[cybourgh]

Today's bidding systems depend on random deals. For example a major game should be bid if partner and you have 25+ total points. But all of us know you can still go down even with these points. This means that bidding game is correct most of the time but only if the hands are randomly dealt. If the hands are skewed like best hand does then the bidding systems have less or little meaning. This is a terrible way to play bridge. I would love to play in ACBL tournaments but if all of them are best hand then I won't play. Also playing defense is 50% of the game. Best hand reduces this aspect of the game. Anyone else feel this way?

[shyams]

Hi cybourgh, welcome to the forums.

I suppose the "best hand" feature for robot tournaments has become quite an established thing. Most of the standard robot tournaments have the feature.

You are right that the proportion of hands where you get to defend reduces tremendously as a consequence. Also, on a few hands, the human declarer gets an unfair advantage. However, one has to admit that the format is very popular on BBO. I suppose people like to bid (instead of passing), like the increased frequency of slams, and definitely feel relieved to have far fewer instances of pass-out deals.

If you prefer other type of ACBL tournaments, perhaps you would enjoy playing all-human tournaments where the hand strength is completely random.

[johnu]

 cybourgh, on 2020-July-16, 10:13, said:

Today's bidding systems depend on random deals. For example a major game should be bid if partner and you have 25+ total points. But all of us know you can still go down even with these points. This means that bidding game is correct most of the time but only if the hands are randomly dealt. If the hands are skewed like best hand does then the bidding systems have less or little meaning. This is a terrible way to play bridge. I would love to play in ACBL tournaments but if all of them are best hand then I won't play. Also playing defense is 50% of the game. Best hand reduces this aspect of the game. Anyone else feel this way?

IIRC, the hands are randomly dealt but the human player gets the hand with the most points. Best hand tournaments do not affect how many points you need for game because the hands are still random. The only difference is that your side will have more HCP on a large majority of hands.

As far as playing defense, GIB doesn't really give signals you can trust when playing defense, and doesn't understand defensive signals. You frequently have to make random guesses on defense in the absence of a signal from GIB. I think for most people, this is not a pleasant experience.

And GIB makes passive leads too often, especially when partner has overcalled or freely bid a suit earlier in the auction. When I play robot tournaments, I bid on the expectation that GIB is going to make a bad lead or will fail to continue a good lead.

[nige1]

I prefer the best-hand format for on-line individuals because you can

• Practice more bidding. Gib's bidding is more predictable than it's play, so there's more opportunity for skill.
• Play the hand, most of the time. When declaring, you control all your side's assets, so it's more skilful than defending, when you control only half. Furthermore, it's hard to understand Gib's defensive strategy so it's hard to co-operate in defence. Gib's defence seems to have become worse over the years. It sometimes chooses tactics that can't possibly win.

[nullve]

I really prefer the non-best hand format. But since being on defence with GiB is so unpopular, how about a format where I or partner gets the best hand?

[johnu]

 nullve, on 2020-July-16, 17:54, said:

I really prefer the non-best hand format. But since being on defence with GiB is so unpopular, how about a format where I or partner gets the best hand?

That has been suggested a number of times. I've never seen BBO comment on the idea.

[pilowsky]

I quite like them although I prefer the group challenge format (Forum challenge is the biggest). This is why.

There are several things to bear in mind when playing in a best hand tournament.

The most important thing to remember about a best hand tournament is that just because you have the most HCP, that doesn't mean you have the 'best hand'.

Also, just because South will always have the most HCP points of any of the individual hands, this does not mean that NS will have the most HCP as a pair.

Once you make this realisation your score will improve dramatically. Part-scores will become your friend.

If you have 14 HCP and East and West each have 11+, then your chances of making a slam are dismal. When you add in distribution, a box of tissues is needed.

When I started playing in robot daylongs I assumed that I was going to get a lot of practice at Declarer play. The results do not bear this out. Here is what happened to me today.

I scored, what was for me, a reasonable 45.74%. I even scored a slam on board 3. But my best board was 4H-2! (85.9%). It turns out that if I had played in defence on 5 out of 8 boards I would definitely have come in the top 25%.

On the plus side, you do have the most HCP. This is a blessing and a curse. Why? Because North definitely does not have more points than you. So, if you open the bidding, don't expect much from North. It will never be there.

Except, in terms of distribution! Sometimes North will surprise you.

Next, when you play in a robot tournament remember that you are actually playing in multiple tournaments that become intercalated to form one big tournament. This is an anti-cheating measure. But it has some negative effects. The big negative is that in matchpoints in a small field a small error in bidding or play can be very costly.

Here's an example from today.

4HN+2.....1 680 100.00% Sample Movie

4HN+1.... 6 650 87.50% Sample Movie

3NS+1.....1 630 75.00% Sample Movie

4HN=.......14 620 48.21%Your Movie

And.......... 7 more people below me.

At IMPS, 1 trick ought to be less costly - at match points, your loss is doubled.

Things can get worse.

In the same tournament as above, on a different board, there were only 18 players. This is where I scored 85.9% with 4H-2 scoring -100. The robots EW made 3S for -140 and 55.88% and the people who made 4H-3 scoring -150 score 23.53%.

Two days ago I was delighted to get a top board but only scored 64.58%. So did 18 of the other 25 people that played the board.

That's the problem with robot daylongs. It looks like you are competing with over a thousand other people but in fact, you aren't. You get to play 8 boards and there is a ceiling on some of your boards. The really good players always do well. They always will. I will always stuff up a board here and there. That's bridge.


The interesting feature of matchpoints scoring is that it follows the rule of 123.

This is a statistical rule that states that 68% of the population sits within 1 standard deviation of the mean. 95% within 2 standard deviations, and just 1 lies outside 99%. Here you can see that my result makes me boringly average along with 14 others, 7 people were mild over-achievers, and one crafty individual found an extra trick.

As an aside, a corollary of the rule of 123 is that to adequately parametrize a curve - that is take enough samples to see something properly - you should take at least 20 samples. This is because at least 1 in 20 could be due to chance. By definition. This is probably why most Club sessions are 24 boards long (at least in Australia).

This sampling theorem was worked out by a guy called Nyquist. The engineers in the audience grossly oversample because they don't want the bridge/building to fall down and kill people. Tragically, even their cautious calculations can go wrong. The Twin Towers in New York were over-engineered to with-stand a direct hit from the largest airliner of the day. Planes got bigger.


And finally, you are going to get hands like this. https://tinyurl.com/y6tlsbb6. You might have the most HCP, but who's having the most fun - if robots had fun. I think the West robot does.

[pescetom]

 pilowsky, on 2020-July-17, 06:20, said:

The engineers in the audience grossly oversample because they don't want the bridge/building to fall down and kill people. Tragically, even their cautious calculations can go wrong. The Twin Towers in New York were over-engineered to with-stand a direct hit from the largest airliner of the day. Planes got bigger.

The planes that hit the towers were Boeing 767-200, smaller (max 136 Tonnes) than the Boeing 707 (max 147 Tonnes) at 600 mph which should have caused only "local damage" according to the engineer responsible in 1964.

[johnu]

 pescetom, on 2020-July-17, 07:07, said:

The planes that hit the towers were Boeing 767-200, smaller (max 136 Tonnes) than the Boeing 707 (max 147 Tonnes) at 600 mph which should have caused only "local damage" according to the engineer responsible in 1964.

I saw a documentary on the engineering behind the World Trade Towers, how well they were built, and why they collapsed.

IIRC, they did survive the crashes fairly well. I don't think they addressed if the buildings could be fully repaired if they didn't collapse. Consulting engineers concluded that the intense heat from a tankful of burning aviation fuel heated up structural steel girders to the point when they weakened and eventually caused catastrophic collapse. There may have been a problem with inadequate or missing insulation on some structural elements that could have contributed to the collapse.

[pilowsky]

You are right, wrong, and focussing on the wrong issue. All at the same time. Quite an achievement really, well done. Of course, I saw the same documentary, but the planes and the building are a tragic metaphor to emphasize the importance of avoiding aliasing which is what the Nyquist theorem and oversampling is all about. If you really want to read the technical specifications about planes and stuff. Here is an article. Discussing the documentary that we all saw. As I recall the guy in charge only scheduled 11.5 minutes for every meeting. Or some similar precise time. This attention to detail is good because it keeps things ticking over, and gets the project completed on time, but it can mean that the fireproofing turns out to be inadequate and lives are lost. Remember the lives? It always amazes me when I visit some of these structures - including the Sydney Harbour Bridge - that there are plaques and signs with information about span, height, nuts, bolts, idiots that cut ribbons, but nothing about the workers that died on the job, who literally sacrificed their lives so that I can smile, as Clive James put it at "the crushed diamond sea and the azure blue sky". Who gives a ***** about the planes?
Edited to remove the first link. Thank U John. Point remains.

[johnu]

 pilowsky, on 2020-July-18, 05:54, said:

You are right, wrong, and focussing on the wrong issue. All at the same time. Quite an achievement really, well done. Of course, I saw the same documentary, but the planes and the building are a tragic metaphor to emphasize the importance of avoiding aliasing which is what the Nyquist theorem and oversampling is all about. If you really want to read the technical specifications about planes and stuff here is an article. Here is a better one. Both discuss the controversy that went into the documentary that we all saw. As I recall the guy in charge only scheduled 11.5 minutes for every meeting. Or some similar precise time. This attention to detail is good because it keeps things ticking over, and gets the project completed on time, but it can mean that the fireproofing turns out to be inadequate and lives are lost. Remember the lives? It always amazes me when I visit some of these structures - including the Sydney Harbour Bridge - that there are plaques and signs with information about span, height, nuts, bolts, idiots that cut ribbons, but nothing about the workers that died on the job, who literally sacrificed their lives so that I can smile, as Clive James put it at "the crushed diamond sea and the azure blue sky". Who gives a ***** about the planes?

Yikes, posting a link to a 9/11 conspiracy site???

[etha]

yes best hand is ridiculous and not bridge.

[Vampyr]

 etha, on 2020-August-13, 06:20, said:

yes best hand is ridiculous and not bridge.

Playing with robots is not bridge, but best hand compensates enough to make it reasonable to enter the odd free tournament.

[Vampyr]

 cybourgh, on 2020-July-16, 10:13, said:

Also playing defense is 50% of the game. Best hand reduces this aspect of the game. Anyone else feel this way?

Playing partnership defence is 50% of the game. You can’t do this with robots; better not to defend at all.

Practise defence with your favourite partners instead.

[Vampyr]

 pilowsky, on 2020-July-17, 06:20, said:

I quite like them although I prefer the group challenge format (Forum challenge is the biggest). This is why.

There are several things to bear in mind when playing in a best hand tournament.

The most important thing to remember about a best hand tournament is that just because you have the most HCP, that doesn't mean you have the 'best hand'.

Also, just because South will always have the most HCP points of any of the individual hands, this does not mean that NS will have the most HCP as a pair.

Once you make this realisation your score will improve dramatically. Part-scores will become your friend.

If you have 14 HCP and East and West each have 11+, then your chances of making a slam are dismal. When you add in distribution, a box of tissues is needed.

When I started playing in robot daylongs I assumed that I was going to get a lot of practice at Declarer play. The results do not bear this out. Here is what happened to me today.

I scored, what was for me, a reasonable 45.74%. I even scored a slam on board 3. But my best board was 4H-2! (85.9%). It turns out that if I had played in defence on 5 out of 8 boards I would definitely have come in the top 25%.

I just played in one and scored 71.9%. I defended on 0 boards. It is definitely best to have the shitty robot defenders as opponents instead of as partner.

[Phil]

Weirdest thread drift in the history of the forums. WD.

I am fine with Best Hand in the NABC Online. It maximizes decisions for the player in the play. Sometimes defense is just following suit. I don't try to game the robots too much, although others have made a cottage industry out of it. I am surprisingly boring and down the middle in these things. Yes, there are inferences about the other hands if you take a few seconds and stay in the present as this isn't really different from watching LHO turn up with a 10 count having passed as dealer and hooking RHO out of the queen. There are more opportunities like this, frankly, so it rewards counting and de-emphasizes playing on general principles.

For the Forum Challenges, I like non-best hand and BAM. It is mano-a-mano and watching two good players slug it out in this format is a great bridge test. IMPs (especially with the sometimes volatile robots) add too much variance.

If you won a Forums non-best MPs, be proud of yourself. Actually, be proud of yourself if you won a Forums Challenge regardless.