[pclayton]

jillybean2, on May 12 2008, 06:26 AM, said:

http://online.bridgebase.com/myhands/hands...rname=jillybean

Funny board, only 3 pairs passed this hand out. Feeling frisky I decided to open 3♦ and managed to get away with it. I didnt think this was a smart bid, hence the post. Playing in a "serious" game I think I would pass but maybe thats not smart either, who knows.

This is a nice little case study to look at.

Also look at the frequency of times your RHO opens (13 out of 47 by my count) the bidding with xxx QJxx Ax Axxx in 3rd seat and what happens. Frequently they were in 3♥ or 3♠ making.

Compare a 3♦ opening with a 1♦ or 2♦ call. So many times the opponents competed successfully. 3♦ bought the pot more often than not.

[pclayton]

hotShot, on May 12 2008, 07:42 AM, said:

jillybean2, on May 12 2008, 04:26 PM, said:

http://online.bridgebase.com/myhands/hands...rname=jillybean

Funny board, only 3 pairs passed this hand out. Feeling frisky I decided to open 3♦ and managed to get away with it. I didnt think this was a smart bid, hence the post. Playing in a "serious" game I think I would pass but maybe thats not smart either, who knows.

Those who open should take a close look at this traveler,
opps make 3♥. Fortunately they can only make 2♠
It seems to be from a BBO tourney where very aggressive bidding is common.

Who should take a call over 3♦?

I can understand a balance with the North hand over 2, but not 3.

[jocdelevat]

east had 11hcp in third seat. i suppose thats why the hand was not pass out for most of the tables.
I saw there some funy openings too like with east hand open 1s with 3 cards. Does 3c transfer to diamonds realy exist? I never heard. I saw 2 or more tables open 3c and alert diamonds. Whats the advantge.

[Mbodell]

jtfanclub, on May 12 2008, 06:32 AM, said:

Mbodell, on May 12 2008, 04:04 AM, said:

On average we expect to make 8.976 tricks with diamonds as trump.  2♦, what I'd bid, makes 89.6% of the time.

In those 250 times, see how many tricks it makes in the opponent's better major, because there's no way in hell they're passing out 2♦. Not unless your partner's the most conservative bidder on the planet. One of your opponents is going to have the points and shape to make a call the second time around.

So forget 2♦. You make 3♦ just under 50% of the time, if it's a bell curve. So if they make 2 of a major, opening this is a loser. Unless your opponents are especially foolish, of course.

Do people not understand what a par score is? Multiple people discussed only part of the data from my simulation but really I think the data is pretty compelling. And this isn't surprising given what inquiry has written in the past about the flaws of Pearson points. I ran 5000 hand double dummy overnight and the results are the same within 1% and we have a positive score (>0) on the board 65% of the time. That means we can take more tricks then the opponents most of the time. Sometimes it is 3d>2M, sometimes we make 2nt>2M, and sometimes the opponents can't even make 2M.

At matchpoints if we pass we get a score of 0. If we don't pass and the par result is obtained (I.e., everybody bids knowing all the double dummy results) we get a positive score 65% of the time. A large percentage of our positive games are either in diamonds or NT with diamonds running. Our getting to make the first statement of bidding diamonds is a winning action.

Anyone who claims that we should pass this has to have a justification about how much worse a position we will be than our opponents at finding the par score when we get to bid first, because if it is just about finding the par score we win. I think a 2♦ bid here is clear. It is a slight overbid in my system as it should be promising 7 tricks even opposite a terrible misfitting hand.

We have a partner and some of the time the opponents will bid 2M. Some of a time this M will be their best strain. Some of the time we will set them. Some of the time our partner will bid 2nt or 3nt and we'll make. Some of the time our partner will bid 3♦. Over all, many more good things than bad will occur when we open 2♦.

If people want to make claims about what kind of hands overcall our 2♦ and what moves our partner makes I can simulate the auction from here, but I maintain that we are in a better position, with respect to par contracts, than our opponents when we bid first. And the par contracts favor us significantly. In fact the actual hand linked in looks to be a case where this would be in the minority where the par result is against us and hence would be in the ~35% of my hands where I'd say pass was the winning option but yet again bidding would be the winning option because EW can't find the par result. From looking at travelers it looks like par result is 4dN/SX-1 bid as a sac against 3HE/W which makes, but almost no EW found the hearts and instead found spades which doesn't make or sold out to 3 diamonds which does make.

I'm not surprised pass out got a score in the ~30% range on this hand.

[mikeh]

I'm not convinced that simulations are much help in trying to estimate likely outcomes in a mp field for various options, because one thing we 'know' is that one or two or all three of the other players have hands on which their earlier or second round decisions will vary, often widely, across a mp field. And any simulation involves the simulator rejecting or accepting hands, for inclusion in the study, based upon his or her view of the degree of probability of any such action. Allowing for the best efforts imaginable, such a simulation will be greatly subject to the simulator's constraints filter, which, in turn, will be idiosyncratic.

Personally, this is a 2♦ opening for me, but a borderline one. 3♦ shows an entirely different hand.. it is highly invitational to 3N, from my pov. I would expect 7 good diamonds (AKJxxxx or the like) and an outside card.. i.e. a full opening short in the majors.

2♦ shows, for me, a hand of about 10-13 hcp with 6 chunky diamonds... here, the suit is weak but the 7th card makes up for it.

My entirely subjective, non-scientific rating is: 2♦ 100, pass 80, 1♦ 60, 3♦ 50.

I like 1♦ less than pass because there is too much risk of us being outbid after 1♦. The main problems with 3♦ are that it may be too high, when partner has the 4=4=1=4 9 count that is consistent with the acution, and it may get us too high when he has the balanced 9 count with the diamond Q or J that warrants 3N by him.

[jtfanclub]

Mbodell, on May 12 2008, 06:20 PM, said:

jtfanclub, on May 12 2008, 06:32 AM, said:

Mbodell, on May 12 2008, 04:04 AM, said:

On average we expect to make 8.976 tricks with diamonds as trump.  2♦, what I'd bid, makes 89.6% of the time.

In those 250 times, see how many tricks it makes in the opponent's better major, because there's no way in hell they're passing out 2♦. Not unless your partner's the most conservative bidder on the planet. One of your opponents is going to have the points and shape to make a call the second time around.

So forget 2♦. You make 3♦ just under 50% of the time, if it's a bell curve. So if they make 2 of a major, opening this is a loser. Unless your opponents are especially foolish, of course.

Do people not understand what a par score is? Multiple people discussed only part of the data from my simulation but really I think the data is pretty compelling.

Wow, that was an impressively long post, but you never answered my question.

If you open 2♦, the opponents will almost certainly bid 2 of a major (with or without an X first). The question is, how often will they make it.

I'm not the least interested in the fact that if partner could psychically know what I have then he could make a call that would win 65% of the time. For example, what are the odds that partner has the heart stops necessary to bid 3NT?

Your partner, on the average, is going to have a 10 count with 2 diamonds. I'm sure that out of pass, double for penalties, 2NT, competing to the three level, doubling the opponents if they go to the three level, and bidding 3NT, at least one of those options will be 65% for any given hand. Unfortunately, you haven't given partner the information he needs to make an informed decision.

This is "last guess" principle. You know it's right for your opponents to bid, and so will they. This puts your partner up for the last guess. The fact that you get 65% if he guesses right won't matter if he guesses right less than half the time.

So, there are two ways to check this.

One is to look at how often they make 2M. If it's a low or high %, then you can calculate success based on partner virtually always passing or virtually always bidding on.

The other is to figure a set of strict rules for what your partner will do over 2M. No peeking in advance, no letting them play in hearts but taking it from them in spades, no "assume partner bid 2♦ with a suit headed by the KT". Then see how well the rules work across this hand, and how well they work across other random 4th hand 2♦ bids.

But you really can't bid based on "well, if partner guesses right, we'll get a good score".

[inquiry]

There is a lot written above about pearson points and being outbid in spades. I will not go over that old ground here, I have posted a long thread about it in the past using a commercial product I am not going to name here. One such thread is this one.... Pearson points and distribution

But I will address a couple of points here. First, NEVER open this hand 3♦ (or 3♣ transfer preempt) in fourth seat. Those bids are aimed at trying to get to 3NT if you catch your partner with balanced hand, a 9 ish hcp and some diamonds.

That leaves either 2♦ or 1♦ (if you open at all). If you play weak twos (that is so that a 2♦ can not be something like flannery, mexican 2♦, multi, SEF GF) you probably should bid 2♦ with this hand. If you do not play weak twos, IMHO you should open 1♦.

I will not post a lot of data from the program I will not mention, but I will throw this observation (study) in. I looked at 4773 hands where the 4th seat held a seven card minor and distributions from 7222, 7321, and 7420 with 10, 11, or 12 hcp. The bidding to this point had to go PASS=PASS=PASS. Note, the bidding didn't go PPP at all the tables these hands were taken from, just that the decision point (pass or bid) was such that the person making the decision was in that position. This is like in real life, you don't know if the biddign went three passes at other tables in your event when you have to make your decision. Maybe RHO opens 1♥ at every other table and your hand bid 2/3♦ each time. Your passing it out, then is a very swingy bid. ESPECIALLY if your opponents are weak. I rather take my chances playing against weak opponents than shoving the cards back unplayed.

Out of 4773 such hands, 587 were passed out, earning -1.18 imps (354 hands) and 34.01% MP (233 hands). Every opening minor bid, whether at the one, two or three level was greater than 50% MP, and positive imps (other than 1D at imps, which average -0.15 imps. This trend was true with 10, 11 or 12 hcp in fourth seat (there were nearly 4200 such bids,

[Gerben42]

I would think the options are 2♦ and PASS. Not 1♦ since that is too easy for the opps, and not 3♦ as it goes down too often.

The whole point of 2♦ is that if opponents get in, they MIGHT NOT PLAY IN THEIR BEST MAJOR!

I would pass against opponents who do not know the value of opening agressively with majors in 3rd seat, and open against most.

[bid_em_up]

Mbodell, on May 12 2008, 06:20 PM, said:

I ran 5000 hand double dummy overnight and the results are the same within 1% and we have a positive score (>0) on the board 65% of the time. That means we can take more tricks then the opponents most of the time. Sometimes it is 3d>2M, sometimes we make 2nt>2M, and sometimes the opponents can't even make 2M.

There is something seriously flawed in your "premises" if you truly find this to be the case.

[Codo]

inquiry, on May 13 2008, 06:02 PM, said:

I will not post a lot of data from the program I will not mention, but I will throw this observation (study) in. I looked at 4773 hands where the 4th seat held a seven card minor and distributions from 7222, 7321, and 7420 with 10, 11, or 12 hcp. The bidding to this point had to go PASS=PASS=PASS.

Not many will disagree with an opening with 12 HCPs and a 2272 hand.
So your data about these hands is irelevant.

How many hands did you have with 10 HCPs a marginal 7 card length and a 2272 shape? These are the hands we should use for any conclusion.

F.E. a xx,xx,AKQxxxx, Jx hand has much more for opening then the hand in the threat. So what was your data for 10 HCPs, weak suit and 2272?

[Ant590]

inquiry, on May 13 2008, 04:02 AM, said:

I will not post a lot of data from the program I will not mention, but I will throw this observation (study) in. I looked at 4773 hands where the 4th seat held a seven card minor and distributions from 7222, 7321, and 7420 with 10, 11, or 12 hcp. The bidding to this point had to go PASS=PASS=PASS. Note, the bidding didn't go PPP at all the tables these hands were taken from, just that the decision point (pass or bid) was such that the person making the decision was in that position. This is like in real life, you don't know if the biddign went three passes at other tables in your event when you have to make your decision. Maybe RHO opens 1♥ at every other table and your hand bid 2/3♦ each time. Your passing it out, then is a very swingy bid. ESPECIALLY if your opponents are weak. I rather take my chances playing against weak opponents than shoving the cards back unplayed.

Out of 4773 such hands, 587 were passed out, earning -1.18 imps (354 hands) and 34.01% MP (233 hands). Every opening minor bid, whether at the one, two or three level was greater than 50% MP, and positive imps (other than 1D at imps, which average -0.15 imps. This trend was true with 10, 11 or 12 hcp in fourth seat (there were nearly 4200 such bids,

Wait a second, do I read this correctly that you took the 4th hands that were suitable and ignored the bidding to that point, changing it to P-P-P. How can you possibly compare scores afterwards?

For instance, if you get a piece of data from a board every dealer opened 1H, say, then this hand can't be scored with the other tables (as it would only simulate dealer mis-valuing their hand and passing). To say this is like real life is missing the point, no?

I get so angry with the pseudo-scientific explanations and simulations... my bridge is often not good enough to always determine what the right course of action on a hand, but my mathematics is good enough not to take notice of simulations - bridge is too complicated a game to reduce to a positivist methodology.

Apologies for the rant, Inquiry.

[hotShot]

Ant590, on May 13 2008, 04:08 PM, said:

Wait a second, do I read this correctly that you took the 4th hands that were suitable and ignored the bidding to that point, changing it to P-P-P. How can you possibly compare scores afterwards?

No you should read that as:

Quote

I'm looking at the travelers of boards where at least at one table the bidding was p-p-p-?

[jtfanclub]

Ant590, on May 13 2008, 09:08 AM, said:

inquiry, on May 13 2008, 04:02 AM, said:

The bidding to this point had to go PASS=PASS=PASS.

Wait a second, do I read this correctly that you took the 4th hands that were suitable and ignored the bidding to that point, changing it to P-P-P. How can you possibly compare scores afterwards?

No, you did not read this correctly.

He's saying that if he looked at the results at a table if if it went P-P-P and 4th hand had a 7 card suit and 10-12 hcp, and compared it to the other tables. For the tables he's comparing to, many of them will not have gone P-P-P.

This does make for a skew. Let's say every board was played 10 times. On board 7, all 10 times it went P-P-P and the 4th player opened, and the auctions and play from there on out were identical. On board 8, only one time did it go P-P-P, the other nine times somebody opened in first, second, or third seat. This makes board 7 worth ten times as much as board 8.

Still, I think with Codo's modification, this is as good as we'll get.

[Mbodell]

jtfanclub, on May 12 2008, 08:31 PM, said:

Mbodell, on May 12 2008, 06:20 PM, said:

If you open 2♦, the opponents will almost certainly bid 2 of a major (with or without an X first).  The question is, how often will they make it. 

I challenge this assumption that we are in a bad spot because we spoke first and that we now have the last guess while opponents have no guess. So what would you want to bid 2M directly and what would you want to X? Describe the types of hands. Do you need a 5cM to bid direct? Do you do it on any 5 card hand even one with a bad 8 points? Do you double if you are 4-3 in the majors? What does partner need to bid 3d? any 3 card support or Hx and 10+ points or shortness in opponents suits?

Be more specific. I bet dollars to donuts that your incorrect in assuming they land in the right spot and that we will sell out to making contracts.

I ran another 5000 deals and tracked the major numbers. They averaged making 7.36 tricks in hearts and 7.03 tricks in spades (presumably because we can take 2 ♥ and often score a ♥ ruff in spades) and when they get to pick their best major they average 7.85 tricks in that major. On fully 63% of the hands we can make strictly more tricks in diamonds than they can make in either major. They make at least 2♥ on 48.6% of the boards and at least 2♠ on 36.6% of the boards. When they get to pick their best major they make 8 tricks about 63% of the time. I still dispute that they will always find their best major over the 2♦ opener, and on the hands where they are most likely to make I think they have diamond shortness and partner is most likely to bid 3♦ and we make that a ton of the time partner bids it.

[jtfanclub]

Mbodell, on May 13 2008, 10:44 AM, said:

So what would you want to bid 2M directly and what would you want to X?

OK.

Any hand with a 5 card major bids 2 of the major.
Any hand with 4-4 in the major doubles.
2nd hand X's with 4-3 in the majors and 10+ hcp, provided the 3 card suit is headed by an honor.
4th hand X's with 4-3 in the majors and 9+ hcp, provided the 3 card suit is headed by an honor.

I expect that will be most hands.

Quote

Be more specific.  I bet dollars to donuts that your incorrect in assuming they land in the right spot and that we will sell out to making contracts.

Will they always land in the right major? Not aways, but a goodly proportion of the time. Often enough that it makes a good baseline for figuring out if we should pass or not.

I have not assumed that we will sell out to making contracts. I'm saying that you haven't shown that we won't sell out to making contracts. That's not the same thing at all.

Quote

When they get to pick their best major they make 8 tricks about 63% of the time.  I still dispute that they will always find their best major over the 2♦ opener, and on the hands where they are most likely to make I think they have diamond shortness and partner is most likely to bid 3♦ and we make that a ton of the time partner bids it.

Well, in general it's better for them to have an uneven split in diamonds. If RHO has a singleton diamond, for example, they're probably better off if LHO has three diamonds than if he has one.

But this one sounds easy to test, based on what you just said.

Assume the opponents bid 2 of a major. We'll do two of the better major for now, although if you can find a good way to figure out when they'll end up in the worse major, go for it. Of course, we're also not factoring in when they bid 3♣ and it's right.

Assume your partner passes with 2- diamonds and bids 3♦ with 3+. Again, if you can find better boolean for when partner should bid over them (and what he should bid), go for it.

How does it work out? If the results are in favor of bidding 2♦, I'm satisfied. If the results are in favor of passing, well, maybe it just needs a smarter switch for when to bid 3♦ (or more factoring in when they end up in the wrong major). But it would certainly show that partner's and the opponent's judgement is going to determine whether it's a good bid.

[bid_em_up]

jtfanclub, on May 13 2008, 11:17 AM, said:

OK.

Any hand with a 5 card major bids 2 of the major.
Any hand with 4-4 in the major doubles.
2nd hand X's with 4-3 in the majors and 10+ hcp, provided the 3 card suit is headed by an honor.
4th hand X's with 4-3 in the majors and 9+ hcp, provided the 3 card suit is headed by an honor.

I expect that will be most hands. .

Dont forget to add in:

4-3-1-5 hands that double in direct
3-3-2-5 that either double or bid 3C (in either seat)
4-3-1-5 hands that double in passout
(54)-1-3 hands that double in direct

and lastly

(54)-1-3 hands that double in passout.

[NickRW]

jtfanclub, on May 13 2008, 04:17 PM, said:

2nd hand X's with 4-3 in the majors and 10+ hcp, provided the 3 card suit is headed by an honor.
4th hand X's with 4-3 in the majors and 9+ hcp, provided the 3 card suit is headed by an honor.

Well, if you double with 43 shapes, then fine, perhaps I would too, but you're kind of making the previous posters point - you won't be playing in your best major fit all the time.

This is exactly the point made on Chris Ryall's weak twos web pages. Everyone knows that a weak 2S is better than a weak 2H for obvious reasons. And the naive assumption is that 2D is weaker still - but it isn't - precisely because people can and do go wrong trying to find which major they fit best in (if any - they could be fitting in clubs).

Nick

[inquiry]

jtfanclub, on May 13 2008, 09:23 AM, said:

Ant590, on May 13 2008, 09:08 AM, said:

inquiry, on May 13 2008, 04:02 AM, said:

The bidding to this point had to go PASS=PASS=PASS.

Wait a second, do I read this correctly that you took the 4th hands that were suitable and ignored the bidding to that point, changing it to P-P-P. How can you possibly compare scores afterwards?

No, you did not read this correctly.

He's saying that if he looked at the results at a table if if it went P-P-P and 4th hand had a 7 card suit and 10-12 hcp, and compared it to the other tables. For the tables he's comparing to, many of them will not have gone P-P-P.

This does make for a skew. Let's say every board was played 10 times. On board 7, all 10 times it went P-P-P and the 4th player opened, and the auctions and play from there on out were identical. On board 8, only one time did it go P-P-P, the other nine times somebody opened in first, second, or third seat. This makes board 7 worth ten times as much as board 8.

Still, I think with Codo's modification, this is as good as we'll get.

Your analysis is correct as far as it goes. However, the result of the 7 is only worth 10 times as much as board 8 if all the auctions go exactly the same way and the play goes exactly the same way at table 7.

For instance, say 4 of the 10 do pass, then that pass counts 4 times, but there are 6 hands where the hand was opened. Some may open 1, some 2, some 3. The opponents may double, overcall or pass. So they six hands could end up in the minor opened, an opponents suit, notrump, and at any level, including much too high if 2nd hand overbids. Each of the good and bad results for the opening bid in fourth position are factored EXACTLY by what happened on them. Only after four passes can the result be the "same" for certain. It is called real world results. No simulation.

[Mbodell]

jtfanclub, on May 13 2008, 08:17 AM, said:

Mbodell, on May 13 2008, 10:44 AM, said:

So what would you want to bid 2M directly and what would you want to X?

OK.

Any hand with a 5 card major bids 2 of the major.
Any hand with 4-4 in the major doubles.
2nd hand X's with 4-3 in the majors and 10+ hcp, provided the 3 card suit is headed by an honor.
4th hand X's with 4-3 in the majors and 9+ hcp, provided the 3 card suit is headed by an honor.

I expect that will be most hands.

OK I did something similar to what you suggested except I had the opponents double when 54 not just bid the 5 card suit. I also replaced honor with top 3 honor so that you would double with the 3 card major Q32 but not with T32. I also allowed our partner to raise us on 3+ diamonds or 2 diamonds including at least 1 of the top 3 (although I tracked this separately so it is as if I allowed them to pass, double, bid a major all before the raise to 3♦). Here's what happened 100,000 hands. Percentages are rounded so may not total 100:

They bid over 2♦ 92% of the time. When they did bid the major they settled on makes 52% of the time. When they didn't bid we make 2♦ just under 4 times as often as we get set. We will bid over their 2M 3♦ around 58% of the time they bid 2M. When we do we will make 3♦ more than 9 times as often as we get set! I didn't track when they'd bid 3M over 3♦ but this all splits into all hands broken down as:

2♦ down 1.5%
2♦ making 5.7%
2M down 25.2%
2M making 13.0%
3♦ down 5.3%
3♦ making 49.3%

We get a + score: 80.2% of the time.

Obviously some of the time they might bid 3M, and I didn't consider 3♣ for them either, but this isn't a marginal result. This is a much better result than pass out.

[jtfanclub]

I'm convinced! Nicely done.