Join Bridge Winners
All comments by Bill Segraves
1 2 3 4 5 6 7 8 9 10
You are ignoring the author of this comment. Click to temporarily show the comment.
We now have a way to handle this as well, though opener can short-circuit it if holding 6 spades. For you, does opener break the transfer when holding 6 spades?
an hour ago
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks, David. You and Craig have me nearly on board. Can you tell me a bit more about the 1H-1S-2R-2S forcing school? What's the invite with 6 spades and what's the GF? Does 2S show 6 for you?

(I think I can deal with 5-5s another way. It can mostly be solved within 2-way checkback.)
an hour ago
You are ignoring the author of this comment. Click to temporarily show the comment.
If I may press you on this (I'm genuinely curious), will you really pass 1H with a heart single, a spade 6-bagger, and a 3-count? And if not, what will you bid with it? 2, widening the range, or the 1 that denies you the possibility of rebidding them if 2 rebid will be forcing?
13 hours ago
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks for the example hand, David. Do you have a solution to my question about making the 1 - 2 bid so wide-ranging?
15 hours ago
You are ignoring the author of this comment. Click to temporarily show the comment.
Interesting link. Thanks, Steve. I believe that Gazzilli largely takes care of the Sparts problem for us, but this is interesting to consider.

We play lots of fit bids, so that came to mind as a possible use, too, but doesn't seem so necessary here. Sub-invitational non-fit would work for our purposes, too, by freeing up another bid in the Gazzilli structure.
19 hours ago
You are ignoring the author of this comment. Click to temporarily show the comment.
Interesting (if I cleared the 6 spade sub-invites out with this, I could show the 5-5s using the bid they would have taken). But how wide a range would this have? If I had 6 spades and short hearts, I might try to improve the contract with very, very little in the way of HCP, but invitational wouldn't start until 9 or 10ish, right? How's opener supposed to know what to do?

I'm not sure I'm seeing the full advantage of making 2 a 5+ GF. it seems to me like maybe if it let us make 2 a 6 card invite and 3 the 6 card GF after 2 (let's say 2, because 2 gets us into Gazzilli auctions) that'd be more helpful. Can you explain further?

We play that opener will rebid 1NT if balanced and will raise spades on three if a non-balanced minimum (via Gazzilli with a non-minimum), so why do we need to show the 5-bagger?
20 hours ago
Bill Segraves edited this comment 20 hours ago
You are ignoring the author of this comment. Click to temporarily show the comment.
Following up now on the question as to whether/when to pull a weak 2 response to Gazzilli. I grouped the main potential 2 responses (5 hearts with short spades, 6 hearts with <3 spades, and the 1=4=5=3), and 2 makes a little less than half the time opposite the hands opener can have at this point in the auction. Unsurprisingly, 2 fares far worse, a little better than 20%, when opener has heart shortness. Next, I asked what happens if opener pulls with shortness. If opener is 5-5 in the blacks, then 3 is now the winning contract (3 makes about 1/3 of the time and things can even be improved further if responder re-pulls when holding 6 hearts and club shortness). If opener has 6 spades, opener pulls to spades (see below wrt level). The only remaining shape with < 17 HCP is 5=1=3=4. With that hand pattern, opposite responder's grouped hand types, either 2NT or 2 come home a little less than 30% of the time. Pulling doesn't really do any good. Therefore, all other things being equal (see below), we'd like to pull from 2 with club shortness when holding 5-5 in the blacks.

*If* we're going to pull 2 using this strategy, should we bid 2 over Gazzilli when holding the 1=4=5=3 death hand that started all of this? Yes! With responder pulling on 5-5s, 2 makes ~ 20% of the time, around twice as often as the runners up.

Unfortunately, all other things are not equal, and we have to give up one of our 2, 2NT or 3 bids if we want to be able to pull 2. This is a frequency and payoff game. 2 is a bid that we'd like to be able to make when holding ~ 16-17 and a max of 2 hearts. That comes up about 3 and a half times as often as the 5-5s with short hearts, and having to introduce it at the 3 level costs us a make ~ 1/3 of the time and removes our ability to show stronger spade hands with 3. If 6-4s are covered with another bid, as in some Gazzilli structures, then it's just 6=2=3=3 hands that might like to rebid the spades. The frequency here is only about 1 and a half times higher than the 5-5s with short hearts, but the cost is about the same (in spade contracts). I didn't run the outcome in 2, but I expect that leaving it in 2 may better than committing to 3.

2NT is a bid that some Gazzilli systems would use for strong balanced hands (e.g., 18-19), and it would be unfortunate to have to give that up if it's being used for those hands or something equally useful. (The strong balanced is around same frequency as the 5-5 with heart shortness.)

Unless it can be shown some other way, 3 is important as the way to show strong 17+ hands with spades and clubs. It occurs about 3 times as often as the black 5-5 with heart shortness, so unless we have some other way to show those hands, it's clearly not worth giving that up.

The best answer will depend on the full context of the rest of the Gazzilli system, but for many, this will come back to the meaning of 1 - 1NT - 2NT. If that shows the strong balanced, then that bid's available. If that shows the strong 6-4s, then 2 is potentially available to show the black 5-5s with heart shortness. One way or another, it looks to me like it's worth it.

Thanks to John and all for good suggestions and comments that helped me to identify, refine and (I hope) solve the problem.
20 hours ago
Bill Segraves edited this comment 20 hours ago
You are ignoring the author of this comment. Click to temporarily show the comment.
I ran the stronger hands: 8-9 HCP with 1=4=5=3 or 1=5=5=2 shape. 1=4=5=3 unsurprisingly continues to be a problem holding, and the three suit contracts (2, 3 and 3) that are in the running are still close to a toss-up (we've bypassed 2 at this point in the auction). 2NT is actually the best of the lot, at a little over 30% (vs low to mid 20s for the others), but we're going to use 2NT for other things (see below).

1=5=5=2 does a *lot* better, with 3 and 3 both making in the mid 40s. Even making four started to show up often enough (~20%) that I started thinking about probes for game on relatively light red 5-5s (see below).

As I suggested in the OP and as Ian Hodges observed is part of AMBRA, we could contemplate showing the 5 diamond hand with 2NT, and I explored the effect of grouping these two hand types into a 2NT bid over 2. 3 is unsurprisingly the most common make if we group them (it was tied for the win with both of the component hands), but opener gets to play, too. I didn't break out the frequencies carefully enough to re-assemble an overall success rate, but it's clear that we can get to our best result if opener passes 2NT with 5=2=2=4, bids 3 with 5-5 in the blacks, bids 3 when holding 3 diamonds, and bids 3 when holding 3 hearts.

Five card heart invites are particularly awkward in standardish 2/1, whether or not playing Gazzilli, and it's easy for the heart suit to get lost in Gazzilli when the opening is in spades. We were already using one of the two 2NT bids available to responder (directly over 2 and after opener's call after 2 positive) to show invites with 5 hearts, and the success of the red 5-5s got me wondering just how far we could push this. I ran our problem hand, 9 HCP 1=5=5=2, against 5 spades, 3+ clubs with 3 hearts and 14+ dummy points (including the weak notrump hands and also black 2-suiters), and it sure doesn't look like a problem hand now: it makes 4 ~ 70% of the time. It is to be expected that the 8 HCP hands will do pretty well, too. What I think we'll do, therefore, is to use one of the 2NT calls to show light distributional heart invites. It'll be a little awkward when there's *not* a heart fit, but I'm thinking that opener can either pass or bid 3 (showing 5) to reject hearts and that responder can then pass or bid 3 (showing 5-5).
Sept. 18
Bill Segraves edited this comment Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
And you play a *non*-forcing notrump.

PS - My pard and I decided it wasn't worth giving up other things to play Flannery. Message me if you want to know the data we based it on and the solution we adopted.
Sept. 18
Bill Segraves edited this comment Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
Interesting points on the 2 contract. It might be 4-2, depending on below, but if it's 4-3, that's almost certainly our best spot.

Wrt when opener pulls and what it shows, this is a complicated decision tree, especially since you and I have different agreements that establish the context (most importantly, I'm not currently playing your 2 openings that cover the 5-4 black suit hands). There's a cost to giving up the use of opener's 2 bid over 2 for other purposes unless I could get a nice gain, but assuming I want to pull to 2, I don't think I'd want to pull with a doubleton heart. Since I can raise clubs more freely than you can, 2 will almost always be 5+ hearts (the instant hand being the only exception) and 2 will be our best spot when opener is doubleton. I'm not sure why you're specifying heart singleton honor or small doubleton. Maybe because you can't generally have 5=1=3=4 or 5=1=4=3? Within context in which black 5-4 are *not* covered elsewhere, when might it make sense to pull? Any heart singleton (or void)? Therefore something like 5=1=3=4 for opener? I'm trying to figure out how to simulate this.
Sept. 18
Bill Segraves edited this comment Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
We are pretty Bridge World Standardish in this regard: “opener may pass with 5-3-3-2 or 4=5=2=2 and a hand deemed no stronger than 12 high-card points.” (We'll also pass on some 13s with 4=5=2=2 since the 2 card club rebid is so potentially problematic, though we do have a way to account for this problem in our Gazzilli structure.)

But that's what I mean (and I *think* most mean) by semi-forcing notrump. That you pass with a bare minimum, but that you honor the semi-force by bidding on some slightly stronger hands that are balanced and that would therefore pass opposite a non-forcing notrump.
Sept. 18
Bill Segraves edited this comment Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
If you open 1NT on 14+ and pass a 1NT response with balanced 11+ - 13, on what hands do you bid again opposite a semi-forcing notrump that you wouldn't bid again opposite a non-forcing notrump?
Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks, John. I appreciate your point that some contracts are likely to be easier to defend than others, and especially with the DD results essentially a toss-up, even a small declarer advantage would be relevant. Would you argue that 3 would be equally hard to defend? For 2 and 3, the values will be known and the long suit length will be known to within about a card. For 2 or 3, the values won't be well-defined (since responder's bid prevented opener's further description on all but truly strong hands), spade length will be generally well-defined (but could include certain hands with 6 spades and ~ 16-17), and clubs will be anywhere from 3-5. I'm not sure those are going to be picnic to defend either.

> You will only play 2 when opener has a balanced weak NT, and it's 2/3 you have a 7 fit in hearts.

It's 1/2 to have a 7 fit in hearts for us (we rebid 2 with 5=3=3=2), but I'm interested to explore your point further. In your proposed scheme, 2 shows 5+ hearts or a good four when lacking another bid, right?

Under what conditions does opener pull 2, what do they pull it to, and what does that show?

>I would not bid my minor at the 3 level without 6 of them though.

I presume you only meant diamonds when you say “my minor,” but if you meant clubs, too, let's explore that as well.
Sept. 18
Bill Segraves edited this comment Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
What range is balanced minimum for you?
Sept. 18
Bill Segraves edited this comment Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
To try to get an idea as to what the best potential resting spot(s) might be on these hands, I started running some simulations. On the first problem (1=4=5=3 weak, 5-7 HCP), it is just plain ugly. On 500 hands, with silent opps and with opener holding the weak notrump hands, along with the semi-balanced minimums allowed at that point in the auction, nothing makes with appreciable frequency, and even down 1 is a triumph.

2 makes 6%, 2 makes 9%, 3 makes 12%, 3 makes 8%.
2 off 1 23%, 2 off 1 17%, 3 off 1 18%, 3 off 1 19%

Obviously, this means we'll usually be down 2 or worse on the substantial majority of hands.

Even with 500 deals tested, these numbers will only be reproducible within about a % or two, so bottom line is “success” is approximately equally likely for all the potential contracts. IMO, 2 is the least likely to have the opponents start up the doubling machine.

The problem is an ugly one in “standard” as well, though it's worth noting that 2 or 2 have more than a 50% chance of making 7 tricks or better. This is a cost of playing Gazzilli.

I will also run the 8-9 HCP bad shapes and include a test of what would happen if using 2NT to describe them.
Sept. 18
Bill Segraves edited this comment Sept. 18
You are ignoring the author of this comment. Click to temporarily show the comment.
Oops! Sorry, John. Got wires crossed with another auction from above. Edited to remove the irrelevant part.
Sept. 17
You are ignoring the author of this comment. Click to temporarily show the comment.
Oooh, I like. :) Thanks, John!

Edited to remove irrelevant stuff.
Sept. 17
Bill Segraves edited this comment Sept. 17
You are ignoring the author of this comment. Click to temporarily show the comment.
See above. Are you really playing it semi-forcing? Or have you just expanded your notrump range enough that you can play it non-forcing?
Sept. 17
Bill Segraves edited this comment Sept. 17
You are ignoring the author of this comment. Click to temporarily show the comment.
Because we're playing it in the context of a semi-forcing notrump, which means we can't pass with 14, and would like to avoid passing with 13. Pard can have a 3 card limit raise, 11 balanced, or various 11 counts semi-balanced or unbalanced (we're playing 2/1).

If you play your semi-forcing notrump differently, or if you really play it non-forcing, that's your prerogative, but I was trying to make it clear that that's a hand type on which we are systemically obliged to bid.
Sept. 17
You are ignoring the author of this comment. Click to temporarily show the comment.
Interesting. Somewhat unexpectedly, the notes I am finding online suggest that is potentially to play (SO). Are you interpreting the same way?
Sept. 17
1 2 3 4 5 6 7 8 9 10
.

Bottom Home Top