[rbforster]

straube, on May 11 2010, 07:15 PM, said:

Personally, I don't like opening 1C with 15 balanced because it makes it harder to get into a GF. 

This isn't true. The lower you open your strong club, even if you correspondingly increase your point requirement for a GF response, the more frequent you will have a GF auction. Say you like to have 25 hcp for your games. What are the differences between a 16+ club and a 15+ club?

16+ vs 10+....... GF for either
16 vs 9 .............GF for bigger club only
15 vs 10 ...........GF for smaller club only
15- vs 9- ..........non-GF for either

Since it's more likely that your points divide evenly, picking up the additional 15-10 GF strong club auction is more frequent than losing the 16-9 one and correspondingly you get more GF auctions. That said, it's not clear that an auction that starts with a strong club is a good thing, given the difficulties with interference, so don't take this as endorsement of a 10+ "strong" club! Also, note that the lower your strong club starts, the more frequent you have the 1C-1D negative auction, which means you need to worry more about 4th hand interference as well.

[OSH]

I checked on the CC, and both Lauria-Versace and Sementa-Duboin played a 2 ♦ to show a 18-19 (20) balanced hand. But this is not a multi, because this opening shows only the strong NT. In this way it could be less pre-emptive for your partner (and for the opps, of course), because he can start the exploration at the 2 level. It would be different with a multi, because the 2 level now is reserved for pass / correct considering the weak opening.

I found a link to a program to calculate the probability of each opening in system, but it doesn’t work, maybe someone can do better and help me, but here are some considerations:
- a 11-15 unbalanced 1♦ hand with 4+ cards (45 in the minors, 4441, ecc), excluding 5332 with 5 ♦ cards, is about 5% (34% for 11-15 and 15% for the am distribution).
- if you open ANY hand with 11+ HCP, it is 44%, in this case you open 1♦ 11 times every 100.
- if you open 1♣ any hand with 16+ HCP (there are exceptions, I know), this is 9,76% in total.
- if you open 1♣ any hand with 15+ HCP, this is 14,18% in total. This means that, including every 15 count in 1♣, you increase the frequency of this opening by 45%. In this case, when you have an opening bid (11+), 33% of the time you have to open 1♣ (if it is 15+ HCP).
- awm suggestion, 1♣ with 16+ and exactly 15 balanced (5332 only minor) is 11,5%, increasing the frequency by 18%.
- a 11-13 balanced (straube question) hand is 9,7%, considering only regular NT distribution and 5332 with a minor. Adding this to the natural 1♦, which is considerably less frequent, it means that now you open 1♦ one time every three. Imo this a bit too often, considering my fear of the "foggy" diamond
- a 14-16 balanced hand is 5,3%, 12-14 is 8,2%, 12-15 is 10% Considering only the hands worth an opening bid (11+ in this contest), they become: 12%, 18,5% and 23%. 5M332 and less regular pattern are not considered. Playing 12-15 NT, one time every 5 the opener will have exactly 15.

[straube]

Rob F, on May 12 2010, 12:12 PM, said:

straube, on May 11 2010, 07:15 PM, said:

Personally, I don't like opening 1C with 15 balanced because it makes it harder to get into a GF. 

This isn't true. The lower you open your strong club, even if you correspondingly increase your point requirement for a GF response, the more frequent you will have a GF auction. Say you like to have 25 hcp for your games. What are the differences between a 16+ club and a 15+ club?

16+ vs 10+....... GF for either
16 vs 9 .............GF for bigger club only
15 vs 10 ...........GF for smaller club only
15- vs 9- ..........non-GF for either

Since it's more likely that your points divide evenly, picking up the additional 15-10 GF strong club auction is more frequent than losing the 16-9 one and correspondingly you get more GF auctions. That said, it's not clear that an auction that starts with a strong club is a good thing, given the difficulties with interference, so don't take this as endorsement of a 10+ "strong" club! Also, note that the lower your strong club starts, the more frequent you have the 1C-1D negative auction, which means you need to worry more about 4th hand interference as well.

We're not disagreeing. I meant that it's more difficult to get into a GF assuming one has opened 1C when that 1C promises less.

[OSH]

awm, on May 11 2010, 08:32 PM, said:

Putting these together, I think it makes some sense to let the 1♣ opening start at 15 for balanced and one-suited hands, but start somewhat higher (perhaps 17) for two and three-suited patterns.

From an aesthetic point of view I don't like it, but I think I can try this one with no impact, and is possible that the practice will convince me. I'm not sure if I can play 1♣ 15+ and 12-14NT NV and keep the old 1♣ 16+ and 13-15NT Vul only.
We have two opening bids for the 2-suited hands with 4,5 losers (that is approx 13-18 HCP), so for the second requirement I'm already in line.

[OSH]

Rob F, on May 12 2010, 06:12 PM, said:

Also, note that the lower your strong club starts, the more frequent you have the 1C-1D negative auction, which means you need to worry more about 4th hand interference as well.

My concern is not (only) for 4th hand interference, sometimes useful in this situation, but for the second negative.
Playing 1♣-1♦-1♥ as strong, now we play that 1♠ is not forced, but is the only (second negative) reply non GF.
In this contest, if 1♥ is 18+, I need 1♠ to be 0-5. Or, if one need 25 to be GF (probably still not enough for my card play...), 0-6. There is the space for a third negative?

[straube]

OSH, on May 12 2010, 12:50 PM, said:

I checked on the CC, and both Lauria-Versace and Sementa-Duboin played a 2 ♦ to show a 18-19 (20) balanced hand. But this is not a multi, because this opening shows only the strong NT. In this way it could be less pre-emptive for your partner (and for the opps, of course), because he can start the exploration at the 2 level. It would be different with a multi, because the 2 level now is reserved for pass / correct considering the weak opening.

I found a link to a program to calculate the probability of each opening in system, but it doesn’t work, maybe someone can do better and help me, but here are some considerations:
- a 11-15 unbalanced 1♦ hand with 4+ cards (45 in the minors, 4441, ecc), excluding 5332 with 5 ♦ cards, is about 5% (34% for 11-15 and 15% for the am distribution).
- if you open ANY hand with 11+ HCP, it is 44%, in this case you open 1♦ 11 times every 100.
-  if you open 1♣ any hand with 16+ HCP (there are exceptions, I know), this is 9,76% in total.
-  if you open 1♣ any hand with 15+ HCP, this is 14,18% in total. This means that, including every 15 count in 1♣, you increase the frequency of this opening by 45%. In this case, when you have an opening bid (11+), 33% of the time you have to open 1♣ (if it is 15+ HCP).
- awm suggestion, 1♣ with 16+ and exactly 15 balanced (5332 only minor) is 11,5%, increasing the frequency by 18%.
- a 11-13 balanced (straube question) hand is 9,7%, considering only regular NT distribution and 5332 with a minor. Adding this to the natural 1♦, which is considerably less frequent, it means that now you open 1♦ one time every three. Imo this a bit too often, considering my fear of the "foggy" diamond 
- a 14-16 balanced hand is 5,3%, 12-14 is 8,2%, 12-15 is 10% Considering only the hands worth an opening bid (11+ in this contest), they become: 12%, 18,5% and 23%. 5M332 and less regular pattern are not considered. Playing 12-15 NT, one time every 5 the opener will have exactly 15.

I think these numbers are pretty damning for the unbalanced natural 1D.

Your frequency distribution then (assuming 12-14 NTs) is something like...

1C 14%
1D 5%
1H 5%
1S 5%
1N 8%

so your 1D is way underloaded. Contrast this to 1C at 16+ (good 15 unbal and bad 17 bal) and foggy diamond and 14-16 NT

1C 9%
1D 14%
1H 5%
1S 5%
1N 5%

In general, one would expect the lower bids to have greater frequency than the higher bids because they leave more room for communication. In the second frequency allocation, 1C appears to be shortchanged compared to 1D, but one has to look at what happens after these openings.

1C is forcing, so there are no 1C-P auctions and typically all of the 1C-1D, 1H and 1C-1D, 1H-1S and 1C-1D-1S even auctions are forcing. So if one wants to save 1C as a forcing bid, then one wants also to restrict its use somewhat. Another way of putting this is that a strong club consumes proportionately more bidding room (sequences) than it would appear if you were just looking at its frequency of use.

btw, what site did you use to run your numbers?

[rbforster]

straube, on May 12 2010, 12:56 PM, said:

We're not disagreeing. I meant that it's more difficult to get into a GF assuming one has opened 1C when that 1C promises less.

Agreed. I realized after I replied that you were talking about the probability of a GF auction, at the point you've opened 1C, rather than the probability of having a GF club auction.

[rbforster]

If you want 1 NT range and a strong club, I've played a 12(+) to 15 NT with no weaker balanced hands being opened and 1♣ being any 16+. This works ok, where we define a 12+ hand as anything you think the field is opening and that you're willing to open Vul. A 3.5 range isn't too bad, but certainly don't play this way if you're tempted to upgrade your 11's!

If you want to come down to a 12-14 NT and a 15+ club (at least one that includes 15+ bal, optional on the unbal hands), I think Free's suggestion below should work pretty well with a 1D/1S negative/double negatives response structure.

Free, on May 11 2010, 01:58 AM, said:

1♣-1♦-1NT shows 15-17
1♣-1♦-1♥-1♠-1NT shows 18-20
1♣-1♦-2NT shows 21-22
1♣-1♦-1♥-1♠-2NT shows 23-24
1♣-1♦-1♥-1♠-2♣-2♦-2NT shows 25+ (2♣ is very strong)
2. I have no experience with this, and I'd advise you not to play it this way 

[zenko]

here is one very interesting analysis on that subject
http://www.migry.com/Articles%20and%20othe...he%20winner.pdf

NT Ranges: The Comparison IMP Gain/Loss analysis on suitable
championship boards from 1992 to 2002 by By Pietro Campanile

In short the conclusion is to pick the most frequent ranges since NT opening is a winner (except 16-18 which is a loser).

Other than that, I would add that the math changes slightly with the seat and with vulnerability (for example after pass-pass you are more likely to hold 14 then 11 points, etc.)

So I would say 1st/nd position non vul 10-12 or 11-13, vul as low as you can stomach it, and 3rd/4th 15-17 always (14-16 also is OK I guess)

[awm]

zenko, on May 12 2010, 02:26 PM, said:

here is one very interesting analysis on that subject
http://www.migry.com/Articles%20and%20othe...he%20winner.pdf

NT Ranges: The Comparison IMP Gain/Loss analysis on suitable
championship boards from 1992 to 2002 by By Pietro Campanile

In short the conclusion is to pick the most frequent ranges since NT opening is a winner (except 16-18 which is a loser).

Other than that, I would add that the math changes slightly with the seat and with vulnerability (for example after pass-pass you are more likely to hold 14 then 11 points, etc.)

So I would say 1st/nd position non vul 10-12 or 11-13, vul as low as you can stomach it, and 3rd/4th 15-17 always (14-16 also is OK I guess)

I've read this article before -- I do not think that the analysis is very convincing.

The issue is that causality in these situations is somewhat complicated. Arguably the better pairs or teams are likely to win IMPs whereas the weaker pairs or teams are likely to lose IMPs, regardless of their methods, simply due to better play, defense, judgement, etc. There is some correlation between the skill level of a pair and their methods; in particular pairs who play very "old-fashioned" methods are probably likely to be weaker pairs since they have not kept their bidding up with the modern trends.

I find it hard to believe that 1NT "16-18" isn't a huge winner when you open 1NT. You get a great description of the hand right away, it helps on your slam auctions a lot, etc. Obviously I'm in no hurry to play this range because of frequency issues, but I find any analysis which indicates that 16-18 notrumpers lose large numbers of IMPs when they open 1NT inherently dubious. Much more likely is that the pairs playing 16-18 notrumps are very old-fashioned, playing outdated methods and perhaps weaker declarers than those who are up with the times (and presumably playing more regularly).

It's also true that there are arguments for/against various notrump ranges based on the rest of the system. For example, it's easy to believe that a 10-12 notrump is a winner when you open 1NT (at least at NV), but especially if your system structure is relatively "standard" it's hard to bolt this method on without substantial issues on hands in other notrump ranges.

[straube]

A couple of other things bother me about that article....

One is that the 14-16 NT range seems like it isn't an imp-gainer as compared to the 13-15 or 15-17 ranges. But why would this be? It's smack in the middle of those two ranges. Perhaps this is statistical noise. Or if not, what is the better explanation?

Another thing is that the data was collected irrespective of vulnerability, but the authors seem to make recommendations based on vulnerability.

Another thing is that what we (I) really want to know is whether to open strong NTs or weak NTs and at what vulnerabilities and whether I should open as light as 10 or 11 points and at what vulnerabilities and the article can't really answer these questions.

For example, 10-12 NTs seem to be imp gainers...so when we open 10 and 11 pt hands that folks at the next table are passing or opening 1m, then we gain. So what? And at what vulnerabilities? How do folks who open 10-12 NTs fare with their 13+ hands? Perhaps they lose more imps on these than they gain with the 10-12s.

[zenko]

awm, on May 12 2010, 02:54 PM, said:

zenko, on May 12 2010, 02:26 PM, said:

here is one very interesting analysis on that subject
http://www.migry.com/Articles%20and%20othe...he%20winner.pdf

NT Ranges: The Comparison IMP Gain/Loss analysis on suitable
championship boards from 1992 to 2002 by By Pietro Campanile

In short the conclusion is to pick the most frequent ranges since NT opening is a winner (except 16-18 which is a loser).

Other than that, I would add that the math changes slightly with the seat and with vulnerability  (for example after pass-pass you are more likely to hold 14 then 11 points, etc.)

So I would say 1st/nd position non vul 10-12 or 11-13, vul as low as you can stomach it, and 3rd/4th 15-17 always (14-16 also is OK I guess)

I've read this article before -- I do not think that the analysis is very convincing.

The issue is that causality in these situations is somewhat complicated. Arguably the better pairs or teams are likely to win IMPs whereas the weaker pairs or teams are likely to lose IMPs, regardless of their methods, simply due to better play, defense, judgement, etc. There is some correlation between the skill level of a pair and their methods; in particular pairs who play very "old-fashioned" methods are probably likely to be weaker pairs since they have not kept their bidding up with the modern trends.

I find it hard to believe that 1NT "16-18" isn't a huge winner when you open 1NT. You get a great description of the hand right away, it helps on your slam auctions a lot, etc. Obviously I'm in no hurry to play this range because of frequency issues, but I find any analysis which indicates that 16-18 notrumpers lose large numbers of IMPs when they open 1NT inherently dubious. Much more likely is that the pairs playing 16-18 notrumps are very old-fashioned, playing outdated methods and perhaps weaker declarers than those who are up with the times (and presumably playing more regularly).

It's also true that there are arguments for/against various notrump ranges based on the rest of the system. For example, it's easy to believe that a 10-12 notrump is a winner when you open 1NT (at least at NV), but especially if your system structure is relatively "standard" it's hard to bolt this method on without substantial issues on hands in other notrump ranges.

The problem with 16-18 is that the frequency drops off very fast (i.e. 16 count is way more likely than 18) so making marginal game invite over 16-19 (with a hand that you would not chase a game opposite 15-17) tends to be a losing proposition.

I agree that superiority of MINI-NT is somewhat suspect, but I do agree with his two main conclusions: a) it does not pay to play wide ranges, either systemically or by liberal widening NT opening is a winner

Nevertheless the issue of "the field" is a serious one, if you notice in his stats 15-17 is superior than 14-16, but 12-14 is also better than 13-15, and thats why I think.

Playing non-standard ranges I think it is crucial to make to set "natural" borders and not playing it to wide. For example what RM are doing, playing 1NT good 13(i.e. at minimum a hand that would in standard accept invitation in 1m-2N sequence) to bad 16 (i.e. at maximum a hand that would not accept invitation in standard 1N-2N sequence) has a lot of merit.

So taking all that in consideration playing "standard" my personal preference is to play something like 14.96-16.95 range, except favorable 1st/2nd seat when I can not resist to push opponent around a bit by opening (11)12-14 NT. Playing strong club I would probably recommend something similar, but I might try to spice it up a bit since I have more options available.

[awm]

HCP 15: 4.42%
HCP 16: 3.31%
HCP 17: 2.36%
HCP 18: 1.61%

If we have 15-17, then the relative odds:

HCP 15: 43.8%
HCP 16: 32.8%
HCP 17: 23.4%
Expected HCP: 15.796


If we have 16-18, then the relative odds:

HCP 16: 45.5%
HCP 17: 32.4%
HCP 18: 22.1%
Expected HCP: 16.766

Honestly the difference in terms of the odds to have a minimum or maximum are very slight here. I just don't see any reason that 16-18 notrump would perform badly when it comes up while 15-17 performs really well. Of course 15-17 comes up about 1.5 times as often so it might be the winner for this reason, but I can't believe that opening 1NT 15-17 wins half an IMP on average every time it comes up while opening 1NT 16-18 loses .2 IMPs on average. That's a really big difference -- and one better explained by the identities of the players involved and their overall methods/age/level of recent experience at the top levels of the game.

[wclass___]

Mini NT - FTW!

[spotlight7]

Hi:

Mini NT may not be a IMP gainer at the top levels.

Meckwell used to play 10-12 1NT.

Regards,
Robert

[glen]

spotlight7, on May 12 2010, 06:21 PM, said:

... Meckwell used to play 10-12 1NT ...

For Meckwell I saw more gains with 13s upgraded into "14-16" than their 9-12 (9 when allowed) 1NT

[rbforster]

I strongly feel that, when NV, the best NT range is the lowest 3 point range against which my opponents play non-penalty doubles .

In this way, 14-16 might well fair better than either 13-15 or 15-17 since the former may suffer some unfortunate penalty doubles while the latter is just as safe as 14-16 from penalties but less frequent (both of the strong ranges winning by occasionally intimidating the opponents out of their partial).

[spotlight7]

Hi:

It is a very cold day in a certain area that is normally very hot when Meckwell downgrades. They state 'freguent upgrades' and they mean just that.

Regards,
Robert

[glen]

spotlight7, on May 12 2010, 08:10 PM, said:

Hi:

It is a very cold day in a certain area that is normally very hot when Meckwell downgrades. They state 'freguent upgrades' and they mean just that.

Regards,
Robert

your post is the first mention of downgrade - what are you discussing here?

[zenko]

It may seem like a bit of an overkill to discuss in detail merits of say 13.5-15.5 range versus 13-15, or 14-16, but I strongly believe that the choice of NT range is the most significant system decision you make, creating a lot of scorecard "traffic".
To answer what's the best range you have to take in consideration (at least) these 7 issues:

1) what is "the field" using, and how much we care about it
2) vulnerability
3) seat
4) how much HPC "flexibility" we allow
5) how much distributional "flexibility" we allow
6) can the rest of our system handle the range changes
7) are we more concerned about IMPs or MPs

Whats the optimal range will in part depend on all of these variables, and how we subjectively feel about them.

What we objectively do know for sure is that to open 1N is clearly beneficial, therefore the first step should be to use the most frequent range as starting point (but beware that most frequent range changes from seat to seat!), and adjust from there, keeping in mind above 7 factors.

As a result of that analysis I am sure one can build a resonably strong argument for using many different ranges, depending of the seat and vulnerability, but on the other hand I am also sure that the same analysis would convince many pairs to stick to mainstream 15-17 or 12-14 ranges.