How to calculate the distributive strength of the hand?

[Zelandakh]

Even untrained monkey would have noticed that the method LTC is not true. I do not know about you. :P

Then please show us how your method evaluates some hands. Take a basic ♠Axxx ♥Kxx ♦Qxx ♣Jxx. How many points for high cards and shape? Then please do the same for 4432, 5332, 4441, 5422, 5431, 5440, 5521 and 5530 hands. That should get us somewhere in comparing whether the MLTC is more or less accurate than your idea. I am not willing to work through your nunbers and do this for myself so if you want some constructive input you are going to have to do the legwork.

[gwnn]

Even untrained monkey would have noticed that the method LTC is not true. I do not know about you. :P

I'm trying to keep my replies short to make it easier for people to read them and to avoid misunderstanding. I regularly fail at this aim unfortunately. So let me try to explain what my point was. As opposed to you, I did not insult you when I brought up monkeys, be they trained or untrained (so let's not go down this rabbit hole where you feel like I insulted you so you reply by a bigger insult to which I'm supposed to reply by an even bigger one, etc - I am not really interested in that stuff, and I trust you are also not). It was not an insult to you or your method, but a compliment to the simple LTC formula (it is so simple that a monkey could apply it). My point was that the formula I gave (LTC_max=min(L1,3)+min(L2,3)+min(L3,3)+min(L4,3)) is equivalent to LTC's basic underlying principle and also equivalent to your formula except for 10-card and longer suits. My point was also that the LTC formula is easier to apply than yours. It's not even really a formula, it's just something you can do with your eyes when you look at your cards (count all the cards that are in the top 3 of any suit). You don't need to worry about P2 and P3, you just look at the top 3 cards of suits and count them. You don't need to know how to add, subtract, anything. You just need to know how to count to at most 12: 1, 2, 3, ... Let me illustrate how a trained monkey would do it. Let's say we have 4432, in that order.

 

xxxx

xxxx

xxx

xx

 

Now you take any card that is in the top 3 of that suit and capitalise it:

 

XXXx

XXXx

XXX

XX

 

11 capital X=> 11 losers. Your method will give 19-8=11, equivalent. Great. Let's check a 3-suiter where you need to apply your correction:

 

xxxxx

xxxx

xxxx

-

 

Enter the chimp:

XXXxx

XXXx

XXXx

-

 

9 losers. You give 19-9-1=9, the same result.

 

Let's check one where we have S2=2, say, a 9-2-2-0:

xxxxxxxxx

xx

xx

-

 

XXXxxxxxx

XX

XX

-

 

7 losers. You give 19-11-1=7.

 

For any distribution, except 10-card suits or longer, the LTC formula will be equivalent to yours.

 

If you don't think so, bring me a counter example. Do not compare me to various animals in the zoo. Do try to read my post this time.

 

I did not say anything whether I agree or disagree with LTC. It may well be that the LTC is very deeply flawed. But then your formula will be equally flawed (except the 10-carders, as noted above).

[mikeh]

I can hardly believe that this thread has persisted for so long. Take another look at the OP's ideas and ask yourself, seriously, would ANY good player ever even consider applying this sort of valuation at the table? Ever?

 

I think it reasonable to infer that the OP is not a good player. That is not meant as an insult. The OP may, for all I know, have the talent to become a good player, so I am not commenting upon inherent traits or abilities. Becoming a good player requires time and exposure to the game as played and understood by already-good players.

 

I have read widely. I have had the pleasure of playing with, as partners and teammates, some truly world class players, and have discussed bridge with them. I have had coaching from Kokish and Fred. I don't say that to stake any claim to personal status or authority, but merely to establish the fact that I do have some understanding of how good players think at the table, even if I can't match them myself.

 

I cannot imagine any expert using the OP method. Given that it adds a layer of complexity to what is actually a pretty simple aspect of hand valuation why should any advancing player want to use it? What edge does it give?

 

As gwnn points out, the method is generally equivalent to LTC, which is incredibly simple, and a method I commend to all improving players as a minor part of their hand evaluation tool kit. So the OP method claims to yield more accurate valuations when we hold a 10 card suit...so what? How often do you get one, and how much difference do you think using the OP method would make to how you bid the hand? Let me guess: very, very rarely and none, respectively.

 

People who create this sort of valuation method are not merely seeing the trees while missing the forest: they are focusing on the twigs when they should be considering the entire ecosystem.

 

Unfortunately bridge players, and especially those of us who frequent online discussion groups tend to fall into rabbit holes and then have a difficult time recognizing that that is where we are :D

[Zelandakh]

Mike, you are obviously right with respect to the Lmax part of his idea and any sensible person would use the method you described rather than the posted formula. I did not want to get into that and essentially lost interest in following the "maths" as soon as I saw this. Instead I wanted to concentrate on practicalities, on how this method is going to evaluate a hand in practise. By doing this it should be easier to show the OP why the method is equivalent to one of the alternatives or where it falls down in comparison. Of course I might be surprised and even learn something!

 

We BBFers are often pretty hard on new posters here. Most that come with new ideas have missed something obvious and get ridiculed for that. Unfortunately in many cases that leads them to becoming trollish rather than constructive posters, or just leaving completely. So forgive me if I am being over-tolerant - I would like the OP to feel welcome here and that the idea was at least taken seriously and understood. Perhaps by understanding why it does not work they will improve and come back with something better, or just become a better player. Or even if not, a productive forum member.

 

It was pointed out recently that the forums are in something of a decline. So let us try to make an effort to make new members feel welcome, even those we do not agree with. Otherwise we may as well all give up and trudge on over to BW.

[jogs]

 

People who create this sort of valuation method are not merely seeing the trees while missing the forest: they are focusing on the twigs when they should be considering the entire ecosystem.

 

Unfortunately bridge players, and especially those of us who frequent online discussion groups tend to fall into rabbit holes and then have a difficult time recognizing that that is where we are :D

 

I never understood why anyone could care about the initial precise valuation of a hand. When the hand is sorted, is it worth opening and if so, what should the opening bid be? The worth of the hand isn't static. It changes as we learn more about the other hands during the auction.

[Fluffy]

I can hardly believe that this thread has persisted for so long.

Because the OP keeps bumping it, adding posts with just a dot or adding nothing, sometimes using a fake acocunt for it. But the moderators won't do anything about it.

[MrAce]

Because the OP keeps bumping it, adding posts with just a dot or adding nothing, sometimes using a fake acocunt for it. But the moderators won't do anything about it.

 

Which poster are you suspicious of using a fake account? As far as I see all the replies came from legit accounts. Do you mean gergana and lovera and lurpoa are same people?

[Fluffy]

yeah lovera, come on, a user with 0 posts that uses the same dot post to bump?, I don't believe in coincidences

[diana_eva]

yeah lovera, come on, a user with 0 posts that uses the same dot post to bump?, I don't believe in coincidences

 

There was a problem with double posts in the forums, that's why there are many empty posts. Lovera is a legit user, she posts in news comments too, since before this thread existed.

[Lovera]

There was a problem with double posts in the forums, that's why there are many empty posts. Lovera is a legit user, she posts in news comments too, since before this thread existed.

1) I have posted initially in italian but acttually i write in english

2) i am male (my name is Carlo)

3) some times i have posted and then cancelled any post that then rested empty.

[MrAce]

Lovera, is legit.

[Lovera]

See post 12 in hand evaluation.

[Lovera]

As you may see in the post indicate i don't use Courtenay losing trick count except in two cases : 2 club open bidding (see my post in Counting Quick Tricks or Difensive Tricks to know i apply) and barrage declaration (considering Ace and King only winners..). So this is my advice : See first if your deal present three colors with 2 cards or less (from 2-2-2 to 0-0-0) or we have a tricolor distribution (4-4-4-1 or 5-4-4-0). These deals are 12 and have a factor of correction of -1 so started in this case with 18 (=19-1) to betray the two longest

suits ( such as 18 - 9 in 7-2-2-2 is 9 losers (= 4x2+1 losers Courtenay ). The last five deals: 10-1-1-1 already defeat (17) whilest 11, 12 and 13 are rare (from 0,0002 to 0,000000006299 per cent) so not to heavy the algoritm them are ignored . All the oterhs deals are more common and have 19 data to calcolate the losers (such as 19 - 9 in 5-4-3-1 is 10 = 3+3+3+1). This in fact is another way to determinate losers (all the deals have 0 winners if any winner is your must detray this number of winner to have laser of your hand ) .Bye .

[Fluffy]

another bump for this useless thread, that shouldn't even be on this forum, will moerators ever do anything?

[diana_eva]

another bump for this useless thread, that shouldn't even be on this forum, will moerators ever do anything?

 

Ben already addressed this in the moderators' thread. None of us will over-rule Ben obv.

[gergana85]

So let me try to explain what my point was. As opposed to you, I did not insult you when I brought up monkeys.....

I accept your explanation of the trained monkey, but with some reservations. The reasons are 3:

 

- I see that you understand that the example of the zoo insult me, but you are continuing with the same examples;

- I accept that it is normal a person unintentionally to afflict someone without any intentions, but I also think it is okay to apologize. You do not apologize;

- Following your logic why you insulted by my example with an animal from the zoo after you put me below level of the trained monkey and I like such (an untrained monkey) answered you? Why you not take my answer as a compliment to your exact definition?

 

I want to say that in the future, when someone passes certain limit against me, he will receive immediate and equivalent reaction in the opposite direction. Until now, I did not allow myself this, even though I had enough reasons.

I believe that the case is closed. Nevertheless, I am glad your questions (they are important for me) and I begin to answer them.

 

1. I never say that I created a method for а general evaluation of the hand (as LTC). I just tried to find the factors affecting on the strength distribution and found (and proved) that it depends of the sum of the two longest suits. It would be possible to imagine, but it has not been proven until now. I determined (and proved) that in some rare cases Lmax is influenced not only by the sum of the lengths of the two longest suits but there is an additional dependency by the lengths of the each three longest suits. Nothing more.

 

2. Continuing in this way, I do not reject the use of LTC. But it must undergo known corrections. Yes, LTC method gives correct results, but as you has noticed only for distributions with no more than 10-card suit. According to LTC, in the distribution 13-0-0-0 has 3 losers, which is obviously not true. My formula is true for all 39 possible distributions.

 

3. What do you find complicated in the formula:

 

Lmax = 19 S_1,2 (P₁ P₂ P₃)

 

This formula in more than 96% of cases is confined to the formula (when talking about percentages, I mean the probability to get such distribution):

 

Lmax = 19 S_1,2

 

To argue that this is difficult is not correct. It is no more difficult than using the formula:

 

LTC_max=min(L1,3)+min(L2,3)+min(L3,3)+min(L4,3)

 

Yes, this formula does the same job but it is not part of the method LTC. According to ypur formula, the maximum number of the losers is not equal to a constant and that each distribution has a different value of Lmax. I say the same. But LTC says otherwise. According to this method always Lmax is equal to 12.

4. Not every player uses the LTC. But each player evaluates the distributive strength of the hand. And every one of them must know the factors affecting on it. This also applies to those who use LTC.

 

Pavel Bogev from Bulgaria. My nick in BBO is gergana85. Sorry for my bad English but.... no one here knows Bulgarian.

[gwnn]

I accept your explanation of the trained monkey, but with some reservations. The reasons are 3:

 

- I see that you understand that the example of the zoo insult me, but you are continuing with the same examples;

But it is not an insult to you. I explained why not and you accepted it. You clearly got insulted but that was not my intention and there was no reason other than you not reading my post properly.

- I accept that it is normal a person unintentionally to afflict someone without any intentions, but I also think it is okay to apologize. You do not apologize;

No, I didn't, because I did not insult you. I cannot apologise for you automatically taking any mention of a "trained monkey" in a post replying to you as an insult. I always assume people read and understand my posts before replying to them.

 

Emphasis mine:

- Following your logic why you insulted by my example with an animal from the zoo after you put me below level of the trained monkey and I like such (an untrained monkey) answered you? Why you not take my answer as a compliment to your “exact” definition?

You still don't understand. Saying "even an untrained monkey can apply method A. That is not true for the equivalent method B." is not putting anyone under the level of the monkey. It puts method B under method A, since it is easier to apply.

 

I want to say that in the future, when someone passes certain limit against me, he will receive immediate and equivalent reaction in the opposite direction.

And isn't this a pity? Wouldn't it be better if you stopped and read my purported insult before you decide that you need to retaliate? There is always a turning point in any internet (or real-life) discussion when replies to opposing arguments turns into reactions to opposing insults. Wouldn't we be better served if we tried to delay this moment as much as possible? For example, by first establishing whether our feeling offended is actually based on anything other than paranoia?

 

The reason why

LTC_max=min(x1,3)+min(x2,3)+min(x3,3)+min(x4,3) is much simpler than your formula is because you don't need to do any maths. The only function here is addition and the minimum function, which is extremely easy to apply visually: you just count the top 3 cards of a suit, no need to count the total length, sort the lengths, subtract from 19. You just look at your hand after sorting them by suits. You may say that your version is also simple, I tend to agree, except that there are already:

-rule of 11 for 4-th best leads

-rule of 10 or 12 for 3/5th best leads

-rule of 19/20/21 for openings

-rule of 1/2/4, or whichever you prefer, for preempts

and a bunch more that I can't recall now. So your formula adds another numerical constant, 19, and people will find it a bit confusing, I think.

 

Of course LTC proper will never give AKQJT98765432 - - - as a 3-loser hand. It will give it 0, since it also accounts for A/K/Q. It will give wrong answers for holdings like AKJT98765432 2 - - where it would give 2 losers instead of one, but hopefully you can see how a) this hand never comes up and b) hopefully anyone can realise how this hand is worth at least slam.

 

You also say that according to LTC, the maximum amount of losers of any distribution is 12. I don't know where you got this from. LTC also considers any 4-th and subsequent card to be a winner. No one who applies LTC to a 5431 will ever get a higher value than 10.

[gergana85]

As you may see in the post indicate i don't use Courtenay losing trick count except in two cases : 2 club open bidding (see my post in Counting Quick Tricks or Difensive Tricks to know i apply) and barrage declaration (considering Ace and King only winners..). So this is my advice : See first if your deal present three colors with 2 cards or less (from 2-2-2 to 0-0-0) or we have a tricolor distribution (4-4-4-1 or 5-4-4-0). These deals are 12 and have a factor of correction of -1 so started in this case with 18 (=19-1) to betray the two longest suits ( such as 18 - 9 in 7-2-2-2 is 9 losers (= 4x2+1 losers Courtenay ).....

I say the same. Hands in which three of the suits have 2 or less cards must make an adjustment. In this case, the length of the longest second suit (S2) is determinant. If it is equal to 2 (S₂ = 2, distributions 7-2-2-2, 8-2-2-1, 9-2-2-0, 9-2-1-1, 10-2-1 - 0 and 11-2-0-0) correction is -1. But if the second longest suit is only 1 card (S₂ = 1, distributions 10-1-1-1, 11-1-1-0 and 12-1-0-0) or 0 cards (S₂ = 0 , distribution 13-0-0-0), the correction is -2, or -3 respectively. For 3-suit distributions (4-4-4-1 and 5440) correction is always -1. Yes, the chance to get a hand with longest suit with 9 or more cards is almost zero and these cases can be ignored. Furthermore, the evaluation of such hands is very easy and requires no formulas.

Pavel Bogev, Bulgaria

[Lovera]

In fact i have tried to clarify better and in more simple explanation what have you said. Fourthemore the subseguent develop for distribution from 10-1-1-1 seems like Goren valutation for short suits: with 0 cards +3 points , with 1 card + 2 points , with 2 card +1 points extended to three short colors : with 2-2-2 until 2-0-0 -1 with 1-1-1 until 1-0-0 -2 with 0-0-0 -3.Bye .

[gergana85]

 

LTC_max=min(x1,3)+min(x2,3)+min(x3,3)+min(x4,3) is much simpler than your formula is because you don't need to do any maths.... And so on.

 

For me, the theme of the monkeys is complete and I will not more to comment on it. You have arguments, me too. They do not fully coincide. OK. Such is life.

I don't agree with your argument that with LTC no need to count, but only to watch. According LTC you need to count the number of the three top honors in all suits and then removing them from 12. Based on your reply, your partner does the same, but he subtract of 24 (to determine the level of contract). And I repeat - you must remember that I do not give a method for evaluation. I want only to show what are the factors that influence and how they affect on the distribution strength. I dont mean nothing more. You ask me why you need to take a new formula when there are so many rules and the introduction of another would confuse the players. Well, some of those are unlikely to help in the evaluation of hand. On the other hand the new formula replaces some of them (e.g. rule 19/20/21 and rule 1/2/3).

You say that there is no information that LTC assumed that Lmax is always 12. Here are some sources:

 

http://www.bridgeguys.com/pdf/LosingTrickCount.pdf.

http://www.bridgehands.com/L/Losing_Trick_Count.htm

http://www.bridge-forum.com/i-trick-1b.htm

https://www.mrbridge.co.uk/library/LTC_2.pdf

http://www.phillipalderbridge.com/LTC.HTM

 

There are many others. Just log into the internet.

 

Pavel Bogev from Bulgaria (nick gergana85)

[gergana85]

In fact i have tried to clarify better and in more simple explanation what have you said. Fourthemore the subseguent develop for distribution from 10-1-1-1 seems like Goren valutation for short suits: with 0 cards +3 points , with 1 card + 2 points , with 2 card +1 points extended to three short colors : with 2-2-2 until 2-0-0 -1 with 1-1-1 until 1-0-0 -2 with 0-0-0 -3.

 

What you say is partly true. But here it comes to tricks, not for points (equivalent to HCP). I'm talking about the maximum number of losers, which means the maximum number of tricks that can be lost.

 

Pavel Bogev

[gergana85]

Do you mean gergana and lovera and lurpoa are same people?

 

Absolutely not understand people who take part in the forum, but did not discuss on the theme, just write sentences and dealing with conspiracy theory.

[MrAce]

Absolutely not understand people who take part in the forum, but did not discuss on the theme, just write sentences and dealing with conspiracy theory.

 

Na, don't get too excited. Only Gonzalo cares about who you are and why you are here and that is probably because he was bored.http://www.bridgebase.com/forums/public/style_emoticons/default/wink.gif Trust me no one is losing their sleep over you or conspiracy.http://www.bridgebase.com/forums/public/style_emoticons/default/smile.gif

[gwnn]

Hey Pavel, thanks, I checked those sites, although not all of them completely, it's quite a bit of reading material that you give me. Why not directly quote from your sources if they say what you say? Could you show me a source which shows any hand other than a 4333 distribution that gives an LTC of 12? Please show me or point me to some direct quotes. Don't send me a book or anything like that. My name is Csaba BTW, originally from Romania, and I think our countries should be friends.

[gwnn]

OK, how about this PDF?

 

https://www.mrbridge.co.uk/library/LTC_1.pdf

 

He shows

543

54

5432

5432

 

And says it has 11 losers. Can you show me any quote where they take the total A,K,Q and subtract the number from 12? Anywhere?

 

The second hand from the very first link you gave me is:

Axxx QJT x xxxxx

 

According to your theory, they should give a LTC of 10, right? One A, one Q -> 10 losers. According to my theory, it is min(4,3)+min(3,3)+min(1,3)+min(5,3)-2=3+3+1+3-2=8 losers. Guess what the PDF gives? eight.

 

Could you read the sources before you link them?