A sequence from a recent deal reminded me of some thoughts that I have had concerning Michaels sequences and gave me some new ideas as well.
Consider a simple start of 1♥-2♥-P, where partner has shown spades and a minor. The classic approach is for 2NT to ask for partner's minor, perhaps to sign off, but perhaps to gain more information for a game try. However, this leaves a strange problem. If you bid 2NT and then 3♠, partner will assume, correctly, that his minor was not interesting to you. Thus, this approach forces Responder to place the contract on the mere information of the suit alone. The problem is that there are many holdings where the minor is immaterial, such as when Advancer has Qxx/Qxx in the minors, but where the overall hand is still invitational.
Alternatively, Advancer might bid 3♥, but this seems to be a GF bid, especially (obviously) if the majors are reversed (1♠-2♠-P-3♠). So, this does not handle GI hands at all.
What about a simple 3♠ call? Well, the apparent usual meaning is "preemptive." This structure, however, seems to leave an inability to make any general invite and only allows a fit-specific invite.
Consider a hand like ♠Axxx ♥Axx ♦xxxxx ♣x. You "know" that partner likely has clubs. OK, but what is his strength? If he has ♥x ♦xx on the outside, you expect two quick losers in that suit. If his spade holding is good (maybe KQxxx) and his club holding decent (maybe Axxxx), then you expect to take five spades on dummy, the heart Ace, the club Ace, and, if trumps split 2-2, two club ruffs in hand, for a net of nine tricks. A tenth will come from establishing a late club when clubs are 4-3. So, if partner has ♠KQxxx ♥x ♦xx ♠Axxxx, game seems to be a fair bet. But, that's a perfect six-loser hand. Can't partner have something less remarkable, like ♠KQxxx ♥x ♦xx ♣QJxxx? Or, ♠Qxxxx ♥xx ♦x ♣KJxxx?
I just don't get the basic approach of simply asking partner for his minor and then guessing. The simple solution, at least for me, is to give up on the wild preference for always preempting to the maximum limit. Preempts are great, but a Michaels call sometimes does the trick without the need for Advancer to add his "I can preempt too" touch to the thing.
Consider this problem, as well, from the standpoint of when Partner has the very strong Michaels hand. You bid 2NT, and partner flags his minor (3♥ for clubs, 3♠ for diamonds). The upside is that partner has shown extra values. The downside is that you have no easy way to force game and agree upon a suit. It seems that about the only way to agree the major is to perhaps bid 4♥, a very ugly call, or four of the other minor, which is almost as ugly. The problem with the latter is that you then lose ability to make any sort of positive supporting bid for the minor.
This very Advancer's hand came up in Detroit. Fortunately, my partner and I had a general agreement ditching preemptive jump raises for constructive jump raises. So, I was able to bid 3♠ with the correctly-interpreted meaning of "I want to bid 4♠, but you might have a P.O.C. 2♥ call, in which case passing is fine." That approach seemed perfect with this hand. I show the game interest with a hand that is really not fit-dependent. If partner has diamonds, I have a side shortness values and an amazing double-fit. If partner has the more likely club holding, I have two clear tricks and a fourth trump coupled with shortness, which is nice. Either way, I have a strongly invitational hand, and partner should do the right thing. He will know that I am not so much fit-dependent, or I would bid 2NT and then select the right contract (hopefully).
As a bonus, if partner has the strong hand, I have focused trumps without ambiguity, such that any further calls are understood as spade-slam approach (if that is our course, as it was).
So, next issue. What are partner's options with the strong hand? Well, to define options, one must first decide what partner needs to be in slam territory. I like to think about this concern from the standpoint of assessing what "worst possible hand" provides slam hope. If we can reach slam opposite that hand, then all others will be easy.
My thought of the "worst hand" was ♠KQxxx ♥x ♦Ax ♣KJ10xx. Wow! A five-loser hand? How is that possible? Well, how about if Advancer has a simple hand like ♠Axxx ♥xxx ♦xxx ♣Axx? Five spade tricks is easy. One diamond makes six tricks. If the club hook works, which may well be odds-on as against Opener, then we have five club tricks, for 11 so far. Two diamond pitches on the clubs allows a diamond ruff on the short side, and 12 tricks. This amazing result occurs because Advancer covers two losers with the two Aces, one cover comes from the probable club hook, and a fourth comes from Advancer having six or fewer clubs+hearts and a fourth trump. Note how moving either Ace to hearts makes this slam hopeless on a diamond lead.
How about a similar holding, changing ♣KJ10xx to AJ10xx? That works, but now Advancer can have the King or Queen of clubs and the spade Ace. ♠Axxx ♥xxx ♦xxx ♣Qxx probably works. Again, the switch of the either cover to the heart Ace, or even to the diamond King, makes this slam hopeless.
It appears to me, then, that Advancer's exact cards are critical. Note how spade Ace and club Ace was necessary on the first, but the club Queen is enough on the second. Exact cards are necessary.
So, how does one work all of this out?
I have thought about this for a few days, and I have a suggestion.
First, a 4♣ call by partner, after 3♠, should show a presumably fit-dependent hand with clubs and spades. This seems obvious. If Advancer has two of the four top club and spade honors (♠AK, ♠A/♣A, ♠A/♣K, ♠K/♣A, ♠K/♣K, or ♣AK), he has it. RKCB seems called for. If Responder has two or three tenuous cards, like perhaps one of these four cards (♠A, ♠K, ♣A, or ♣K), plus a secondary blacks-suit card (♠Q/♣Q), plus the diamond Ace or King, he cues 4♦; if two tenuous cards only, he declines a 4♥ LTTC bid. If Responder has two tenuous cards but no diamond control, he makes a "LTTC" call of 4♥. Note that Advancer has no way to show or deny a heart control, but the heart Ace might be a third tenuous card justifying acceptance of a LTTC call. In other words, Advancer will note cooperate in a fit-dependent sequence unless he has at least one clear key cover and one tenuous card, but the holding of these two PLUS the heart Ace will merit acceptance of a LTTC call. Note also that 4♣ clearly implies heart control.
Second, partner bids 3NT with a diamond-oriented fit-dependent hand. This is huge, in my opinion. Advancer needs to be given space to cue a control in the other minor, and 3NT gets below 4♣. If Advancer does cue 4♣, this shows the same tenuous holding as before. However, partner can cue 4♦, repeating his suit, to suggest a need for a heart control (Advancer can bid 4♥ to show the heart control but suggest a tenuous internal card), or 4♥ as LTTC with a heart control. Same stuff, then, but an ability to check on a heart control when tenuous. Advancer can also bypass 4♣ (no club card) to cue 4♦ with two tenuous cards, allowing 4♥ as LTTC, presumably needing also a heart control (heart Ace, spade King, and club Queen, for instance). Advancer's 4♥ would seem to show the heart Ace and one key card.
These two bids and follow-ups might very easily need tweaking and improvement, but the idea is to handle fit-dependent hands in a way that maximizes space for description.
What, then, do 4♦ and 4♥ cover? Well, the one gap is the club-fit-dependent without a heart control. It seems to me that these should be somewhat split between 4♥, 4♦ with a hand that would decline a 4♥ LTTC response, and 4♦ with a hand that would accept a 4♥ LTTC response. One should seek primes, one should seek a high tenuous count, and would should accept a relatively low tenuous count. Just to agree on something, then, it seems that 4♥ is strongest, 4♦...P weakest, and 4♦...OK middlish.
So, what about handling hands that are not fit-dependent? Simple. One option is RKCB directly, obviously. The other is to get whatever information you can from whatever call is most likely to lead to an intelligent sequence. In other words, treat hands that are not fit-dependent as if they were fit-dependent.
1♥-2♥-P-3♠ = constructive invite
3NT = diamond slam try
4♣ = club slam try with hearts controlled
4♦ = mild to middling slam try with hearts not controlled
4♥ = strong slam try with hearts not controlled
Advancer then cues his values and value types.
So, what if hearts is the major? That's next...