[straube]

We use queen points, also called relay points.

A=3, K=2, Q=1 but stiff K=1

I'm often disappointed to learn pd has a stiff K. We're up pretty high and I've usually based my exploration
on the likelihood that partner has a more useful card. Same thing though for a stiff ace. I'd usually rather
learn that patient has those 3 queen points elsewhere.

So my question is whether anyone has tried counting a stiff ace as 2 QPs instead of 3. The stiff ace is still useful, but not as useful as an ace in a longer suit.

Also, I would think that pd could work out whether the stiff honor is the A or K and work out the other cards.
Or can he? Perhaps the difference of only 1 QP would create difficulties.

Opinions?

[Free]

I rarely had problems when counting a stiff Ace as 3 QP's and not counting stiff Kings and Queens. If a stiff King is useful, then usually low level RKC would work better, so you made the mistake of asking QP's. If you don't need the stiff King, then it's a nice bonus and a surprise for everyone.

The biggest drawback of this is that you can have a gigantic number of HCP's and you're representing the hand much weaker because of the QP's. But in my experience you can usually set things straight later in the auction.

[straube]

 Free, on 2012-October-22, 13:29, said:

I rarely had problems when counting a stiff Ace as 3 QP's and not counting stiff Kings and Queens. If a stiff King is useful, then usually low level RKC would work better, so you made the mistake of asking QP's. If you don't need the stiff King, then it's a nice bonus and a surprise for everyone.

The biggest drawback of this is that you can have a gigantic number of HCP's and you're representing the hand much weaker because of the QP's. But in my experience you can usually set things straight later in the auction.

Well that's a thought not counting the stiff king. We do have RKC available and I've never found the stiff king to be an unexpected positive. I still am interested in counting the stiff ace as 2. It just doesn't pull its weight.

awm, if you're around, what do you think?

[awm]

My first reaction is that this is a bad idea. There are many auctions where you can't use an RKC ask in partner's singleton (either because there was a zoom, or because you wanted to resolve shape and it resolved too high, or because you are worried about controls in some other suit too, or because you don't want that suit to be trump but still need help there). If you are skipping stiff kings entirely (or even stiff queens) you're in bad shape on these hands. Of course, showing two different honors the same way is bad (i.e. if A=2 and K=2 also its bad).

This style also makes partner's job in deciding whether to relay quite a bit more difficult, because certain shortcuts like subtracting the total RP from 24 no longer work. In some sense this may not matter, but I find that mistakes (and/or time pressure if you double check all your computations) are not unusual when playing a relay system.

Of course, you could always try a large number of example hands in bidding practice and see what happens.

[straube]

 awm, on 2012-October-22, 19:16, said:

My first reaction is that this is a bad idea. There are many auctions where you can't use an RKC ask in partner's singleton (either because there was a zoom, or because you wanted to resolve shape and it resolved too high, or because you are worried about controls in some other suit too, or because you don't want that suit to be trump but still need help there). If you are skipping stiff kings entirely (or even stiff queens) you're in bad shape on these hands. Of course, showing two different honors the same way is bad (i.e. if A=2 and K=2 also its bad).

This style also makes partner's job in deciding whether to relay quite a bit more difficult, because certain shortcuts like subtracting the total RP from 24 no longer work. In some sense this may not matter, but I find that mistakes (and/or time pressure if you double check all your computations) are not unusual when playing a relay system.

Of course, you could always try a large number of example hands in bidding practice and see what happens.

Curious, but do you count stiff Qs as 1 in your total QP count? And do you stop when you have a stiff queen (as well as a stiff king or queen)? I suppose one could count the stiff queen as 0 but stop to show a stiff honor. In many cases partner could work out that you have a stiff queen (like obviously if he is holding both the ace and the king). In other cases, partner might get confused.

[awm]

 straube, on 2012-October-22, 23:01, said:

Curious, but do you count stiff Qs as 1 in your total QP count? And do you stop when you have a stiff queen (as well as a stiff king or queen)? I suppose one could count the stiff queen as 0 but stop to show a stiff honor. In many cases partner could work out that you have a stiff queen (like obviously if he is holding both the ace and the king). In other cases, partner might get confused.

Yes, I count stiff Q exactly like any other queen. Remember that partner knows my total QP count, so he can often infer my holding via knowledge of other suits.

[straube]

 awm, on 2012-October-22, 23:25, said:

Yes, I count stiff Q exactly like any other queen. Remember that partner knows my total QP count, so he can often infer my holding via knowledge of other suits.

So you must count the stiff K as 2. Somehow I thought you counted it as 1 which is what we've been doing. Well, you get a more complete picture of the hand, and there have been times where we missed a slam by not knowing about a stiff Q. OTOH, we just get so high before we realize partner has a stiff and unwanted honor that it's not always possible to stop shy of slam.

If we counted stiff A=2, stiff K=1, and stiff Q=0 but stopped to show a top honor, do you think we could disambiguate the hand? I'm feeling that stiff honors deserve a severe downgrade.

Side note, but we've had some success using RKC in the order of clubs/diamonds/hearts/spades as opposed to longest to shortest. PCB works as well or better usually for the majors, but we usually can't disambiguate the hand before risking going past 5m.

[akhare]

I would draw the line at stiff K=1 and leave it there. IMO, discounting stiff Aces simply isn't worth the mental effort and the hands on which it matters for the better are likely far and few in between.

[Zelandakh]

I treat a stiff ace as an ace and a stiff king as a queen within an AK point method and see no reason why that would not work perfectly well in an AKQ point method. There was a thread here a while back where Cascade used simulations to suggest that a stiff ace was worth approximately 1/2 a hcp less than a small singleton; something over 1 hcp worse for a king; something just under 1 hcp for a queen; and nearly 3/4 hcp worse for a jack, albeit under a restricted set of hands. To the nearest integer, that would correspond to a QP evaluation of A = 3, K = 1, Q = 0 for singletons.

[rbforster]

 straube, on 2012-October-22, 23:01, said:

And do you stop when you have a stiff queen (as well as a stiff king or queen)? .

I don't scan singleton suits at all. You just count the stiff honor at full QPs and partner figures it out eventually.

[awm]

 rbforster, on 2012-October-26, 23:21, said:

I don't scan singleton suits at all. You just count the stiff honor at full QPs and partner figures it out eventually.

This works fairly well if you are scanning for controls, but I've found it doesn't work at all for QPs. The problem is that partner will often have ace-only in a particular suit and needs you to hold the king, and cannot distinguish between your holding that queen + stiff queen versus you holding the king (and small stiff). This comes up a lot more than the equivalent ace vs. king+king with controls, because if you're missing enough cards that partner has trouble distinguishing ace from two kings you probably don't have slam anyway.

[DinDIP]

 straube, on 2012-October-22, 12:31, said:

We use queen points, also called relay points.

A=3, K=2, Q=1 but stiff K=1

<snip>

So my question is whether anyone has tried counting a stiff ace as 2 QPs instead of 3. The stiff ace is still useful, but not as useful as an ace in a longer suit.

Bo-Yin Yang recommends discounting all singleton honours by 1QP (although this scale doesn't actually appear in Terrorist Moscito) and this was the scale used by GIB when it played Moscito. My testing suggests that, while a singleton ace is sometimes worth less than an ace elsewhere, as others have noted, there is a big difference on many deals between an ace and a king. So, while discounting a singleton ace is accurate for hand evaluation purposes (although probably not by one-third of its value), it's inaccurate for control purposes on a sufficient proportion of hands. This contrasts with both a kingleton and a singleton queen, which are clearly worth 1QP less than their normal on a very large (in the case of a queen, very, very large) percentage of hands.

As to the question of whether it's best to scan singletons or not: I'm unsure. I think it depends in part on the method of scanning the system is using.

David