Join Bridge Winners
Forcing or Semi-Forcing 1NT

Traditionalists mostly play a forcing NT.

In recent years, more and more play semi-forcing (experts so favored, for the first time in BWS 2001) which means non-forcing on a minimum unbalanced hand.  The influential teacher Larry Cohen so advises, moderately so, in LC Standard for advancing players.

I was first exposed to non-forcing (then sometimes called "almost forcing") in "Better Bidding by Bergen," circa 1985). I was a newer player and that notion made great sense to me and I've often so played (sometimes, at that time, unknown to partner) with pretty good results. I still prefer that treatment in a regular partnership. Marty's view, in part, was that he would be rejecting a game invitation in three-of-a-major and not bothered by playing in 1NT for overtricks as opposed to a possible touch-and-go three-level partscore with two balanced hands.  I doubt that his view has changed.

Later in this piece, I've pasted in most of the Bridge Winners articles on the subject to date. As I recall, they are peppered with comments from a number of notable experts. I'm not sure, but I hope I've included the thread that includes comments on the sometimes expert practice of responding 1NT on "air."

My purpose here is to simply display, as a baseline for consideration, what a simple double-dummy simulation shows in regard to the observations of "we might miss a major-suit game" or "we should score better in spades with an 8-card fit."

My software cranked out 10,000 deals, with these limited specs:

Version A: Any opening hand (presumably first not to pass) with exactly five spades and 332 otherwise and 11-13 HCPs.  Any responding hand with exactly three spades and a balanced hand (could be 5332) with 10-12 HCPs.

Version B: Same as above except responding hand is limited to 10-11 HCPs.( I'm interested in how much one point matters.)

I did not try to constrain either opponent from entering the auction which of course is not the real world.

Here's what I found:


In A:

Responder if in 1NT and Opener in 3S, cumulatively:

          7 tricks  8 tricks  9 tricks  10 tricks  11 tricks

1NT:    79%     53%       24%      6%         1%

3S:      99%     90%      60%      23%       4%


In B:

Responder if in 1NT and Opener in 3S, cumulatively:

1NT:     75%    47%     18%        4%          1%

3S:       98%    87%     53%       16%         2%


Clearly, there is considerable risk in making 3S. I am somewhat surprised that there is a fair amount in making 1NT.

You'll note that we're looking at two balanced hands with a spread of 21-25 pts in (A) and 21-24 pts in (B).

You're aware, too, that "hand evaluation" is ignored.

Importantly, eliminating opposition constraints is likely to be a major flaw in drawing any firm conclusions.

(Surely, a goodly number of deals would be eliminated.)


Perhaps a more experienced simulator would be willing to help out here.


These are the other links I referred to:

(1) Integrating the Semi-Forcing NT Bob Ciaffone Jan. 8, 2012 26 Comments

(2) Revisiting a "Semi-Forcing" INT Discussion Ed Judy Oct. 11, 2014 20 Comments

(3) An obscure forcing NT sequence Michael Albert Nov. 11, 2013 14 Comments

(4) A question regarding the semi-forcing 1NT Szőts Gábor Dec. 12, 2015 30 Comments

(5) Finding hearts after a semi-forcing 1NT Ed Herstein June 11, 2013

Getting Comments... loading...

Bottom Home Top