[Stephen Tu]

---

To me:
1. 2c is horrible, 1nt is clearcut. 1nt should be preferred when in range for 1nt and balanced and 2-3 cds other major. 2c should be reserved for too strong for 1nt or more shape with like stiff H and diamond fit or sixth club or something like that.

2. 2H is horrible, 2nt is clear. Partner shouldn't have 4 hearts, would usually negative double in preference to 2c, if not should have GF strength and can bid hearts himself.

3. 2H should be redefined as strong and forcing this type of reverse sequence. If it were 1h-(1s)-2c-2d instead, I feel like 2d ought to be F1 but could be min.

[virgosrock]

Stephen Tu, on 2018-March-24, 09:29, said:

---

To me:
1. 2c is horrible, 1nt is clearcut. 1nt should be preferred when in range for 1nt and balanced and 2-3 cds other major. 2c should be reserved for too strong for 1nt or more shape with like stiff H and diamond fit or sixth club or something like that.

2. 2H is horrible, 2nt is clear. Partner shouldn't have 4 hearts, would usually negative double in preference to 2c, if not should have GF strength and can bid hearts himself.

3. 2H should be redefined as strong and forcing this type of reverse sequence. If it were 1h-(1s)-2c-2d instead, I feel like 2d ought to be F1 but could be min.

Re 1.
I think in GIBBO system 2c is 5+ clubs and 10+ HCP. Quality of clubs is irrelevant. At this point no idea what pd has so one bids the hand per system. I am ok with this treatment.

vrock

[Stephen Tu]

virgosrock, on 2018-March-24, 12:20, said:

Re 1.
I think in GIBBO system 2c is 5+ clubs and 10+ HCP. Quality of clubs is irrelevant. At this point no idea what pd has so one bids the hand per system. I am ok with this treatment.

vrock

I am claiming this system clearly should be changed. 1nt is a better bid than 2c when both fit, 1nt should be a higher priority rule when balanced and in range. 2c is unlimited and F1, 1nt isn't. So if you bid 2c, you often end up having to declare 2nt instead of 1nt when partner has no interest in game. This is a HUGE loss when 2nt is down 1 or more. Bidding clubs just doesn't have a lot of upside, NT often scores more as a partial, you might end up in 3c making 3 instead of 1nt making 2. You also get a huge loss if you end up in 3c down 1 instead of 1nt making 1, when opener has some blah 3343 type of hand and nothing other than 3c makes any sense. If you really belong in clubs opener will usually bid it.

[wbartley]

One more thing. The 2♥ bid, regardless of the strength requirement should show five or more diamonds, not 4-4.

[virgosrock]

Stephen Tu, on 2018-March-24, 12:35, said:

I am claiming this system clearly should be changed. 1nt is a better bid than 2c when both fit, 1nt should be a higher priority rule when balanced and in range. 2c is unlimited and F1, 1nt isn't. So if you bid 2c, you often end up having to declare 2nt instead of 1nt when partner has no interest in game. This is a HUGE loss when 2nt is down 1 or more. Bidding clubs just doesn't have a lot of upside, NT often scores more as a partial, you might end up in 3c making 3 instead of 1nt making 2. You also get a huge loss if you end up in 3c down 1 instead of 1nt making 1, when opener has some blah 3343 type of hand and nothing other than 3c makes any sense. If you really belong in clubs opener will usually bid it.

Hi Stephen
I used to do the Bridge World bidding challenge and there was a guy who would run simulations on the bidding problems. Simulations sometimes gave results which were counter-intuitive. May I suggest similar simulations for the bidding changes you wish were implemented.

Merely a suggestion.

vrock

[Stephen Tu]

I don't need a damn computer simulation to know that it's losing bridge to force to the 2nt+ level with an aceless balanced 10 count, in order to show a Txxxx suit, with a double stop in the opp's suit, no ruffing value, no weak suit other than the one partner opened. I have plenty of experience at the table to just know this.

99% of the bugs I post in this forum are not at all borderline, and completely egregious like this one.

I do run simulations for things I consider borderline or unclear, when I'm deciding what bids are best in my human partnerships, second guessing my choices from my offline play. But GIB bugs I complain about are nearly always things I don't think are at all close decisions, and usually consider completely ridiculous, if they are subtle/borderline I note it in my post.

Why are you so strongly in favor of 2c instead of 1nt? What great triumphs do you think this will achieve on average?

[virgosrock]

Stephen Tu, on 2018-March-26, 10:15, said:

I don't need a damn computer simulation to know that it's losing bridge to force to the 2nt+ level with an aceless balanced 10 count, in order to show a Txxxx suit, with a double stop in the opp's suit, no ruffing value, no weak suit other than the one partner opened. I have plenty of experience at the table to just know this.

99% of the bugs I post in this forum are not at all borderline, and completely egregious like this one.

I do run simulations for things I consider borderline or unclear, when I'm deciding what bids are best in my human partnerships, second guessing my choices from my offline play. But GIB bugs I complain about are nearly always things I don't think are at all close decisions, and usually consider completely ridiculous, if they are subtle/borderline I note it in my post.

Why are you so strongly in favor of 2c instead of 1nt? What great triumphs do you think this will achieve on average?

I show partner 10+ HCP and 5 clubs. This is a wealth of info for partner as compared with 1 NT.
Initially I was not going to respond since the tone is contentious (at the end of post) and borderline rude.

vrock

[steve2005]

virgosrock, on 2018-March-26, 10:27, said:

I show partner 10+ HCP and 5 clubs. This is a wealth of info for partner as compared with 1 NT.
Initially I was not going to respond since the tone is contentious (at the end of post) and borderline rude.

vrock

Personally, I would prefer partner have something useful in a suit they bid.
I know Gib doesn't lead partner's suit by i do!

If 2♣ shows any 5 clubs i really don't want to know about it I'd rather they pass if they won't bid 1N or 2N.

[Stephen Tu]

virgosrock, on 2018-March-26, 10:27, said:

I show partner 10+ HCP and 5 clubs. This is a wealth of info for partner as compared with 1 NT.

When is this particular info going to be useful to partner?

Pro 2c: you show that you have 10+ unlimited pts rather than only 8-10/bad 11. On what hands is this going to make a positive difference in your final contract? On almost all hands, if you have game, you want to be in 3nt. If partner only wants to be in game opposite 10 rather than 8, he can bid 2nt. I don't see why you having bid clubs will have helped him make a decision. If anything, it will cause him to overvalue decent club holdings like AQx/KQx/AKx, and push to 3nt on less HCP (he cues 2s, you bid 2nt, he bids 3) than he normally would over just a 1nt bid, because he thinks you have almost certain 5 or 6 club tricks.

If you don't have game, 1nt is going to be the best scoring partial an overwhelming majority of the time, because they give you more points for NT tricks than minor tricks. 2nt never outscores 1nt, and occasionally goes down when everything is foul. And if you play clubs, you have to play 3 of them, which likewise will sometimes go down when 1nt makes, and if it makes, NT will usually outscore it for +120 or +150.

Pro 2c: you show a fifth club immediately. How often do you really prefer to play in clubs rather than NT, if partner doesn't bid clubs themselves? What hand do you posit from partner where showing the fifth club enables him to make a better decision?

Con 2c: It's unlimited, and you are forcing yourself to 2 level+. What if 1nt was the limit? What if partner raises to 3c, having nothing better to do (no spade stop themselves, 3 clubs), and 1NT/2nt was a better spot?

[johnu]

virgosrock, on 2018-March-26, 10:27, said:

I show partner 10+ HCP and 5 clubs. This is a wealth of info for partner as compared with 1 NT.

The description of 2♣ says 11+ total points. Only in Gibberish is that hand 11+ total points. 10 HCP, a 5 card suit that looks like a 4 card suit, and distribution points without a fit are a GIB thing.

1NT would show a spade(s) stopper with an upper range of 10 (or 11? in GIB) HCP and NT looks like a likely contract.

IMO, 1NT is a lot more descriptive than 2♣ and in practice, I would consider 2♣ and grade it as a 10% or less bid.

[virgosrock]

Stephen Tu, on 2018-March-26, 12:39, said:

When is this particular info going to be useful to partner?

Pro 2c: you show that you have 10+ unlimited pts rather than only 8-10/bad 11. On what hands is this going to make a positive difference in your final contract? On almost all hands, if you have game, you want to be in 3nt. If partner only wants to be in game opposite 10 rather than 8, he can bid 2nt. I don't see why you having bid clubs will have helped him make a decision. If anything, it will cause him to overvalue decent club holdings like AQx/KQx/AKx, and push to 3nt on less HCP (he cues 2s, you bid 2nt, he bids 3) than he normally would over just a 1nt bid, because he thinks you have almost certain 5 or 6 club tricks.

If you don't have game, 1nt is going to be the best scoring partial an overwhelming majority of the time, because they give you more points for NT tricks than minor tricks. 2nt never outscores 1nt, and occasionally goes down when everything is foul. And if you play clubs, you have to play 3 of them, which likewise will sometimes go down when 1nt makes, and if it makes, NT will usually outscore it for +120 or +150.

Pro 2c: you show a fifth club immediately. How often do you really prefer to play in clubs rather than NT, if partner doesn't bid clubs themselves? What hand do you posit from partner where showing the fifth club enables him to make a better decision?

Con 2c: It's unlimited, and you are forcing yourself to 2 level+. What if 1nt was the limit? What if partner raises to 3c, having nothing better to do (no spade stop themselves, 3 clubs), and 1NT/2nt was a better spot?

If the context is MP nothing more from me.
At IMP I clearly remember reading in some text by some famous folks that c fit partial was more % than 1 NT or 2 NT. I wish I remembered the text.
I came back to MB May of last year and probably played 1000's of hands. I don't remember a single instance where this approach worked out badly. In fact GIBBO has even done this with 9 HCP. Anecdotal experience no doubt. In fact the bidding went to 3C and opponents had trouble deciding if they should go to 4.

vrock

[Stephen Tu]

virgosrock, on 2018-March-26, 19:23, said:

If the context is MP nothing more from me.
At IMP I clearly remember reading in some text by some famous folks that c fit partial was more % than 1 NT or 2 NT. I wish I remembered the text.

I have no doubt that there are some hand types where club partial, 2c, is safer than 1nt and a better spot at IMPS. But not this one where partner opened the other minor, not clubs, you have kqx in both opponent +unbid suits, one of those behind the overcaller, and your suit is ten high. This simply isn't one of them, and furthermore you are comparing THREE clubs to 1nt which is an entirely different animal than comparing 2c vs. 1nt/2nt. Bidding 1nt, you might be able to buy it on a good day for 2c if opener rebids them and opps don't balance. Bidding 2c, you are self forcing to 3c and the opps don't have to do anything.

You asked for a sim so I gave in, dealing out a 1000 hands. Giving partner a weak nt, 43/53/44 in diamonds/clubs, which are by far the most common hand type, and North a spade overcall:
1nt made 97.2% of the time.
3c made only 68.3% of the time.
3c made while 1nt goes down 0.9%
1nt made while 3c goes down 29.8%

Also note that you get to play 1nt not 2nt when partner has doubleton club. In those cases 2nt is failing some 36.5% of the time, 1nt is making in some 32.3 out of those 36.5.

[steve2005]

virgosrock, on 2018-March-26, 19:23, said:

If the context is MP nothing more from me.
At IMP I clearly remember reading in some text by some famous folks that c fit partial was more % than 1 NT or 2 NT. I wish I remembered the text.
I came back to MB May of last year and probably played 1000's of hands. I don't remember a single instance where this approach worked out badly. In fact GIBBO has even done this with 9 HCP. Anecdotal experience no doubt. In fact the bidding went to 3C and opponents had trouble deciding if they should go to 4.

vrock

What good is a fit when partner raises or passes on 3 small and you have 4 trump losers?
I agree a fit is useful but 76543 opposite 982 is useless.


In any case, at imps you goal for game is 3N not 5m.
Yesterday i had a had with 5-card support and A as stopper in opps suit. I bid 2NT not 3m and partner had enough to bid game.

[virgosrock]

Stephen Tu, on 2018-March-26, 20:12, said:

I have no doubt that there are some hand types where club partial, 2c, is safer than 1nt and a better spot at IMPS. But not this one where partner opened the other minor, not clubs, you have kqx in both opponent +unbid suits, one of those behind the overcaller, and your suit is ten high. This simply isn't one of them, and furthermore you are comparing THREE clubs to 1nt which is an entirely different animal than comparing 2c vs. 1nt/2nt. Bidding 1nt, you might be able to buy it on a good day for 2c if opener rebids them and opps don't balance. Bidding 2c, you are self forcing to 3c and the opps don't have to do anything.

You asked for a sim so I gave in, dealing out a 1000 hands. Giving partner a weak nt, 43/53/44 in diamonds/clubs, which are by far the most common hand type, and North a spade overcall:
1nt made 97.2% of the time.
3c made only 68.3% of the time.
3c made while 1nt goes down 0.9%
1nt made while 3c goes down 29.8%

Also note that you get to play 1nt not 2nt when partner has doubleton club. In those cases 2nt is failing some 36.5% of the time, 1nt is making in some 32.3 out of those 36.5.

Thanks for running the sim Stephen. Great results. Can you teach me how to run similar simulations?

thanks
vrock

[virgosrock]

steve2005, on 2018-March-27, 07:21, said:

What good is a fit when partner raises or passes on 3 small and you have 4 trump losers?
I agree a fit is useful but 76543 opposite 982 is useless.


In any case, at imps you goal for game is 3N not 5m.
Yesterday i had a had with 5-card support and A as stopper in opps suit. I bid 2NT not 3m and partner had enough to bid game.

an interesting counter example. doubt if it proves anything. merely presenting.

---

vrock

[Stephen Tu]

virgosrock, on 2018-March-27, 08:17, said:

Thanks for running the sim Stephen. Great results. Can you teach me how to run similar simulations?

If you have some skill with computer code, you can use Thomas Andrews Deal, or the port of that program by Antony Lee (forum member here) to Python if you prefer Python over the more obscure Tcl. They both utilize a double dummy solver by Bo Haglund. Both programs easily found via a google search.

I actually use a different package, Hans van Stavaren's dealer, but DD analysis with that requires a copy of GIB's bridge.exe, which I happen to have since I bought it some 20 years ago or so. This uses the deal scripting language BBO also uses at its bidding/teaching practice tables, the scripting language is somewhat more bridge specific and less like computer code so it's somewhat faster/easier than the above. But need to track down the GIB.

If you don't have coding experience, well, it'll be harder, but all of the above software have some example scripts so maybe you can work it out.

[johnu]

virgosrock, on 2018-March-27, 08:17, said:

Can you teach me how to run similar simulations?

You can buy Dealmaster Pro. http://dealmaster.com/

There is almost no learning curve and there is an interface with checkboxes and drop down lists where you set the conditions of the simulation. You don't have to learn a scripting language. The interface is very dated, but it works well enough. I'm very happy with my purchase.

[thomas c]

. for what its worth, stephen is correct. 1nt is the ONLY bid that makes sense on that hand.

[A2003]

GIB Doubles

[Stephen Tu]

A2003, you should start a new thread for each of your robot reports, rather than posting what you think are problems to random unrelated threads.