[42]

Hi!

This suit combination caused some trouble among 3 experts (you see, I was not really involved...):

Contract is 6, reached in an uncontested auction, and the trump suit comes down:

A7xx

Q108x (you must make 3 tricks, the rest is easy).

Do the 8 and 7 make a difference? Normal play with small cards instead of 8 and 7 is A and then low to 10. That is what 1 expert said, the other argued that running the 10 is correct, the 3. was indifferent.
______________________________________________________________
In general:
I find it extremely tiring trying to learn from books how to play a specific suit combination, I cannot really focus for a longer time So I read as much as I can and try to figure out principles.
How did you experts learn this stuff? Did you go the mathematical path and counted out all possible suit layouts? This would take me a lot of time and at the table it is impossible without a piece of paper and a pen....

Thx!
Caren

[inquiry]

It looks on the surface like you have to guess who has what cards... for example, cash Ace and lead low, if RHO plays low do you hook the Ten or play the Queen. But on this one, I would want to protect against four fo this suit with RHO by playing ACE and low to the TEN. If the suit is divided 2-3 with two behind my QT, the second one could be the King or the JACk, so the plays are equal..but I gain (bonus) if the suit is divided 1-4 with small singleton behind my QT...

I don't think on this hand the 8 or 7 makes a difference unless you expect RHO has a singleton for some reason. Then if it was a singleton nine, or jack and you guessed correctly you could hold the suit to one loser... but that is not realistic line.

[luis]

The normal play is to cash the ace and lead low to the ten next.

[cherdano]

Running the 10 seems very wrong. If the Jack is onside (and covers), you still have a guess for 3 tricks; if it is off, you need doubleton king, i.e. a lot of luck.

I have never read lists of suit combinations, and it sounds like a waste of time to me. Instead, completely working out suit combinations yourself is not so hard and once you have done it a couple of times, it gets faster. More importantly is, and I think that is why Fred has been teasing us to work them out many times, that you also learn which holdings which play is catering for. That is after all what matters, since in real life you almost never have pure suit combinations (hints from the bidding, entry restrictions etc.).

Arend

[luis]

I wonder how can somebody claim to be an expert and suggest running the ten here? Life is full of surprises.

[cherdano]

luis, on Feb 7 2006, 07:25 PM, said:

I wonder how can somebody claim to be an expert and suggest running the ten here? Life is full of surprises.

I think that is entirely not surprising. More surprising is that someone suggests running the 10, and Caren calls him an expert!

Sounds like a discussion among 3 "experts", so Caren was indeed not involved.

[han]

In general:
I find it extremely tiring trying to learn from books how to play a specific suit combination, I cannot really focus for a longer time sad.gif So I read as much as I can and try to figure out principles.
How did you experts learn this stuff? Did you go the mathematical path and counted out all possible suit layouts? This would take me a lot of time and at the table it is impossible without a piece of paper and a pen....


I love suit combinations, every time I see a suit combination that I don't know for sure I try to work it out myself. I try to remember them afterwards, but it is not uncommon for me to redo a suit combination, even if I already know the answer. I bet that many experts share this enthousiasm for suit combinations (after all, how can you ever become a true expert if you don't love most aspects of this game), but I'm sure that not all do.

As a really good player you should not only know the best lines for the most common suit combinations, you should also know some alternatives, and roughly how much worse these lines are. At the table you often have extra information or lack in entries or reasons to keep a certain player from getting in.

For example, if you have AJ109xxx in hand and xx in dummy then the best line is to finesse twice. However, cashing the ace first is not a whole lot worse, and if this is the trump suit and there might be a defensive ruff then it could be better to play diamonds from the top. If RHO has made a preempt (or is a passed a passed hand and already took two aces) then this line might also be superior, as KQ doubleton on your left could be more likely then KQx(x) on your right.

That's a suit combination I think I know fairly well, because I have thought about it many times. I still have a long way to go on many others.

[42]

Jlall, on Feb 7 2006, 06:54 PM, said:

BBO experts

*hehe* just a mistake in typing, correct is "expart"
Please don't think that I made it wrong, when they asked ME how the suit should be played

And thx Hannie! Which is the step-by-step WAY to work a suit combination out? Especially when time is money? One takes into account the missing honours and the most likely suit break, or? I would, with a piece of paper and pen, write down all possibilities and do some counting. That is easy when I have 9 cards together, ok with 8 but more difficult with only 7 or less, at least for me Again easier when I need a special combination (both honours onside for example); but when both opps were silent and nothing happened so far, and I need to draw trumps... *phew*

[luis]

42, on Feb 7 2006, 09:38 PM, said:

Jlall, on Feb 7 2006, 06:54 PM, said:

BBO experts

*hehe* just a mistake in typing, correct is "expart"
Please don't think that I made it wrong, when they asked ME how the suit should be played

And thx Hannie! Which is the step-by-step WAY to work a suit combination out? Especially when time is money? One takes into account the missing honours and the most likely suit break, or? I would, with a piece of paper and pen, write down all possibilities and do some counting. That is easy when I have 9 cards together, ok with 8 but more difficult with only 7 or less, at least for me Again easier when I need a special combination (both honours onside for example); but when both opps were silent and nothing happened so far, and I need to draw trumps... *phew*

I think that there are some dozens of combinations that you have to learn, working them at the table would be difficult or even impossible.
For example:
Qxx
K9xx
To get 2 tricks here the correct play is low to the 9, then low to the K (yes, from Qx) this is very hard to work at the table and there are others.
If you are interested you should study the more useful combinations and the rare ones you may need your time to think about other aspects of the play.

Luis

[Stephen Tu]

Hopefully you know enough basic principles of combinations to instantly recognize the 2 or 3 best candidate lines. At that point, you concentrate exclusively on imagining layouts where one line succeeds while the other line would fail. Then you toss out the one that fails in a higher percentage of layouts, keeping mind things like each individual 3-2 break is more common than each 4-1 break.

[han]

Instead of starting with all possible holdings, I usually start with possible lines and calculate for how many holdings that line would work.

Take for example the suit combination that Luis gave:

Qxx

K9xx

His line (low to the nine then low to the king) is not quite complete, you would also need to say how to play when RHO plays the 10 or 9. Let's say that in that case you play the king at trick 1 and play a low card to your 9 the next round (seems best to me at the moment).

The line works for any 3-3 split, so you could start by counting how many 3-3 splits there are (6*5*4/6 = 20). However, any reasonable alternative line will also work for all 3-3 splits, so we don't need to count those splits, they will cancel.

In addition, his line will work when RHO has ace doubleton or singleton but not when RHO has J10 (LHO should duck). There are 6 such holdings. Are there any other holdings? Yes, if RHO has AJ10x(xx) then you will also make 2 tricks. There are 3 AJ10x holdings, also 3 AJ10xx holding and 1 AJ10xxx holding. I don't see any other holdings, so the total number of holdings is (I think) 20 3-3's, 8 4-2's 4 5-1's and only 1 6-0 split for a total of 33 out of 64 holdings. The odds should be much higher than 50% though, as we succeed mostly with balanced distributions.

These holdings are not equally likely but the difference between a 3-3 split and a 4-2 split is not that large. Unless a different play has almost exactly the same number of possibilities you can be sure that the play with the highest number is best.

Luis is right, this is not a combination I would be able to work out at the table.

[FrancesHinden]

42, on Feb 7 2006, 05:45 PM, said:

_____________________________________________________________
In general:
I find it extremely tiring trying to learn from books how to play a specific suit combination, I cannot really focus for a longer time So I read as much as I can and try to figure out principles.
How did you experts learn this stuff? Did you go the mathematical path and counted out all possible suit layouts? This would take me a lot of time and at the table it is impossible without a piece of paper and a pen....

Thx!
Caren

A few of the most common ones you should really just learn. But because they are the most common ones you have a chance of remembering them the second time if you didn't learn them the first... But I would imagine you know most of these anyway.

Some combinations are very hard to work out at the table - the ones that get put here are often the difficult ones where the right answer is not the obvious. But there are a couple of guidelines that work in most cases:

- if the relevant card will drop if the suit is split as evenly as possible, then play for the drop (e.g. 9 cards missing the Queen, 7 cards missing the Jack) otherwise finesse (e.g. 10 cards missing the K, 8 missing the Q)

- When there is only one key card missing, you can usually work out what the best line is by comparing key cases, rather than working out the initial probabilities. Just assume that all relatively even splits are equally likely (they aren't, but you normally end up with the right answer).

To take two examples that I have rarely seen in textbooks but seems to come up quite often at the table: KJxxx opposite Ax in a side-suit, and you need four tricks from the suit, cannot lose a trick, but can afford to ruff one round. There are two possible lines: A, K and a ruff or A and then finesse the second round. To see which is better, we just work out which cases each of them gains on. We can ignore 5-1 and 6-0 because we can never make 4 tricks. So, A, K and a ruff will gain with Qx or Qxx offside. Finesse will gain with Qxxx onside. The ruffing line picks up 2 cases the finesse doesn't, the finesse one, so the ruffing line is better. (They aren't all exactly equal, but it doesn't matter, the conclusion is correct.)

Now suppose we have KJ10xx opposite Ax.
There are three possible lines: A, K and run the Jack; A, K and ruff; first round finesse. Again we can work out which is best step by step:
i) missing the 98 we can't pick up any 5-1 or 6-0 breaks, except singleton Q, so still ignore them
ii) A, K and run the Jack is better than a first round finesse, because while they each pick up half the holdings (Q on the left vs. Q on the right) cashing the AK first picks up Qx in either hand. So forget the first round finesse.
iii) Comparing A, K, run the Jack with A, K ruff: the first gains with Qxxx on our right, the second with Qxx on our left. Everything else is equal. That's one case all, so we don't know the answer. But the difference between these two lines is only 1.6% so it doesn't matter much if you can't decide.

- if you need something for a tie-break, remember that
i) a more even break is always more likely, so Qxx on our left (most even, 3-3) is more likely than Qxxx on our right (less even 2-4) and you should ruff the third round.
ii) an honour card is more likely to be with the length, so Qxxx is more likely than Qx

By the way, very often the difference between two possible good lines is tiny. As long as you are doing something sensible, don't worry about the odd %. (For example, take AJ109xx opposite xxx. There is very little to choose between two finesses and ace and another; the thing to avoid is one finesse and then cash the ace and even that is only marginally worse.)

[kenrexford]

I have found that great gain exists in studying patterns, not necessarily in memorizing odds. It pays to learn the lesser-odds plays well also, just to know them, in case of specific situations.

The classic, of course, if the intrafinesse or the backwards finesse. Odds are against this in a vacuum, but knowing those lines can be invaluable.

Two of my favorite "weird" positions:

Qxx on dummy, A10 in hand. Opponents who do not know what you have precisely, how many tricks you need from this suit, and how many losers you have external to this suit. The "winning" line for two tricks, no losers, is small toward the 10, planning to hook, IMO. Sure, that gives up on dropping the stiff King to the left. However, it gains when RHO does not split (very plausibly) or, even better, if he DOES split, you cross back to dummy and lead another small card, and he suspects Qxx opposite stiff Ace. Tested, it works.

9xx on dummy, Q8xx in hand. You need a trick from this suit without giving up three. Easy step #1 is small toward Queen. If RHO pops 10 or Jack, covered, and lost to King, this seems hopeless. However, you have a fair shot of inducing a smother if RHO has the stiff Ace remaining and LHO has (10/J)xx remaining. Set up LHO's suit (NT), to create GREED, and then lead small toward dummy, pretending to sneak a trick. Your chances increase if you play a suit on the side in such a way as to induce LHO to count your hand wrong and be assured that you have the missing Ace in this suit.

Many other patterns like this exist. Small toward 10xx from xxx wins occasionally, when RHO ducks with KJx and LHO ducks SMOOTHLY with AQxx. Spotting these weird patterns occasionally creates tricks out of air. The simpler non-magic finesse positions, studied well, suggest lines to take when you have knowledge of point spreads, suspicion of stacks, etc., and these are often not in textbooks.

I suspect that abberant plays on the example hand might have been induced by these concerns. The simplest explanation is an assumption that LHO held KJx(x), with the 10 working well. Without seeing the extire hand and entire auction, it seems difficult to assess the play as absurd. For that matter, lead might have suggested this play, perhaps through restricted choice?

[42]

Thank you all, your replies are very helpful for me, especially Frances'. I work for mathematicians and I understand some "principles", but it is still hard Cooking a fancy dinner is much easier
Caren