BBO Discussion Forums: Upgrading - BBO Discussion Forums

Hello, i wanted to ask, how do you decide upon if a hand is worth a upgrade to a higher strength level. Mostly im concerned with balanced hands, because that seems to be the most difficult(for me). I know i can just stick to flat face point count, but sometimes that does not feel right to me. So my question would be-what does it take for you to upgrade a hand to a higher NT point range, despite its face point count?

Hand that got me thinking: AKxx;KQxx;Axx;xx(14-16 and 17-19 NT ranges)

I generally look for at least two of:

1. Good spots
2. A five card suit
3. No wasted honors in short suits

This hand would be 1nt for me in a 14-16 range but AKTx KQ98 Axx xx would upgrade.

 phoenix214, on 2015-February-16, 15:27, said:

Generally speaking, point count evaluation works better for balanced hands than for unbalanced hands. So, all other things being equal, there should be fewer upgrades and downgrades for balanced hands than for unbalanced hands.

That said, the location of your honors and the proportion of aces affects the overall value of your hand. Furthermore, additional "body," i.e., 10s and 9s, which are not included in the point count, may improve the value of your hand.

The easiest things to look for are honors in short suits. QJ doubleton is of very questionable value. Qx of less value still. And Jx is almost valueless. So these holdings are not worth the point count that would otherwise be assigned to them.

If you have 7 cards distributed between 2 suits in this manner: xx AKxxx It is usually worth more than if they were distributed in this manner: AK xxxxx. Furthermore, if the long suit had "body" it would add more value to the hand. AKT9x is much better than AKxxx, while T9xxx is often worth more than xxxxx even though the point count assigns zero to both holdings.

Along the same lines, connected honors, especially in long suits, often produce more tricks than widely scattered honors. The AKxxx referred to in the preceding paragraph is an example of connected honors. KQJxx will often produce four tricks, which is more than most 6 HCP holdings will do.

A 4333 hand usually has less trick taking potential than a 5332 hand, especially if the long suit has some values.

And, since the point count system undervalues Aces and overvalues Queens and Jacks, a hand lacking aces is worth less than a hand rich with aces. But this should not be taken to extremes. Axx Axx Axx Axxx is valued by the point count system as a strong NT opening (indeed, some would even add a point for the possession of all four aces). But without anything behind the aces, it may be difficult to develop additional tricks (the phrase "aces and spaces" is used to describe such a hand).

This is a quick run through of some of the more common upgrade and downgrade situations. With experience, you will learn when a hand is worth more than its point count value and when a hand is worth less than its point count value.

I think the answers have already been given, but an additional factor is consideration of the most suitable contract.

For example, your quoted 16 point hand (and I'm deliberately not giving it "good spots") of AK53, KQ64, A87, 92 with two 4-card majors, is a hand I would rather play in 2M than 1NT. The values are "suit-suitable" rather than "NT suitable". If you open 1NT 14-16, and partner is 6-8 count, you are likely to play there. But if you open 1♣ (or 1♦ if that is your taste) then there is a reasonable chance that you will play in 2♥ or 2♠. Without the majors you would be happy to open 1NT with no upgrade. With both majors I would upgrade.

 fromageGB, on 2015-February-16, 17:13, said:

Almost zero chance for me - my systematic rebid with 17-19 balanced and 4-card support is 3M.

I would probably upgrade this hand anyway, but it's close.

My hand evaluation is complex and based on DD research.
AKxx;KQxx;Axx;xx is closer to 17 than 16, so I would upgrade.

It turns out that AKxx;KQxx;Axx;xx is "only" a good 16 for notrump purposes but in the 17-19 category should you end up in a trump contract.
A good evaluator of bridge hands can be found at http://bridge.thomas...submit=Evaluate

binky-hcp.suit and binky-hcp.nt show the value of a hand in a trump or notrump contract.

Rainer Herrmann

ArtK78 gives some excellent pointers.

I would just add that you need to continually re-evaluate the hand as the auction progresses. A holding such as Qx, which you initially view as being of "questionable value", might become less problematic (and worth a full 2 points) if partner bids the suit strongly. Conversely KQJXX is a holding with "good connected honours" that you may well upgrade; until partner splinters in the suit - when it is reduced to "a lot of wasted values" - now you probably want to down-grade!

Opposition bidding (or passing) can also affect our valuations. Is our KJ9X holding sitting over or under the opponent's suit?

Hand valuation is a continual process and judgement is required at each call.

THanks for the answers. I personally usually look at it similar to the way awm does, but then again after reading some of Kit´s articles on bridgewinners have noticed that he sometimes upgrades even without spots, so figured to try this work on the example hand, because the hand seemed quite good to me, but still some of the time I get these decisions wrong.
Also i usually check my upgrades with KnR, although i do not know if it is such a good idea. The example hand is given as 17.

The main factors for upgrading are shape, honour type, honour placement and honour structure. You could argue that all of the last 3 are the same thing but I am going to split it up for the purposes of this post.

First of all shape, which has been covered above. A decent 5 card suit is a plus, typically worth about half a point (0.4 average); 4333 shape is a very small minus even though it often performs quite well DD.
For honour type I am going to say essentially the opposite of Rainer's binky points - aces are good while unsupported quacks (queens and jacks) are bad. Kings and supported quacks are neutral. Supported tens (and even nines) should also not be underestimated.
By honour placement I am referring to Adam's point 3. AKxx is better than AK bare; KJxx is much better than KJ. I downgrade a full point for some short suit honours.
Finally honour structure, which builds on the previous points. This is the reason why supported quacks get upgraded to neutral instead of being a minus. KJTxx is a good suit; having minors of KJTxx and xxx is generally better than JTxxx and Kxx. This last category is probably the most difficult to use because it is more situational than theothers. For example, if Kxx is our only stopper opposite partner's shortage then we prefer it to xxx in 3NT! Similarly, Kxx in partner's suit is a good holding.

There is another aspect of upgrading that is often overlooked too. Partner has to be in on the scheme if they are to evaluate properly. For example, the OP hand is pretty borderline on whether to upgrade or not. I see it as a bad 17. If partner will downgrade their quacks then it makes upgrading more appealing, since the chances are they hold a few when we have so many aces and kings. The less likely they are to downgrade, the less willing we should be to upgrade this hand type.

Finally, I will give some guidelines that I think are helpful as a starting point. You will probably not use them for long if at all but they at least give an objective evaluation to refer back to when unsure until you develop your own feeling for this.

good 5 card suit: +0.5 (treat a bad 5 card suit as 4 cards; a 6th card is worth at least another 0.5)
non-sgl ace: +0.5 (minus) (so 4.5 total; this overvalues aces a little so add a minus to each)
sgl K/Q/J: -1 (yes, J bare counts as zero here)
other unsupported Q or J: -0.5 (if dbl then add an additional minus)
AJ/KJ: -0.5
"useful" supported 10: up to +0.5 (but just a plus value if in a normal suit)

This is probably a lot less involved and complex than Rainer's scheme but is at least reasonably practical. You will probably find you want to adjust it as you go too and the simple rules provide a framework for adding more subtle adjustements. Finally, you will find you no longer need to think in terms of these rigid numbers but can just look at a hand and feel whether it is right to upgrade or not. I only suggest the purely point count approach for B/I players that need the extra help and can use the rules to stimulate thought about the subject.

 rhm, on 2015-February-17, 04:33, said:

I suspect if this is true. The value for binky-hcp.suit is not evaluated on the same scale as the traditional point count. In fact it is a bit too large.
Consider a fairly normal 10 count which, I think, no one would think of an upgrade:
J432
K432
Q32
A2
The value for binky-hcp.suit is 11.2, which suggests an upgrade.
So if you want to use this method for your hand evaluation, probably you should subtract the value by 1.

In my homegrown evaluation method it will be 7.25 for AKxx, 4.5 for KQxx, 4.5 for Axx, and 0 for 4432 shape so the hand has a total value of 16.25 thus no upgrade.

 yunling, on 2015-February-18, 02:16, said:

Seems clearly wrong if evaluating for NT.

 yunling, on 2015-February-18, 02:16, said:

Hand evaluation for trump contracts is different to notrump contracts.
Some say the same thing by claiming that point count is more accurate for balanced hands.
But there are still differences even for balanced hands, which can be worth significant more in a suit contract than notrump and vice versa.
Realizing this has made my judgement when to look for 3NT in preference to a major suit game better.
For example claiming that aces are worth more and queens less than point count suggests is generally true, but the impact of this fact in suit contract is much higher than at notrumps.
One problem when somebody asks how to evaluate a specific hand as opener is that you do not know whether you will end up playing in notrumps or in a suit contract, because the major determinant is whether you will have a sufficient number of trumps in a major suit and judgement is only a secondary consideration.
The conclusion is that initial hand evaluation should be a compromise between notrumps and suit contracts.
If a hand like above is a good 16 at notrumps but worth 18 at a suit contract it is sensible to upgrade it into the 17-19 category, particularly when holding two four card majors.

One thing which is different in trump contracts is that distribution plays a far larger role, while in notrump distribution's impact is generally small.

The value for

J432
K432
Q32
A2

in a suit contract comes mostly from the value of the doubleton in a trump contract. By the way the DD research I have seen does not distinguish between the above hand and

J876
K876
Q87
A8

The lowest spot card which changes the value of a hand is the 9.

Rainer Herrmann

 rhm, on 2015-February-18, 02:46, said:

While this is true, the value of being able to control when to take the trick in a suit makes the ace more valuable in NT too, especially single dummy. That is less so for Ax and, especially, a bare ace, which relates back to points made earlier about honour location.

I like AWMs list but I also update hands with a good 6 card minor in and out of NT without regard to negative factors.

IMHO, the given hand is worth an upgrade but just barely. Make the shape 4-3-3-3 and I'd never consider upgrading; make the shape 5-3-3-2 and I'd never consider NOT upgrading.

Ugh, I wrote this 2 days ago but it has been lost on the cloud.

There is one aspect nobody is talking about: You want to upgrade/downgrade into 1NT opening more than you want to updrade/downgrade out of it because 1NT is a very good opening for natural systems (all systems actually).

On the other hand 2NT opening is very bad in natural systems for slam purposes, so you should rather rebid 2NT when in doubt. This is not true when partner is a passed hand though.

 Fluffy, on 2015-February-19, 03:10, said:

On the contrary, as opening 1NT is bad for natural systems. Playing matchpoints it is vital to get to the major part-scores when there are any, and only then if not, play in NT. If you open 1NT, partner is pre-empted and will not be able to speak unless he is at least invitational. Conversely, opening a minor enables a weakish partner to speak, and then you can play in a major.

Playing matchpoints you want to open 1♣ more often, so if you have a 3-point NT range, up/downgrade out of it as much as you can. Better to adopt a 2-point range, such as 15/16, and then when you are forced to open 1NT and partner has a 7/8 count he can try for a major fit, and if there is not one, subside in 2NT to play.

Of course internationals play IMPs where such nuances are lost, and at this scoring part-score differences of 10 can be ignored. Different game; different methods.

There is a bit of swings and roundabouts about it. Opening 1NT makes it harder for us to find 4-4 major fits on marginal hands but easier to find 3-5 fits and harder for the opps to find their fit. Conversely a 1m opening is great for our 4-4 fits but also makes it a lot easier for them to compete, which they will at MP. I would much sooner be playing a 3 (or 4) weak NT range than a 2 point strong NT as the latter puts too much pressure on the rest of the system (yes I am aware a weak NT has other issues and also adds pressure on the system in comparison to a 3 point strong NT).

 rhm, on 2015-February-18, 02:46, said:

No I'm not talking about upgrade or downgrade on a specific hand, I'm explaining about the nature of this evaluator.
If you put a lot of hands into the evaluator, you will find that binky-hcp.suit can never be lower than binky-hcp.nt, thus "every hand values more in suit than NT", which is clearly not the way we play everyday.

When we say 25HCP=Game, the game can be 3NT or 4M, so it can be used for "general purpose". It is not the case for binky points.

A certain number of binky points, adding two hands together, is correlated to a certain number of tricks taken. So while 25 binky-hcp.nt=3NT, with 25 binky-hcp.suit, you are only expected to win 9 tricks in a suit contract. To make 4M you need something like 27.

Thus, the value for binky-hcp.suit cannot be directly used to adjust our "general purpose" evaluation, unless you do a subtraction.

 mgoetze, on 2015-February-18, 02:20, said:

It's not quite exact——to be more precise, it's something like KQxx=4.6 and Axx=4.4, but we can't be so exact during the play——but it is not "clearly wrong" either.
Even in NT, A is undervalued and touching honors is always a negative.
Besides, opening 1NT does not mean that you belong to NT contracts. You need a mixed strategy for both NT and suit contracts.