covering with the Queen is a no lose play since the J or T fell at trick one.

[inquiry]

West with Jx might or Tx might get cute and drop a high card on first round but then the dummy card wins

If West has Jx or Tx then East has KTx or KJx and the T or J will pop on the second trick. In this case covering with the Queen is a no lose play since the J or T fell at trick one.

Which is what i meant with dummy card wins... i just cover if east play honor, duck otherwise.... not worded so well, but i see you got the point.. play H from Hx is not so good.

[sceptic]

if left hand o has 3 honours you cant make any way

 

so finese the 9 ( cover with ace if K is played) if not let it win with 10 or J and play ace when you get lead back hope Kj or K10 is doubleton

[Deanrover]

Instinct says A then low to Q. Will work out real answer today. Pls don't post solution for another 24 hours :P

[flytoox]

Small from dummy, leading to my Q. This will wins, whenever, RHO has K, singleton or doublteont with one of J or T.

[myfish]

my first instinct:ace and if nothing special then small to Q

[Fluffy]

My instict said run the Q for 3/4 tricks, now I really need 4, A then low to the Q without thinking.

 

Thinking a bit more low to the Q to catch singleton 10/J at the right (but losing to doubleton J10), 1 position improvement.

 

In practice intrafinese is very good because few players play low smothly with Kx, so leading the 6 and running is probably the best in practice (sadly on bbo most players think for 5/6 sconds anyway).

 

the cases that matter between intrtafinese and low to Q are:

 

104,105,J4,J5

against

K4,K5,K104,KJ5,K54

 

In thery low to Q seems bettter, but will depend on oppoent's level also.

 

And then also some more cases where you make 2 instead of 3 tricks, wich will matter at MP.

[pclayton]

Thought about this on the drive home last night:

 

Into the rabbit hole we dive................ :) :D

 

____________________________

 

First thoughts:

 

Arbitrarily playing for Kx, or leading the Queen out of hand to try for a pin on round 1, or intrafinesse or anything else seems silly.

 

Cash the Ace; if only small cards appear, continue with card toward to Queen. forget about the odds being 50-50 mid-play; a priori, its more likely to have Kxx than Kx; so follow through with plan and play to queen.

 

If honor appears on left, I can't pick up KJxx or KTxx on my right, so the only relevant combos are KT / KJ / JT. So apply restricted choice and play for the finesse,...er drop.

 

_______________________

 

Second thoughts:

 

Better is low to Queen first; now I can pick up stiff 10 or J on right, versus picking up stiff K on left with cash of Ace first.

_______________________

 

Third thoughts:

 

Whats really weird about this is that the restricted choice with JT (on left) doesn't apply (with play #2 - low to Q), since I'm now playing back toward to the Ace and the JT are popping up! So effectively, I'm losing out on the restricted choice applications of KT / KJ by playing low to the Queen 1st, but I get to pick up the combo of JT on left! But: I'm now LOSING to JT on right by playing low to Queen, since I'm now taking the restrictive choice play of hooking LHO for KJxx or KTxx.

 

Bizarre: TWO restrictive choice combinations with ONE card combo, depending on my choice of plays!

 

So: Cashing the Ace is best with K, KT or KJ (all on left) or JT (on right). Up to Queen is best with T, J (on right) or JT (on left).

 

So, it appears the the Ace is a 4:3 favorite. If only for weird reasons.

 

I can smell something else going on; but I can't quite figure it out. Of course this took me about three tries. Oh, and I'm not taking into the restrictive choice issues of the 4 and 5 spot. :P

 

And I wouldn't be the least bit surprised if Fred said: Sorry; wrong answer.

Edited January 6, 2005 by pclayton

[Stephen Tu]

Both my instincts and the excellent program Suitplay confirm that best line is ace first, then low to Q unless an honor dropped on the left.

[EricK]

Both my instincts and the excellent program Suitplay confirm that best line is ace first, then low to Q unless an honor dropped on the left.

This is very interesting.

 

According to Suitplay, Ace first is the best play for 4 tricks, trying to run the 9 is the 100% play for 3 tricks, but playing up to the Q is the line which maximises your expected number of tricks.

 

So at IMPs, needing 4 tricks you should play the A, and needing 3 tricks you should run the 9. But what about if you need 2 tricks?

 

And at MPS, assuming you are in a normal-looking contract and not doubled and not in desperate need of a top and so on, what line should one take? Does it depend on how many tricks you need for the contract or should one always try for 4 tricks regardless or always go for the maximum expected number regardless?

 

Eric

[Free]

~snip~ But what about if you need 2 tricks? ~snip~

If you have a 100% line for 3 tricks, you'll always have an overtrick. ;) I probably read your post wrong, but I can't see any sensible meaning in that sentence... B)

 

Btw, with this suit combination, there isn't even a line of play which gives you less than 2 tricks anyway, since you can only lose tricks to K, J and T...

[EricK]

~snip~  But what about if you need 2 tricks?  ~snip~

If you have a 100% line for 3 tricks, you'll always have an overtrick. ;) I probably read your post wrong, but I can't see any sensible meaning in that sentence... B)

 

Btw, with this suit combination, there isn't even a line of play which gives you less than 2 tricks anyway, since you can only lose tricks to K, J and T...

The 100% line for the overtrick may not be better than the line which increases your chances of making two overtricks at the risk of making 2 exactly.

 

Eric

[Stephen Tu]

So at IMPs, needing 4 tricks you should play the A, and needing 3 tricks you should run the 9. But what about if you need 2 tricks?

 

A 100% line for 3 tricks is obviously also 100% for 2 tricks. For a problem you'd have to specify how many tricks you can afford to give up before getting those 2 tricks.

 

And at MPS, assuming you are in a normal-looking contract and not doubled and not in desperate need of a top and so on, what line should one take? Does it depend on how many tricks you need for the contract or should one always try for 4 tricks regardless or always go for the maximum expected number regardless?

 

You should take the line that maximizes your matchpoint expectation, which depends on many factors. If you feel you are in a normal contract with practically the entire field and have gotten normal defense, then you should take the "mp-best" line, which is often but not always the "max expected tricks" line. It's not always equivalent, because max expected tricks will be effected by the number of trick difference between two lines on a combination, while "mp-best" is only affected by there being any difference.

 

E.g. the max-trick line of AQT9xx opposite xxx is low to the Q, but mp-best is low to the T (assuming you have entries to pick up the 4-0 split, and no particular reason to discount the possiblity of a 4-0 split).

 

But sometimes you can estimate that there won't be any appreciable MP difference between down 1 / down 2, so if you need 4 tricks you just go ahead and take the 4 trick line, or maybe you think just making it will be a great score so you take the safety play for 3. MP is often a tough game to estimate where you are relative to the field.

[arrows]

I still believe low to Queen is better

 

According to SUITPLAY,

the chances of getting 4 tricks by playing low to Q are roughly

4% less than those of playing Ace first.

 

But the chances of East holding Kx are about 6.8%,

I believe few of opponents will duck the King smoothly

[Free]

Whenever it's not a 100% line, there will always be some holdings which the line will fail. The problem with suit combinations is to get the highest %. Ofcourse you'll lose against Kx in West, but playing low to the Q also loses against any stiff K... Percentages don't lie, A first is best ;)

[fred]

OK I think it's time for the answer. I have never known Suitplay to be wrong before, but either it is here or it is being misused or I am missing something about this suit combination.

 

When I saw this combination at the table, the intuitive play for me was to lead low to the Queen without cashing the Ace first. Evidently most of the field in the Blue Ribbon Pairs felt the same way since the actual layout was:

 

KJxx 10

 

Once the Queen lost to the King, most of these people presumably knew enough about restricted to choice to (correctly) take the 2nd round finesse (since this play picks up stiff J or 10 on your right and loses to J10 doubleton it is roughly a 2 to 1 favorite versus playing for the drop on the second round).

 

So most of the actual field made 3NT by playing this way.

 

I have enough experience with similar combinations to know that the "intrafinesse" is never the % play so I did not consider that. However, I decided to compare my instinctive line to the line of cashing the Ace first. This is what I learned:

 

Cashing the Ace first loses in each of the following layouts:

 

KJxx 10

K10xx J

J10 Kxx

 

Cashing the Ace first gains in each of the following layouts:

 

Kxx J10

K J10xx

KJ 10xx

K10 Jxx

 

If the cases in which LHO has J10, KJ, and K10 seem to be misplaced to you, the reason you should duck the second round after having cashed the Ace on the first round is due to "restricted choice" (gaining in 2 cases, KJ and K10, and losing in 1 case, J10 doubleton).

 

So it is clearly right to cash the Ace first. Seems that the BBO Forums crowd has better instincts than me (and better instincts then the average player in the Blue Ribbon Pairs), but in retrospect I wish I had stuck to my instincts!

 

Fred Gitelman

Bridge Base Inc.

www.bridgebase.com

 

PS If this suit combination discussions are to continue, I think it would be better if people did not post Suitplay output until everyone has had a chance to think about the problems and express their thoughts should they choose to. Being told the answer right away ruins the problem. Besides that, you are doing yourself a disservice by using this program - trying to figure these things out on your own can only be good for your bridge!

[Stephen Tu]

OK I think it's time for the answer. I have never known Suitplay to be wrong before, but either it is here or it is being misused or I am missing something about this suit combination.

 

Suitplay is correct here. Best play for 4 tricks is indeed cash ace. But max-trick & mp-best (if you are in field contract) is low to Q without cashing ace, as it confirms. So the Blue ribbon field is correct to lead low to Q, if most everyone is in the same boat, and there is some reason LHO would be hesitant to lead from 4/5 to the K (e.g. your side bid the suit). But they are wrong if the goal is to make, and down 1/down 2 has little difference. It's often impossible to know at the time which of these situations actually hold. You often only know after seeing the score distribution afterwards ...

 

Cashing the Ace first loses in each of the following layouts:

 

KJxx 10

K10xx J

J10 Kxx

 

It also loses a trick to

KJTx x

KJTxx -

 

which doesn't help you make but might leave you down 2 rather than

down 1. But if you didn't bid this suit, the possibilities of these last holdings are greatly diminished, since opening leader might have led from such a suit. So the MP-best probably is cash the ace, given lack of other information. Suitplay doesn't take into account such inferences, as it operates in a vacuum.

 

So it is clearly right to cash the Ace first

... If you think there is a large MP difference between making / down 1, but small difference between down 1 / down 2, which depends on how much of the field is in the same contract you are. Or if you think there is a reasonably strong inference from the lead that LHO doesn't hold KJTx(x), which is probably correct.

 

Besides that, you are doing yourself a disservice by using this program - trying to figure these things out on your own can only be good for your bridge!

I disagree. One should think through what the answer is first on one's own, and why, but then use the program to double-check your thinking. You may discover you are in error; it's very easy to overlook something.

 

Also sometimes suitplay gives a result or line for a combination that you don't find to be intuitive, it doesn't really explain it for you, but if you think about it you can reason out what the defense is supposed to be doing in terms of falsecards and such, So it doesn't do everything for you, you still have thinking to do.