[david_c]

Below is a copy of an article I've written, which I was hoping might provoke some discussion here. Sorry if it reads like a thesis (well, I need the practice!). Any comments out there?

Why I don't like Precision

I play Precision myself, and, like most other people who play it, I'm well aware of its strengths and its weaknesses. The strengths are the 1H and 1S opening bids, and constructive auctions after 1C. The weaknesses are competitive auctions after 1C and 1D, and constructive auctions after a 2C opening. I hope this is all very uncontroversial! As such, it would be easy to say, "I don't like Precision because the 1C and 1D bids are vulnerable to pre-emption." But that would be to overlook the fact that all systems have their strengths and weaknesses, and without considering both of these it's not possible to come to any conclusions about whether the system is a good one.

So I hope to do something different - my argument is that at the heart of the system there is a slight design flaw, which results in the system not being as good as it ought to be.

I want to look at the choice of opening bid with various hands. It will help to divide opening hands into five categories. One category is "balanced"; unbalanced hands are divided into "spades", "hearts", "diamonds" and "clubs" hands, depending on the longest suit (or more precisely, depending on which suit would be opened playing a natural system - so 5=2=5=1 hands are "spades" hands rather than "diamonds"). Let's ignore 4441 hands for the time being. Now, in the version of Precision that I play (which is fairly standard), the opening bids are defined as follows:

balanced: 11-13HCP open 1D; 14-16HCP open 1NT; 17+ HCP open 1C.
spades: (10)11-15HCP open 1S; 16+ HCP open 1C.
hearts: (10)11-15HCP open 1H; 16+ HCP open 1C.
diamonds: 11-15HCP open 1D; 16+ HCP open 1C.
clubs: 11-15HCP open 2C [or 1D]; 16+ HCP open 1C.

The thing to notice here is that the strength required to open 1C is almost independent of the distribution of the hand. This is what I perceive to be a problem with the system. Observe that the five different types of hand described above essentially correspond to the five different limited opening bids 1NT, 1S, 1H, 1D and 2C. A priori, there is no reason why the upper limits for these five bids should all be equal. But let's first consider the reasons why making the upper limit the same for all types of hand might be a good thing. As I see it, there are only really two reasons:

1. Simplicity. That is, it makes the system easier to remember, and it makes the description of the 1C bid nice and short when you alert. Needless to say, in terms of pure bidding theory, this doesn't come into it at all.

2. When 1C is opened, it helps responder to judge the likely strength of opener's hand. For example, responder has a fairly good idea of what strength he needs in order to force to game. This certainly is very important - it's very easy to make the system unplayable by tinkering with it so that this advantage disappears. For example, suppose you were to decide that you hated the Precision 2C opener so much that you would open those hands 1C (in addition to the 16+ HCP hands). Leaving aside for a moment the fact that it's virtually impossible to come up with a sensible way of bidding after such a 1C opening, you would have terrible problems in competitive auctions. The main reason is that a 1C opener would nearly always have at least 16+ HCP, but it would be extremely dangerous for responder to bank on that because of the weaker possibility. [More generally, it helps if the most frequent hand type possible for a bid is also the weakest. The 1C opening in Polish Club is particularly good in this respect; the 1C opening in weak-NT-short-club is bad.]

There is a third possible reason - coincidence. Perhaps it really is true that the right limit to put on the 1NT, 1S, 1H, 1D and 2C bids is approximately 15HCP in each case. Let's investigate this by considering each of the five hand types separately. There are two factors to consider.

1. How useful is it for the limited opening to be limited?

balanced hands - It is essential that the stength required for opening 1C with a balanced hand is relatively low. If it was made higher, then either you would have to play a wide-range 1NT opening, or you would have to open a nebulous 1D on lots of hands. This would be highly undesirable.

"spades" and "hearts" hands - Limited 1S and 1H opening bids are very useful indeed. Indeed you could say that these bids are the reason you play a strong club in the first place. Even so, you would have to rate a limit of the order of 15-16HCP as "useful", rather than "essential" as in the case of balanced hands. (You'd find it difficult to argue that a Standard American 1S or 1H opening bid is unplayable.)

"diamonds" hands - These definitely get the wooden spoon. It is certainly useful to have a small range for 1D, but much less so than for 1S and 1H, the reasons being:
- responder often needs to check on major-suit fits before knowing what the final contract should be, which is not a problem when the opening bid is 1S or 1H;
- the shape is less well defined (another reason why it is difficult for responder to set the final contract quickly);
- there is more space available after 1D than after 1S or 1H.

"clubs" hands - It's so difficult to bid constructively after a Precision 2C opening that it's absolutely essential for 2C to have a narrow range - even more so than in the case of balanced hands.

2. How easy is it to bid the hand after opening a strong 1C?

balanced hands - These are fairly easy to bid, particualrly if playing some sort of relay system. You do get some problems if the opponents pre-empt, of course, but sometimes you have the option of passing, which is obviously more attractive with a balanced hand than with an unbalanced hand.

"spades" hands - These are easy to bid: you usually bid spades at the lowest available level.

"hearts" hands - These are not as easy to bid as "spades" hands, because you lose a level of bidding if someone bids spades.

"diamonds" and "clubs" hands - These are relatively hard to bid after 1C, particularly if partner makes the negative response. "Diamonds" is slightly worse than "clubs" because you have less space after your diamond rebid.

In summary,

Is a very limited opening useful?
clubs - very useful
balanced
spades
hearts
diamonds - not very useful

Are you likely to do well if you open 1C?
spades - relatively well
balanced
hearts
clubs
diamonds - relatively poorly

If these two factors were mostly working in opposite directions, then playing the 1C requirement as roughly the same for all five hand types would make a lot of sense. But immediately we can see that one of the five sticks out - diamonds. You wouldn't want to rush to conclusions, but there is a clear indication here that it might be a good idea to make the upper limit for a 1D opener higher than for 1NT, 1S, 1H and 2C.

So suppose we change our 1C and 1D opening bids to take this into account. Would we now be in trouble because responder is unable to judge the strength of a 1C opener? Clearly not - taking out minimum hands with diamonds makes hardly any difference to the expected strength of 1C. So although we found earlier that there was a good reason to make the strength required for 1C the same for all hand types, we're keeping that advantage here.

In practice, however, requiring 18 or 19HCP to open 1C with diamonds is rather pointless. All the sequences after 1C would be unchanged, except that responder could rely on a few extra points when opener turned up with diamonds. Sometimes you would gain by opening 1D with up to 18HCP, and sometimes you would lose. (I think you would gain more than you would lose, but that's rather difficult to prove.) In order to make a real differece to the system, we'd have to go further and take enough hands out of 1C that the system after 1C could be improved.

Here we enter the realms of personal preference. One idea is to require a 1D for all "diamonds" hands short of a game force. Certainly the system would no longer be Precision - maybe you could call it "strong club with unlimited diamond". Bidding after this 1D opening would not be too difficult - certainly better than in natural systems because you have a 2NT rebid available to help with some of the strong hands. The real advantage is that you've freed up opener's 2D rebid after a 1C opening, which could be used for, say, strong 4441 hands.

Personally, I like to look a bit closer at each of the five lilmited opening bids. My feelings are that the requirement for opening 1C ought to look something like this:

balanced - 15HCP
clubs - 15HCP
spades - 16HCP
hearts - 18HCP
diamonds - almost game

[Note in particular that the low limit for balanced hands means that you don't need to open 1D with any balanced types.]

However this is in a perfect world where you don't have to worry about constructing a system that works over 1C! In practice life is easiest if you increase the requirement for "spades" and "hearts" hands to something more like Polish Club:

balanced - 15HCP
clubs - 15HCP
spades - 18-19HCP
hearts - 18-19HCP
diamonds - GF

And finally we've reached my pet system which is a good deal closer to Millennium Club than it is to Precision. But, as I said, I don't like Precision.

[mr1303]

Some interesting thoughts there David.

However, there are a few difficulties, especially since I play strong club with you.

1) Having up to 18-19 HCP hands in 1S and 1H will presumably restrict that hands you open these to unbalanced hands. However, how do you define unbalanced?
Are 5422 hands unbalanced? I would like to open 5422s as NT hands, but I can see this leading to difficulties

2) Putting GF hands with diamonds into a nebulous 1D opening would, IMO, force partner to respond on virtual rubbish opposite a likely 11-13 NT. This makes defenders lives easier if the 1D opening becomes almost forcing as well.

3) If you take out the 11-13 NT hands out of 1D and put them into 1C, this will kill the symmetric relays that we play, since we either require a GF opposite 11-13 balanced, and our 1D response becomes WAY overloaded or we revert to complete Polish club, which has no relays

[Double !]

I believe that raising the range of 1M opener to 18-19 defeats one of the primary purposes of playing a forcing club (or diamond) system. The advantages of playing limited openings are 1) eliminating rebid problems with strong 1-suited hands that are too strong for a nonforcing jump-rebid of the suit, and 2) addressing problems with having to open 1 on strong 2 (or 3)-suited hands that aren't good enough for a 2C opener. IMO, if you eliminate that advantage, then you defeat the purpose of playing a strong, forcing 1C or 1D system that allows opener to handle these in-between types of hands that are problematic in sayc and 2/1-style systems. I agree, playing a forcing C or D system has multiple design flaws, many of which are attempts to compensate for the loss of the ability to bid certain hands naturally. By the way, there are many problems inherent in playing 1m as 3+ as is done in sayc or 2/1 or K-S, and there appear to be many contributors to the forum who view 4-card majors positively. Will be curious to read your suggestions for improving precision. BTW, I am a precision player from eons ago.

[MickyB]

Mark, I think David was suggesting a 12-14 NT, with 15+bal opening 1♣.

David:
IMO 16 point hands are THE hands you want to open 1♣. You have already shown the strength of the hand, which might be difficult if you had opened with a suit.

I'm not convinced that a natural 1♦ opener doesn't benefit much from being limited - it allows partner to raise with a weakish hand and an outside 4 card major, whereas playing unlimited openings you normally have to bid the 4 card major for fear of missing game.

I showed you a system a while back which had 1♥ as 5+ cards 11-19 and 1♠ as 4+cards 10-15, which allowed 1♥ to be a strong relay after a strong club and negative response. Tysen has said on here that his simulations suggest that the 1♥ and 1♠ openings are actually better reversed; I don't know to what extent this took into account the effects on the 1♣ opener, or how a response structure would work.

I played something similar to Millennium Club a few months ago. I decided that I preferred 1♥ and 1♠ to be unlimited to simplify the 1♣ opener; I think including strong but not GFing hands with a primary major could make the opening much less successful. Another interesting alternative is to make 1♣ 11-13 or 17+; 1NT and 2♣ are 14-16 nat. IIRC Misho said that Kalin-Rumen played this vul, with Millennium Club or similar NV.

[david_c]

mr1303, on Jan 22 2005, 03:27 PM, said:

Some interesting thoughts there David.

However, there are a few difficulties, especially since I play strong club with you.

lol don't worry Mark, I wasn't intending to put any of these ideas into practice - our system is still a huge improvement over what everyone else in Manchester plays

[david_c]

MickyB, on Jan 22 2005, 06:45 PM, said:

I played something similar to Millennium Club a few months ago. I decided that I preferred 1♥ and 1♠ to be unlimited to simplify the 1♣ opener; I think including strong but not GFing hands with a primary major could make the opening much less successful.

Weren't you playing Millennium Club with transfer responses to 1C, so that 1D, 1H and 1S showed 0+ HCP? If so I can see why it's a lot simpler if the major-suit hands are restricted to GF. My 1H and 1S responses are natural and promise enough for game opposite an 18HCP hand, so a lot of the problems disappear.

Anyway, I wasn't intending to try and defend my own system here, just to say what I didn't like about Precision The final remark I made, where I said I prefer to require 18HCP to open 1C with a major suit, doesn't follow logically from the preceding arguments - it's just my personal preference. Yes it has problems - it makes the major-suit openings a lot worse (well, they're supposed to be identical to Polish Club, so they're not that bad!), but of course opening 1C on the 16-18HCP hands has its down side too.

[david_c]

mr1303, on Jan 22 2005, 03:27 PM, said:

2) Putting GF hands with diamonds into a nebulous 1D opening would, IMO, force partner to respond on virtual rubbish opposite a likely 11-13 NT. This makes defenders lives easier if the 1D opening becomes almost forcing as well.

Ach, yes, I forgot to say that. I wouldn't advocate opening 1D on more than about 18HCP unless it promised a 4-card suit, for precisely those reasons. But anyway, GF hands would still be opened 1C (when I said diamonds=GF that was intended to be the lower limit for 1C, not the upper limit for 1D).

Now, I'm finding it very difficult to resist the temptation to reply in detail to each line of each post. I'd better make this the last one for the evening.

[the hog]

Some interesting thoughts here. The fundemental design flaw in Precision is the attempt to meld 5 card majors and a strong C:

We are currently playing the following:
1C 15+ most
1D/H 10-14 single suited in H/S or 10-17 if 2/3 suited, 4+ cards. The latter higher range is so as to be able to cope with intervention
1S = as above with D. denies a M
1NT = all 12-14 balanced, including ALL 5332 shapes.

This allows for a vastly simplified symmetric structure, where single suited resolution starts at 2H for high shortage. Resolution also excludes all 5332 shapes, and thus you are able to incorporate 6322/7222 min and max and 6331 min/max responses. (in case you are wondering, 1C 1N includes all 4432/4333/5332/4441 shapes).

The structure works well and has been tried in international competition with success.

Ron

[flytoox]

The_Hog, on Jan 23 2005, 12:55 AM, said:

The fundemental design flaw in Precision is the attempt to meld 5 card majors and a strong C:

Ron, can you explain why this is so? My impression is that this is the advantage of precision. People talk abou the flaw of precision in 2C, 2D opening. I have never heard of this claim before.

[EricK]

flytoox, on Jan 23 2005, 10:03 AM, said:

The_Hog, on Jan 23 2005, 12:55 AM, said:

The fundemental design flaw in Precision is the attempt to meld 5 card majors and a strong C:

Ron, can you explain why this is so? My impression is that this is the advantage of precision. People talk abou the flaw of precision in 2C, 2D opening. I have never heard of this claim before.

I don't wish to pre-empt Ron's response as I am not a Strong Club player, but aren't these bids necessary in Precision because you open 5 card majors. So any weakness in them is as a result of trying to play a strong club with 5 card majors.

Eric

[the hog]

Yes, Eric is correct. The nebulous 1D opening and the opening 2C together with a possible 4 card major could both be alleviated by opening 4 card majors. The 2D opning is also a "fix" based on the fact that 1M has to be a 5 card suit in Precision.

For a better system, and one not based on say Moscito style relays, have a look at von Arnim's - Auken's strong Club system.

Ron

[flytoox]

I understand your point. However, this logic applies to most system. If you open 4-card major then there are other problems. If you play standard or 2/1, there is still problem. Both minor openings are semi-neublus. All one level openings are extremely unlimited. Those 16-18 vs 7-9 hands are hard to bid game.

No system is perfect. The only problem is which tradeoff you are willing to take. Precision is by far one of the most successful systems in competitive bridge. I dont think a system with a fundemental flaw could acheive such success.

[david_c]

The_Hog, on Jan 22 2005, 07:55 PM, said:

Some interesting thoughts here. The fundemental design flaw in Precision is the attempt to meld 5 card majors and a strong C:

That may well be true, though it's quite difficult to persuade people to believe it if they've grown fond of 5-card majors. My point was that even if you restrict your attention to strong-club-five-card-major systems, it's simply not possible that Precision is optimal. Interestingly, if you try to do the same analysis on moscito, you find it doesn't have the same problems. Which just reinforces the view that moscito is an extremely well-designed system.

Having said that, I don't think I'll ever be a moscito player. I can see it's a very good system, quite possibly the best there is at the moment, but it doesn't suit my temperament. Also, most of my opponents in real life are hardened Acol players, and providing disclosure of transfer openings, relays and suchlike is absolute hell.

[Free]

Yes, MOSCITO is a very well designed system imo. However, there are other problems than with precision: you'll have to be a good player before you'll get the advantages out of the system. For example, you'll have to learn how to play 4-3 Major fits as good as possible, because you'll end up in a lot of them

I once decided to create a Real Diamond Precision system, which is still quite natural to keep explaining stuff to opps quite simple. I posted it in one of the threads here, but here's a summary:

1♣ = 15+ any
1♦ = (9)10-14, unbalanced, 4+♦, can have longer ♣
1♥ = (9)10-14, unbalanced, 5+♥ (or 4-4-1-4 as only exception)
1♠ = (9)10-14, 5+♠ (can be balanced)
1NT = 11-14, balanced, no 5 card ♠ (5 card ♥ is possible)
2♣ = (9)10-14, 5+♣ & 4M / 6+♣ singlesuited

After ALL openings I have a full relaystructure, to get the some of the advantages from MOSCITO, and have nice slam bidding options.

Result: you open all suits with a high probability of being a 5 card suit (♦ is only a 4 card when you have a 4441 or longer ♣, otherwise it's a 5+ card), the 1♠ and 1NT openings are used also as a preemptive weapon, and ♦ hands are naturally bid.
Biggest disadvantage: still a lousy 2♣ opening...

I noticed that, when you want to combine 5 card Majors with strong ♣, you either have to sacrifice the 1♦ opening to become non-natural, or you'll need to use an overleaping 1NT range, so the balanced hands get out of the 1♦ opening, and an extra trick to bid 4-4-1-4 hands (2♦ opening or bid it as a 5 card ♥).

The only way to get rid of that 2♣ opening is to put it in the 1♣ opening (which makes this quite unbiddable imo), or play 4 card Majors (which MOSCITO does).

[luke warm]

flytoox, on Jan 23 2005, 06:12 AM, said:

I understand your point. However, this logic applies to most system. If you open 4-card major then there are other problems.

that might be true fly, but what if the 4 card opening guaranteed another, usually longer, suit?
1c= 16+ if balanced, 17+ if a Major, 18+ if a minor oriented hand
1d= 5x4x; x54x; xx44+
1h= 54xx; x45x; x4x4+
1s= 45xx; 4x5x; 4xx4+
1nt= 12-15 balanced, all 5332 and no 54xx hands
2c= all 3 suited hands
2d/h/s= all one suited hands, 6+ long and limited (11-16/17)
2nt= one suited club hands, 14-17
3c= the same but 11-14

[mike777]

In my limited experience with Honeymoon Moscito it seem the advantage is the ability to open very light 9hcp hands and have hidden longer minors. Also since I love 4 card major, this is very nice. It also has all the advantages and weakness of relay system. A disadvantage seems to be in many competitive auctions I am unsure of opener's length in their suits.

Have only seen Auken club in action a couple of times in The Bridge World magazine but at first blush looks great. Where can I find more detailed information?

[EricK]

mike777, on Jan 23 2005, 05:05 PM, said:

Have only seen Auken club in action a couple of times in The Bridge World magazine but at first blush looks great. Where can I find more detailed information?

http://public.aci.on...en-vonarnim.pdf

[flytoox]

luke warm, on Jan 23 2005, 04:52 PM, said:

flytoox, on Jan 23 2005, 06:12 AM, said:

I understand your point. However, this logic applies to most system. If you open 4-card major then there are other problems.

that might be true fly, but what if the 4 card opening guaranteed another, usually longer, suit?
1c= 16+ if balanced, 17+ if a Major, 18+ if a minor oriented hand
1d= 5x4x; x54x; xx44+
1h= 54xx; x45x; x4x4+
1s= 45xx; 4x5x; 4xx4+
1nt= 12-15 balanced, all 5332 and no 54xx hands
2c= all 3 suited hands
2d/h/s= all one suited hands, 6+ long and limited (11-16/17)
2nt= one suited club hands, 14-17
3c= the same but 11-14

Jimmy, 4-card will do well in uncontested auctions. However, in the competitive auctions, 5-card major will have some advantage over 4-card major.

Of course, 5-card major has its flaws. My point is that to say precision has a fundemental flaw is too exaggerate to say the least.

[Flame]

Nice post, Im playing polish and i was trying to convince my partner to play the 1D opening to have a higher limit then our 1h/1s. I have seen some polish ccs for that.

edited
-------
here is a link to "your" system.
http://www.ecatsbridge.com/documents/files...z-miszewska.pdf

[the hog]

"Jimmy, 4-card will do well in uncontested auctions. However, in the competitive auctions, 5-card major will have some advantage over 4-card major. "

Disagree strongly! It is precisely in contested auctions that opening 4 card Majors has a huge advantage over 5 card major systems. 5 card Majorites have to open 1D, whereas 4 card Majorites can immediately open the Major. Note that this also has a slight pre emptive effect.

Off topic: One of the most effective of all openings was in the early versions of Moscito, where 1D or in later versions 1NT, showed both majors. When you opened something else you immediatley knew whether the opps had a Major suit fit or not.