Splitting by second hand

[gnasher]

If declarer leads a suit and you split or cover from touching honours, what do you think is the normal card from:

 

KQxx

KQJx

KQJ10

QJxx

QJ10x

QJ109

 

Does anyone do something different depending on whether they're covering or splitting?

 

Assume that you don't have to worry about deception or suit-preference.

[Codo]

I always thought that the smalles card is standard, but since some month I play the highest card and surprisingly it works much better for me.

[mikeh]

3+: top, 2 bottom

[whereagles]

I've been trying to find a good rule for this. Trouble is, I think it matters whether you're 1st, 2nd or 3rd to play, and whether the lead is towards dummy or the hand.

 

Anyone got a full set of rules, with motivation?

[hanp]

3+: top, 2 bottom

yes

[pclayton]

3+: top, 2 bottom

I've been playing this and it works very well.

 

Of course, there are many situations where you want to vary your play too.

[cyc0002002]

You may make a second hand high in this suit

      AJ9xx

QTx

        x

 

But it is unreasonable to play K in

      AJ9xx

kTx

         x

when you don't know who is holding Q

 

so in

      AJ9xx

KQx

         x

is better to split Q

 

I believe that's the reason of 2 bottom

[kenrexford]

I have never seen it matter.

[hanp]

You may make a second hand high in this suit

      AJ9xx

      QTx

        x

Yes

 

But it is unreasonable to play K in

      AJ9xx

      kTx

         x

when you don't know who is holding Q

 

No it isn't.

 

so  in

      AJ9xx

      KQx

         x

is better to split Q

 

It is usually better to duck.

 

I believe that's the reason of 2 bottom

 

I doubt that.

[matmat]

I believe that's the reason of 2 bottom

 

I doubt that.

heh. I believe it's a reason for many more than 2 bottoms.

[paulg]

Rules in one partnership are:

• When following to dummy's lead with a sequence, play the second highest of equals (highest if dummy has honour).
  
• When following to declarer's lead with a sequence, play the cheapest if you can beat dummy, and the highest of equals if you are not beating dummy.
  

Like Ken I don't think it makes a huge amount of difference.

 

Paul

[mikeh]

You may make a second hand high in this suit

      AJ9xx

      QTx

        x

 

But it is unreasonable to play K in

      AJ9xx

      kTx

         x

when you don't know who is holding Q

 

so  in

      AJ9xx

      KQx

         x

is better to split Q

 

I believe that's the reason of 2 bottom

Whether you play the K from k10x in front of AJ9xx is a matter for you to decide as soon as dummy hits... because there will be times when it is correct and times when it is not. And whichever is may be, it is important, usually, to play smoothly.

 

As for Q10x or KQx.. (and K10x when contemplating the K), the best strategy depends on how good your declarer is and, oddly enough, how good you think he thinks you are.

 

Advancing players are taught to pop the H from H10x, so that declarer will play them for KQx...and, in order for this to work, they have to play low from KQx... so that declarer will hook the 9.

 

But good declarers 'know' this as well, so when an advanced player plays low, they should usually insert the J, not the 9... because with k10x or Q10x, the advanced defender would have played the high card.

 

Experts know that other experts will vary their play, so we are back to a guess. With K10x or Q10x, play low some of the time and high others, and similarly with KQx...split some of the time and duck some of the time, and when you split, split randomly if you can.. most of us can't.

 

 

Of course, which honour you play from KQx gives rise to restricted choice issues... with K10x you always play the K, if you play a high card, but with KQx you may play either top card. All of this means that optimal strategy is complex... my sense is that one should split KQx less than half the time and play H from H10x somewhat more than half the time, but I confess I haven't studied this as thoroughly as I should. Since we will not have this situation against any particular opps with great regularity, our non-random choices won't really help declarer sense a pattern so long as we do vary our plays.

 

This, btw, is not a situation that falls within the OP intent, I think... because our decision is not intended to convey info to partner but to create uncertainty for declarer.

[JLOL]

If you play UDCA splitting low creates a problem because the 9 could be T9 or 9xx or T9x. If you play the T from T9x then partner doesn't know if you have JT or T9. This could lead to problems. Have you discussed what you play with T9x in these situations with your partner?

 

If you split high then when you have AK you have a problem because the ace might be AK or AQ or ace empty. Do you play the K from both KQ and AK to solve this?

[gnasher]

If you play UDCA splitting low creates a problem because the 9 could be T9 or 9xx or T9x. If you play the T from T9x then partner doesn't know if you have JT or T9. This could lead to problems. Have you discussed what you play with T9x in these situations with your partner?

 

If you split high then when you have AK you have a problem because the ace might be AK or AQ or ace empty. Do you play the K from both KQ and AK to solve this?

I wanted answers, not more questions.

[Mbodell]

I currently play split with the top card of a sequence (2+) when we are playing 2nd to the trick, and play the lowest honor in a sequence when we are 3rd to the trick. I haven't given it much thought about what is best.

[dake50]

I wanted answers, not more questions.

 

*** But the answers SHOULD include these discussions as JLOL.

 

I've been in this same quandary.

UDCA signals in an unblocking situation clouds the card played.

 

*** That needs to be answered in this theoretical discussion,

and surely discussed in any building partnership.

[benlessard]

Ive always play the highest when I play before partner because ive never seen a convincing argument why I should play the 2nd from JTx, QJx, or KQx

 

Maybe it has to do when you play the lowest of equal and expect to win the trick.

 

 

If you jump with K and declarer duck partner might play you from AK. While if you had KQx and play the Q declarer duck is less misleading.

 

With KQJ its a bit unlikely declarer is going to attack the suit