Jacoby 2NT

A raise that settles the game in one bid and leaves the three level free.

Get it for iPhone & iPador play in your browserboth free

What the Bid Promises#

Partner opens 1♥ or 1♠. You hold four-card support and enough for game. 2NT is Jacoby: it is artificial, it has nothing to do with no trumps, and it says two things at once — four or more of your major, and values enough to force to game.

The alternative, raising straight to 4♥, tells partner exactly the same thing about the contract and nothing at all about the hand. Jacoby 2NT arrives at the same place with the entire three level still available to explore, which is the difference between bidding a game and finding out whether there is a slam.

About thirteen support points is the working floor — that is high-card points with the value of your shortage added, since a fit has already been established and ruffing values are real once it has.

How Opener Answers#

The auction is game-forcing, so opener is not deciding whether to bid game. Opener is describing, and does it in a fixed order of priority:

Read the table top to bottom and the logic is one rule: say the most useful thing you have, and shortness beats points. Rase's robots reply in exactly this order, so an auction of 1♠ – 2NT – 3♦ is not a diamond suit; it is a singleton or void, and it is the most important card in the hand.

The last row says "four of the trump suit" because that is where game is from a standing start. At a rubber table it may not be — see what a part-score does to all of this below.

Why Shortness Outranks Strength#

It looks wrong to show a singleton ahead of a good hand, and it is one of the few places in bidding where the counter-intuitive rule is clearly right.

Once a fit is agreed and the values are known to be enough for game, the only question left is whether there is a slam, and slams are not decided by the thirteenth point. They are decided by whether your losers can be got rid of. A singleton opposite partner's three small cards turns three losers into one; the same singleton opposite partner's A-K-Q is worth nothing at all and may be worse than nothing.

Nothing else opener can say carries that much information, and no amount of point-counting recovers it. So it goes first, and the extra values wait for the next round — where, the auction being forcing, they are still perfectly safe.

What a Part-score Does to the Ask#

Jacoby 2NT buys information, and it pays for it with a round of bidding. At duplicate that round is free, because game is always four of the major. At a rubber table it is not: a part-score moves where game is, and once it has moved far enough there is nothing left to buy.

The arithmetic is the whole argument. With 40 or more below the line, two of your major scores the game outright. But 2NT is the cheapest bid the convention has, and the cheapest answer to it is three of the major — so asking the question guarantees you a level higher than the contract you were trying to reach, and hands the defence an extra trick to attack for a trick score you had already banked. Raise to the level that scores it and take the game.

With only 30 on, game is three of the major, which is exactly where the reply ladder settles anyway. The relay costs nothing there, so it is still worth bidding.

The exception is a genuine slam. A made game is worth bidding past when thirty-three points are there opposite partner's minimum — not opposite their maximum, which is an invitation, and an invitation does not buy the room. On that hand the convention is still the right start.

Rase's engine applies both halves. Responder checks the pad before choosing between 2NT and a direct raise, and if the relay does get started — on a slam hand, or from a part-score small enough to leave the room — opener's reply stops at the level that scores the game rather than at a reflex four. Shortness is still shown when it is cheap enough to stay under that level, and swallowed when it is not: over a heart fit, 3♣ and 3♦ are free and 3♠ is not.

It Does Not Apply in Competition#

Jacoby 2NT lives in an uncontested auction. If your right-hand opponent has overcalled between partner's opening and your call, the convention is off for that hand — Rase's engine will not bid it over interference, and does not read a 2NT there as Jacoby.

There is a good reason. Competition gives you better tools for the same message: a cue-bid of their suit shows a good raise, a jump raise is preemptive, and a negative double handles the hands with the other major. It also takes away the room that made Jacoby 2NT worth playing — the point of the convention is the free three level, and an opponent bidding at the two level has already spent some of it.

So in competition, raise. Show the fit and the level directly, and leave the describing ladder for the auctions that have room for it.

On in 2/1, Off in ACOL#

Jacoby 2NT is on by default in 2/1 and off in ACOL, and unlike most of the switches in this app it is not offered in ACOL at all by default.

The reason is what the two systems' openings mean. In 2/1 a 1♥ or 1♠ opening promises five cards, so four-card support is a known nine-card fit and a game-forcing raise is a safe thing to describe. In ACOL, where a major opening can be four cards long, four-card support is only an eight-card fit and might be less; a bid that force to game on that basis is doing more forcing than the information justifies.

With the switch off, 2NT over partner's major reverts to its natural meaning — a balanced invitational hand, around 10 to 12, with no guarantee of support at all. That is a genuinely different bid, and it is the one your robot partner will read if you have ACOL selected.

The Hands That Should Not Use It#

Two limits are worth stating, because the convention is attractive enough that people stretch it.

Three-card support is not enough. The bid promises four. With three and game values you have other routes — a two-over-one in a side suit, then support — and taking the Jacoby route instead tells partner about a fit that is one card shorter than advertised, exactly when partner is deciding on a slam.

A flat thirteen is a poor choice even when it qualifies. The bid launches a slam investigation, and 4-3-3-3 opposite a balanced opening has no slam in it. Nothing is gained by two rounds of describing before signing off in the game you could have bid at once, and something is lost: you have told the defence which suits you are short in on a hand where you needed them guessing.