Hand Evaluation on 5332 hands

[blackshoe]

Hand evaluation is more art than science. I get that. I also get that the "rules" of hand evaluation are not really rules, they're guidelines. And I've not run into any problems (that I've noticed) ignoring length points when opening 1NT. Still, it's always bothered me that I don't really understand why, when I have a five card major in a 5-3-3-2 hand, I'm supposed to count one extra point for length (and so the HCP range is 14-16 with these hands), and when I have a five card minor in such a hand, I'm not supposed to count it. Why is that? Or am I totally out to lunch?

[Vampyr]

I am pretty sure that in most jurisdictions, if you frequently vary your NT (in HCP) range from the one which you disclose, you are giving MI.

[sfi]

In terms of hand evaluation there's no real difference between the major and the minor. A reasonable 5 card suit means the hand is stronger than it would otherwise be, so it's worth adding some value, and a point is probably around the right amount (i.e. the 5th card is worth around 1/3 of a trick).

 

I think the approach you are talking about arises from the idea that opening 1NT with a 5 card major is a compromise option. Therefore if you have a 17 count, most of the time you can upgrade because of the 5 card major and get the best of both worlds by rebidding 2NT. I don't tend to see too many people always upgrading balanced 14 counts with a 5 card major - judgement tends to come in much more on this end of the NT spectrum. Good 17 counts with 5 card minors are often upgraded as well. People are less dogmatic about spelling out the rules relating to that, which may explain the apparent discrepancy.

[JLOGIC]

Agree with sfi. You are supposed to upgrade with 17 much more often as a compromise, you overbid a little but ensure finding the major suit fit, you don't want 1N AP and partner to have like 4 trumps and a stiff when you have 17 and 5M. Treating it as 18 is not that far off for NT purposes and you gain finding the fit.

 

With 14 and a 5 card major, I would tend to be less likely to upgrade for the same reason, I'd like to find the major suit fit and even if it's worth a bad 15 I'm not underbidding by that much for the gain.

[lowerline]

I think a 14hcp 5332 needs honour concentration in the 5crd suit, intermediates and not too many quacks before I upgrade it to a 15-17 1nt opening. As already mentioned by others, there are other considerations when the 5crd suit is a major.

 

Steven

[blackshoe]

I am pretty sure that in most jurisdictions, if you frequently vary your NT (in HCP) range from the one which you disclose, you are giving MI.

I suppose that's true, but... if you upgrade good 17s with a five card suit out of 1NT, but don't usually upgrade good 14s with a five card suit into 1NT, how would you write that on the card? How would you announce your range in the ACBL? What I've seen seems to suggest people think upgrading a good 17 with a five card suit is "just bridge" and need not be explicitly disclosed unless specifically asked. Also, how would "giving MI" by not mentioning the possibility cause damage? In what circumstances would it make a difference to what an opponent bids?

 

sfi and JLOGIC make sense. I don't know why I couldn't have thought of those reasons myself. Blind spot, I guess. :blink: :D

[Vampyr]

I suppose that's true, but... if you upgrade good 17s with a five card suit out of 1NT, but don't usually upgrade good 14s with a five card suit into 1NT, how would you write that on the card? How would you announce your range in the ACBL? What I've seen seems to suggest people think upgrading a good 17 with a five card suit is "just bridge" and need not be explicitly disclosed unless specifically asked. Also, how would "giving MI" by not mentioning the possibility cause damage? In what circumstances would it make a difference to what an opponent bids?

 

It could very easily make a difference in the play. I do think that the MI problem is much more of an issue if you are upgrading hands into your 1NT range than if you are upgrading them out of it.

 

In any case I have heard world-class players in the ACBL announce with their range that they frequently upgrade.

 

As for writing it on the card, you would put something like (14)15-17.

[RSClyde]

My simulations have found that balanced hands with 5 card suits aren't particularly stronger than those without for play no trump. Has anyone found an data to the contrary? Those of you who do this, what makes you think that a 14 count with a 5 card suit is worth 15 without one?

 

I understand that discussions like this are difficult because it is impossible to weigh all of the possible ways that a hand can play. Having an extra card is often an extra winner, but sometimes that isn't useful (there's nothing left to pitch). It also means that you have short suits, which means weak spots in your hand for them to attack. Plus there's a loss of flexibility: If I have AKxxx of clubs then it looks great but it's also kind of unilateral, if we're in a 3nt contract that partner stretched to bid because I upgraded, I almost have to make the clubs work and when partner's fit there is poor I have little chance. People also like "honor concentration" in long suits, but those cards can be wasted. AKQJx looks great. But it also ties up 10 of your HCPs. AKQxx frees up a point to go elsewhere in the hand and still plays for 5 tricks much of the time. Even AKxxx might play for 5 tricks and that frees up an entire K!

 

Of course this all just depends on frequency, you could go back and forth and possibilities all day. Which is why having some hard data would be helpful. I'm open to changing my view on this topic, I'd just like something more than intuition.

[RSClyde]

I agree when you have fast controls complimenting your long suit:

I mean

Ax

xxx

Axx

AKQxx

Is a very fine 17 count and I'll agree is too good for 1nt. I mean you basically have 7 tricks in your hand, partner could easily have the goods for game and not find a call over 1nt. But when you start to turn some of those aces into softer cards:

Qx

QJx

Kxx

AKQxx

It kind of feels like I need just as much help from my partner for game as with an 17 count.

[blackshoe]

It could very easily make a difference in the play. I do think that the MI problem is much more of an issue if you are upgrading hands into your 1NT range than if you are upgrading them out of it.

 

In any case I have heard world-class players in the ACBL announce with their range that they frequently upgrade.

 

As for writing it on the card, you would put something like (14)15-17.

If they're adding information to the mandated announcement, technically they're doing it wrong. <shrug>

 

If you're not upgrading into 1NT but are upgrading out of it, that won't work. How do you show the rare upgrades out of the range? I don't think there's a "standard" notation for it, and I can't think of a good one.

[rhm]

Hand evaluation is more art than science. I get that. I also get that the "rules" of hand evaluation are not really rules, they're guidelines. And I've not run into any problems (that I've noticed) ignoring length points when opening 1NT. Still, it's always bothered me that I don't really understand why, when I have a five card major in a 5-3-3-2 hand, I'm supposed to count one extra point for length (and so the HCP range is 14-16 with these hands), and when I have a five card minor in such a hand, I'm not supposed to count it. Why is that? Or am I totally out to lunch?

 

In terms of hand evaluation there's no real difference between the major and the minor. A reasonable 5 card suit means the hand is stronger than it would otherwise be, so it's worth adding some value, and a point is probably around the right amount (i.e. the 5th card is worth around 1/3 of a trick).

 

I think the approach you are talking about arises from the idea that opening 1NT with a 5 card major is a compromise option. Therefore if you have a 17 count, most of the time you can upgrade because of the 5 card major and get the best of both worlds by rebidding 2NT. I don't tend to see too many people always upgrading balanced 14 counts with a 5 card major - judgement tends to come in much more on this end of the NT spectrum. Good 17 counts with 5 card minors are often upgraded as well. People are less dogmatic about spelling out the rules relating to that, which may explain the apparent discrepancy.

 

My simulations have found that balanced hands with 5 card suits aren't particularly stronger than those without for play no trump. Has anyone found an data to the contrary? Those of you who do this, what makes you think that a 14 count with a 5 card suit is worth 15 without one?

 

I understand that discussions like this are difficult because it is impossible to weigh all of the possible ways that a hand can play. Having an extra card is often an extra winner, but sometimes that isn't useful (there's nothing left to pitch). It also means that you have short suits, which means weak spots in your hand for them to attack. Plus there's a loss of flexibility: If I have AKxxx of clubs then it looks great but it's also kind of unilateral, if we're in a 3nt contract that partner stretched to bid because I upgraded, I almost have to make the clubs work and when partner's fit there is poor I have little chance. People also like "honor concentration" in long suits, but those cards can be wasted. AKQJx looks great. But it also ties up 10 of your HCPs. AKQxx frees up a point to go elsewhere in the hand and still plays for 5 tricks much of the time. Even AKxxx might play for 5 tricks and that frees up an entire K!

 

Of course this all just depends on frequency, you could go back and forth and possibilities all day. Which is why having some hard data would be helpful. I'm open to changing my view on this topic, I'd just like something more than intuition.

Contrary to popular believe the fifth card tends to be overjudged. It is worth about as much as a ten and at notrumps 2 HCP are worth about a trick.

Treating an average 17 count with a 5 card major as a balanced 18 is an overbid, but justified by the better chance of finding your major suit fit.

I do not subscribe that hand evaluation is art, neither do I subscribe that hand evaluation should not differ between major and minor. Hand evaluation is not about tricks it is about contracts. It is science.

Even if you do it properly, you will sometimes be wrong since the chances getting to a good contract depends on statistical frequencies.

If my hand has trick potential with a major suit that is different news to an otherwise equivalent hand with a minor suit.

The science is applied correctly, when you are right more often and wrong less often than others.

Why people confuse this with art I do not know. Probably because few people understand probability theory and statistical evidence.

 

Rainer Herrmann

[PhilKing]

My simulations have found that balanced hands with 5 card suits aren't particularly stronger than those without for play no trump. Has anyone found an data to the contrary? Those of you who do this, what makes you think that a 14 count with a 5 card suit is worth 15 without one?

 

The simulations are not bridge:

 

1. Real opponents do not unerringly lead your doubleton - the lead is less likely to be critical when we are, say, 4333. In actual play, the stats show that "declarer advantage" occurs mainly on the opening lead.

 

2. Real declarers don't get 100% of 2-way finesses right or pick the correct suit to establish when they have a 4333 opposite a balanced dummy. In actual play, decent defenders drop fewer tricks after the lead than decent declarers when compared to simulated perfection, and this is particularly true on hands that require good guessing.

 

3. Real opponents occasionally throw the wrong thing when we run our five-card suit - these are the kinds of hands where declarer advantage still applies after the lead.

 

In response to the second question, without wishing to state the bleeding obvious, the fifth card is a potential extra trick.

[sfi]

at notrumps 2 HCP are worth about a trick.

 

On the face of it this statement seems wrong. Otherwise a combined 18 count would expect to give you play for game most of the time. I was going with the simple 40/13 = 3 points/trick calculation, so where does this figure hold?

[Zelandakh]

My simulations have found that balanced hands with 5 card suits aren't particularly stronger than those without for play no trump. Has anyone found an data to the contrary? Those of you who do this, what makes you think that a 14 count with a 5 card suit is worth 15 without one?

A long time ago, some French guy did some analysis and came to the conclusion that the 5 card suit is worth around 0.4hcp. That is close enough to half a point that it is what I personally use. That is, I think a 5 card suit is not enough of itself to upgrade but in combination with the hand being generally above average it is. That said, Justin's tactical considerations go beyond a mere evaluation and are quite likely more important from a practical point of view.

[kenrexford]

A long time ago, some French guy did some analysis and came to the conclusion that the 5 card suit is worth around 0.4hcp. That is close enough to half a point that it is what I personally use. That is, I think a 5 card suit is not enough of itself to upgrade but in combination with the hand being generally above average it is. That said, Justin's tactical considerations go beyond a mere evaluation and are quite likely more important from a practical point of view.

 

Ah, but I bet that Justin would also concede another point. Upgrading depends upon who your CHO is. Having the lead come into the stronger declarer is probably worth the missing 0.6 tricks (or more), eh? LOL

[PhilKing]

A long time ago, some French guy did some analysis and came to the conclusion that the 5 card suit is worth around 0.4hcp. That is close enough to half a point that it is what I personally use. That is, I think a 5 card suit is not enough of itself to upgrade but in combination with the hand being generally above average it is. That said, Justin's tactical considerations go beyond a mere evaluation and are quite likely more important from a practical point of view.

 

But not all five card suits are equal.

 

If you add on a point holding KQJ92 added half with KJ753 but nothing with 65432, you would probably be about right - and it might even match the stats from real play if they were broken down further.

[Zelandakh]

If you add on a point holding KQJ92 added half with KJ753 but nothing with 65432, you would probably be about right

So half a point for a 5 card suit, half a point for good honour texture and minus half a point for very bad honour distribution?

[rhm]

On the face of it this statement seems wrong. Otherwise a combined 18 count would expect to give you play for game most of the time. I was going with the simple 40/13 = 3 points/trick calculation, so where does this figure hold?

 

A combined 21 HCÜP will on average be sufficient for 1NT and 7 tricks.

A combined 23 HCP is sufficient for 8 tricks

A combined 25 HCP is sufficient for 9 tricks and so on.

 

at the edges (slam level or for the first tricks) you need a little bit more for the remaining or first tricks.

 

Rainer Herrmann

[mycroft]

I am happy to upgrade (although I play a weak NT) 11s with decent or better 5 card minor suits. KJTxx, QJTxx, AT9xx, that sort of thing. The kind of suit that if partner has median support, I should eventually set up. I also upgrade A, A, K (and especially A, AK).

 

I don't know if that's enough to be "good 11" or not in announcement (and I've heard "good 14 to flat 17" for the people who play that). I don't know if that's enough to be (11)12-14 on the card.

 

I agree with all, it's difficult to disclose. And it's difficult to disclose effectively for our local National Champion, our local "expects to make the playoffs in the Nationals", our local "never seen a weak NT before", and our local Walruses (A players though they be).

[blackshoe]

And IMO at least disclosure is complicated by the ACBL's "state the range" with examples that only say "X to Y" with no caveats. So is "good 14 to flat 17" proper under this regulation, not withstanding that it gives more information than "14 to 17"? The problem, if there is one, is of course that it not only gives more information to opponents, it also gives more information to partner.

[RSClyde]

The simulations are not bridge:

 

1. Real opponents do not unerringly lead your doubleton - the lead is less likely to be critical when we are, say, 4333. In actual play, the stats show that "declarer advantage" occurs mainly on the opening lead.

 

2. Real declarers don't get 100% of 2-way finesses right or pick the correct suit to establish when they have a 4333 opposite a balanced dummy. In actual play, decent defenders drop fewer tricks after the lead than decent declarers when compared to simulated perfection, and this is particularly true on hands that require good guessing.

 

3. Real opponents occasionally throw the wrong thing when we run our five-card suit - these are the kinds of hands where declarer advantage still applies after the lead.

 

In response to the second question, without wishing to state the bleeding obvious, the fifth card is a potential extra trick.

These things may cause the simulator to be a little off center, but it's not clear that this would be a significant error. The things you mention can be true when you have a 5 card suit as well just to a slightly lesser extent.

 

I'm open to the possibility that upgrading is right, I'd just like a clear reason to believe this beyond something like "all the good player's know..." or "It's always worked for me" or "It's obvious that this hand is worth more than...". I get that the logistics of testing such a thing would be difficult, so like I say I opened minded on this. I do however find it suspicious that the one test I can run doesn't bear out the that upgrading is right.

[rhm]

The simulations are not bridge:

 

1. Real opponents do not unerringly lead your doubleton - the lead is less likely to be critical when we are, say, 4333. In actual play, the stats show that "declarer advantage" occurs mainly on the opening lead.

 

2. Real declarers don't get 100% of 2-way finesses right or pick the correct suit to establish when they have a 4333 opposite a balanced dummy. In actual play, decent defenders drop fewer tricks after the lead than decent declarers when compared to simulated perfection, and this is particularly true on hands that require good guessing.

 

3. Real opponents occasionally throw the wrong thing when we run our five-card suit - these are the kinds of hands where declarer advantage still applies after the lead.

 

In response to the second question, without wishing to state the bleeding obvious, the fifth card is a potential extra trick.

 

These things may cause the simulator to be a little off center, but it's not clear that this would be a significant error. The things you mention can be true when you have a 5 card suit as well just to a slightly lesser extent.

 

I'm open to the possibility that upgrading is right, I'd just like a clear reason to believe this beyond something like "all the good player's know..." or "It's always worked for me" or "It's obvious that this hand is worth more than...". I get that the logistics of testing such a thing would be difficult, so like I say I opened minded on this. I do however find it suspicious that the one test I can run doesn't bear out the that upgrading is right.

Philkings objections highlights the differences between simulation and bridge.

However, these objections do not explain why simulation should be further off base with a five card suit present in declarer's hand.

All simulations I know attach little value to the fifth card.

However, you could interpret Richard Pavliceks analysis at http://www.rpbridge.net/9x42.htm from real play as an indication that long suits are a positive sign and worth close to a full HCP..

 

Rainer Herrmann

[PhilKing]

Philkings objections highlights the differences between simulation and bridge.

However, these objections do not explain why simulation should be further off base with a five card suit present in declarer's hand.

All simulations I know attach little value to the fifth card.

However, you could interpret Richard Pavliceks analysis at http://www.rpbridge.net/9x42.htm from real play as an indication that long suits are a positive sign and worth close to a full HCP..

 

Rainer Herrmann

 

I think the data overstates the case. For instance, the statisitic for bidding 3NT with a combined 23 and 24 count with one five-card suit are not representative - they are just cases where one team decide to bid up with those values, usually for a good reason. If you look through the hands, generally the hand without the five-card suit was bidding up on the basis of good intermediates. Board 1 in the 24 point group and hand 13 in the 23 point data are good examples.

 

As to the difference, I suggest that the opening lead is the main difference, since it is the first blow in what is essentially a sprint race. When we are 4333, the lead is less critical.

[NickRW]

A long time ago, some French guy did some analysis and came to the conclusion that the 5 card suit is worth around 0.4hcp.

 

That's about right (for NT purposes, according to the SIMs I've done).

 

However it shouldn't be forgotten that 4432 hands compared to 4333 hands are also worth about 0.2hcp (also for NT purposes). Further, 4333 compared to both 4432 and 5332, for play suit purposes, is worth at least a full point less (and you don't always end up in NT just because you open NT).

 

So I think people have it the wrong way round: Instead of mentally upgrading the 5332 hands, downgrade the 4333 ones.

 

Nick

[RSClyde]

That's about right (for NT purposes, according to the SIMs I've done).

 

However it shouldn't be forgotten that 4432 hands compared to 4333 hands are also worth about 0.2hcp (also for NT purposes). Further, 4333 compared to both 4432 and 5332, for play suit purposes, is worth at least a full point less (and you don't always end up in NT just because you open NT).

 

So I think people have it the wrong way round: Instead of mentally upgrading the 5332 hands, downgrade the 4333 ones.

 

Nick

I think there is a fallacy that occurs here however. Which hand is better if partner has a long suit to play in? 4333 or 4432? It's not automatically the latter, because that suit is frequently your short one. Yes the 4432 shape is better opposite the 5(+) card spade suit (than the 4333) if spades isn't the short suit, otherwise the 4333 is better.