[pilun]

Symmetric relay is simplest and maybe best when the limited openings are based on 4-card majors. (Like the symmetric responses to 1♣)

There are various opening structures.

In our method a 1♥ opening shows 4+♥, denies four spades, 11-15 pts
Over that, 1♠ is GI relay (11+), which messes a bit with the spade suit.
This impacts the responding structure and continuations.

1NT response is 6-10, which is a bit wide. 7-10 would be better.
A flat 6-7 count with three hearts can pass 1♥. However

♠QT65 ♥93 ♦KT6 ♣J954

is a bit awkward, particularly vulnerable, or at matchpoints. Opposite ♥Q852, maybe -200 vs nothing.
4-1-4-4 is even worse.

There are arithmetic issues. Opener with a flat 15-count will wonder about the 15 + 10 3NT that "everyone" else is bidding after a 15-17 1NT opening. Against that, raising with a 15-count will more often look silly, watching partner fail in 2NT with 15 opposite 6.

11+ hands relay, which leads to

1♥ - 1♠
2♣ - 2NT

where 2♣ is balanced. 2NT on 11 + 11 is not a good look. With that in mind, same flat 11-counts pass, rather than open 1♥. On marginal hands, we look at the 4-card heart suit. How will you feel if partner raises on three?

This is where the Precision 1♦ excels, with the nice split of 11-13, 14-16, 17+.

Another candidate for 1♥ - 1NT is a hand with five spades:

♠KT765 ♥64 ♦Q94 ♣K42

2-level responses are non-forcing "less than invitational", 6-carders, maybe a chunky 5-carder for 2m. Think "weak two".
They work well enough but the suit needs to be up to scratch because opener will sometimes pass with a singleton.
We have been known to make these bids with 3-card heart support but probably shouldn't.

♠KT43 ♥62 ♦J5 ♣AJ654

is an instructive hand. You don't fancy 2♣, so respond 1NT.
That won't be good if partner has a singleton club. What if partner bids 2♦?

The likely shapes for 2♦ are 3-5-4-1 and 3-4-5-1. Which is it?

With 2-2 reds, responder has to guess well to avoid the 4-2 fit.
Rather than guessing, we insist that opener's 2♦ rebid over 2♣ shows five of them.

When will responder raise to 2♥ with three trumps?
This is not Acol, where a 1♥ opening is either five of them, or a good hand or good suit.
Our 1♥ opening is a 4-carder roughly half the time, could be four small.

With that in mind, the three 4333s will almost always prefer 1NT.
The 5332s will tend to raise, more so with a heart honour and small doubleton.
2-3-(5+3) will always raise because the opponents are known to have 8+ spades. (1♥ denies four spades)
3-3-5-2 has the option of 1NT, then 2♥ over 2♣ by opener.
What about 3-3-2-5, or 4-3-2-4? (The balanced hands with a doubleton diamond)

Opener will always remove 1NT (to 2m) with a singleton. The four 5422 shapes have a choice. See later.
Responder can have a problem with 2-2 reds.

1♥ - 1NT
2♦ - ?

2♦ could be 5-4 either way (compare 1♥ : 2♣ : 2♦). What does responder do with

♠KT43 ♥62 ♦J5 ♣AJ654 ?

Our agreement is that responder always bids 2♥ with 2-2 reds (4-2-2-5 or 5-2-2-4)
Half the time that will be right. The other half, opener will want to escape the 4-2 fit.
2♠ shows 3-4-5-1, which will be fine because responder has a least four spades.
2NT on 1-4-5-3 is not great but "no fit, notrumps" plus responder can remove to 3♣ with five.

What of 2-4-5-2? Perhaps that hand should pass 1NT.
If you choose to rebid 2♦ and see 2♥, follow up with 2♠ and hope for the best.

This agreement means that a hand with three hearts and a doubleton diamond should always raise directly, because preference over diamonds shows 2-2.

Similar principles apply after 1♥ : 1NT : 2♣

Here responder can bid 2♥ with 3-2 (or 3-3) in opener's suits.

With 4-2-5-2 or maybe 5-2-4-2, responder bids 2♦ over 2♣ to show 2-2.

We play 1♥ : 2NT as GOOD limit plus with 4 trumps.
3♣/♦/♠ are invitational, good 6-carders.

3♥ is mixed. Typically 5=card support or chunky four, or short spades,

Transfer openings are popular these days, so

1♦ = 4+ hearts, maybe not denying spades(?).

It's great for responder to be able to respond 1♠ with spades.
1♥ as a relay is good too. Right-sides heart contracts.

Transfer openings give the other side more competitive options, which is not great.

Also, responder can't sit in 1♥. This is a more serous issue.

We like our 1♥ - 2♥ to show a good hand, say 7-10 TP.

With transfer openings, a weaker responder chooses between pass, 1NT & 2♥ A bit of a lottery.

[awm]

It seems like you have an awful lot of problems -- even without competition you're having awkward sequences when responder doesn't have an immediate game force. This opening style will also run into trouble if opponents intervene (responder with three-card major support is often a difficult one). Are you sure that this opening structure is worthwhile just to simplify the relays? I've always found that five-card majors work quite a bit better in real bridge (as opposed to a slam bidding contest where opponents always pass).

[pilun]

awm, on 2026-April-13, 01:30, said:

It seems like you have an awful lot of problems -- even without competition you're having awkward sequences when responder doesn't have an immediate game force. This opening style will also run into trouble if opponents intervene (responder with three-card major support is often a difficult one). Are you sure that this opening structure is worthwhile just to simplify the relays? I've always found that five-card majors work quite a bit better in real bridge (as opposed to a slam bidding contest where opponents always pass).

Thanks for that.

We have issues, though not an awful lot. We've played this method for 30 years.
The MAFIA opening style makes most relay auctions straightforward and "doubly" symmetric.

Relay pairs tend to spend too much time and effort on relay auctions at the expense of the more common natural auctions.
Bridge is a 4-handed game, so we don't mind the 2-level in the 4-3 fit, which is dangerous for both sides.
The denied major can be useful in a contested auction. Knowing there is no 4-4 fit in the other major, double can focus on the 5-3 and the 3-5.

We haven't noticed relay as GI being a negative. Having all 2-level responses constructive has proved useful.

Of course a Precision 1M opening is a winner when it occurs - better than us and a Standard 1M - but you pay by farming out the 40% of hands with a 4-card major. A 5-4 fit is good but so is a 4-5 or 4-6.

[foobar]

pilun, on 2026-April-12, 23:59, said:

Symmetric relay is simplest and maybe best when the limited openings are based on 4-card majors. (Like the symmetric responses to 1♣)
...

Transfer openings are popular these days, so

1♦ = 4+ hearts, maybe not denying spades(?).

It's great for responder to be able to respond 1♠ with spades.
1♥ as a relay is good too. Right-sides heart contracts.

Transfer openings give the other side more competitive options, which is not great.

Also, responder can't sit in 1♥. This is a more serous issue.

We like our 1♥ - 2♥ to show a good hand, say 7-10 TP.

With transfer openings, a weaker responder chooses between pass, 1NT & 2♥ A bit of a lottery.

Nick,

I am 100% behind awm's observation that five card majors simply work better in competition. IME relays work equally well with 5CM (a la Tarzan for example), and the same relay scheme (say 1♠ - 1N as GFR) can be repurposed over 1♣ - 1♥ (spades) - 1♠ (GFR) and so forth. So, it seems like the only debate here is 4CM vs. 5CM, and that's been going on for decades (with one clear victor IMO).

Someone like Richard is better suited to chime in on transfer openings, and several versions of Moscito did use natural openings (and were probably better off for it).

[foobar]

pilun, on 2026-April-13, 02:27, said:

Thanks for that.

We have issues, though not an awful lot. We've played this method for 30 years.
The MAFIA opening style makes most relay auctions straightforward and "doubly" symmetric.

Of course a Precision 1M opening is a winner when it occurs - better than us and a Standard 1M - but you pay by farming out the 40% of hands with a 4-card major. A 5-4 fit is good but so is a 4-5 or 4-6.

Once again, I posit that while not achieving the "doubly symmetric" standard, it's possible to construct a symmetric scheme over 5CM that comes really close. Regarding the 4-5 or 4-6 fit, aren't some of those hands already opened 1M in Precision? It seems like you are trading those with having to guess that to do over 1M - (2x/3x) - with a 3-card fit.

Quote

This is where the Precision 1♦ excels, with the nice split of 11-13, 14-16, 17+.

This is probably the one good thing that can be said about the Precision 1♦ opening. Several systems, including IMPrecision have tried to address the issues with it, and my preference is for the Nystrom-Upmark approach (balanced or major-minor canape).

If insisting on 4CMs I would strongly advocate removing the balanced hands out of the 1M openings, thereby making 1M unbalanced with 4+.