[steve2005]

finally get a hand where Gerber applies and Ben screws it up
responded 4♦ 0 or 4 aces as i have 1 that means 0
but it had 3 aces. if it shows 3 aces i bid 7N
My link

[pescetom]

From a Ben point of view, I seem to remember this bug was already documented here.
Seeing as Lorserker has miraculously resuscitated, maybe he could confirm

From a bridge point of view, I think Gerber was (yet again) inappropriate.
You know there are 32-34 points on the line and you have 6 running diamonds.
6NT looks almost inevitable and does not depend upon number of Aces (barring the freak of opps holding 2 Aces and no other honour, even then they have to find both).
But 7♦ also looks likely and well worth investigating.

[steve2005]

I thought i saw that all bids above 3NT are not defined. If so hard to bid slams scientifically.

[johnu]

pescetom, on 2026-June-10, 14:25, said:

From a bridge point of view, I think Gerber was (yet again) inappropriate.
You know there are 32-34 points on the line and you have 6 running diamonds.
6NT looks almost inevitable and does not depend upon number of Aces (barring the freak of opps holding 2 Aces and no other honour, even then they have to find both).
But 7♦ also looks likely and well worth investigating.

Why is Gerber inappropriate when there is a chance you are missing 2 aces, especially playing with Ben? OK, one reason would be that Ben makes random responses to Gerber. BTW, sometimes you won't have 12 tricks even if the opponents don't cash their 2 aces.

[thepossum]

Is this AI? Near enough is good enough

[pescetom]

johnu, on 2026-June-10, 17:37, said:

Why is Gerber inappropriate when there is a chance you are missing 2 aces

We've been through this before, here is what I wrote at the time

Quote

#
# Does it make sense to check if we are missing 2 cashing tricks?
#

produce 1000

Nmin=15
Nmax=17
Smin=17
Smax=17

Open1NT = shape(north, any 4333 + any 4423 + any 5332) and hcp(north)>=Nmin and hcp(north)<=Nmax
Raiseto6NT = shape(south, any 4333 + any 4423) and hcp(south)>=Smin and hcp(south)<=Smax

EtopS = top2(east,spades)
EtopH = top2(east,hearts)
EtopD = top2(east,diamonds)
EtopC = top2(east,clubs)
WtopS = top2(west,spades)
WtopH = top2(west,hearts)
WtopD = top2(west,diamonds)
WtopC = top2(west,clubs)
EWaces = aces(east) + aces(west)
NSaces = aces(north) + aces(south)
EWkings = kings(east) + kings(west)
EWhcp = hcp(east) + hcp(west)

OffAKany = ((EWaces == 1) and (EWkings == 1))
OffAKsame = (EtopS + WtopS == 2) or (EtopH + WtopH == 2) or (EtopD + WtopD == 2) or (EtopC + WtopC == 2)

action frequency "EW hcp" (EWhcp,0,9), frequency "EW Aces" (EWaces,0,4), frequency "EW have an A and a K" (OffAKany,0,1), frequency "EW have AK in same suit" (OffAKsame,0,1)


condition Open1NT and Raiseto6NT

#### for debug or skeptics, uncomment and reduce Produce to a small number
#action printes (" EtopS: ", EtopS, " EtopH: ", EtopH," EtopC: ", EtopC," EtopD: ", EtopD," WtopS: ", WtopS," WtopH: ", WtopH," WtopC: ", WtopC," WtopD: ", WtopD, " OffAKsame:", OffAKsame, " NSaces:", NSaces, " EWaces:", EWaces, \n), printew

[johnu]

pescetom, on 2026-June-13, 14:30, said:

We've been through this before, here is what I wrote at the time

Also, in real life, many players will upgrade very good 14 counts to 1NT. So there are a lot more hands where 2 aces are missing. But why not ask for aces if that is part of your system, even if missing 2 aces doesn't happen often?

As far as making 6NT missing 2 aces, half the time one opponent will have both aces. I estimate the chances of 6NT going down as 100%. When aces are split, sometimes you can't ever make 12 tricks without losing 2 aces first. Or the defense may just get lucky and cash both aces, or make informed guesses.

[pescetom]

 johnu, on 2026-June-13, 16:02, said:

Also, in real life, many players will upgrade very good 14 counts to 1NT. So there are a lot more hands where 2 aces are missing. But why not ask for aces if that is part of your system, even if missing 2 aces doesn't happen often?

If you and your partner are both unemotional and highly concentrated all the way through a competition then it's a possible tiny marginal gain, agreed. In my experience many people in real life get excited by Ace asking (and/or confused between Blackwood and Gerber replies) and screw up more frequently than they would like to admit, certainly more often than being off two Aces. There is also the small risk of a successful lead directing double.

[mycroft]

Ace asking bids are a necessary evil. They are also - especially over 1NT or 2NT openers - massively overused by all but the highest class of players (IMO, mostly either 'don't have an alternative to "we might have slam"' or 'don't have patience to work out a better way of investigating slam'). I have an occasional partnership with a national-level player (who, TBFair, treats me as her "relax and have fun, maybe experiment" partner, but still) - we play Gerber. She's winning, 3-0 in the bid it stakes. She was also *clearly* wrong to do so - as in, "the information you needed wasn't 'Aces', or at least not yet" - at least two of those times.

Yes, if I was going to go all in on "31 in two balanced hands is 50+% to make 6NT DD", there is a definite argument for some kind of ace-control mechanism in a flat invite/"31-33 blast". I'm absolutely willing to argue that.

Now, some of this is my matchpoint bias (I just don't get to play IMPs very often any more), but I find that the loss from no G - even the loss of "31 minimums should at least invite" because we could be off two cashers and I have no way of finding that out (*) - is absolutely made up for by our alternative uses of 4♣ in NT auctions to find the best game, or in fact as a different kind of slam investigation (**).

Slams are fun. Slams are cool. Pulling out toys is fun. I will absolutely admit that; and if that is your goal, I'm happy for you. I kind of have to be - I'm the person who says "bring your card, I'll see what I can't play on that. You don't want to play mine" because, well, toys, at least some of which are "fun but clearly inferior" :-). But with most partners, especially ones interested in learning *from me*, getting them to slow down and have work on other ways to look for slam until they can decide whether to pull out the OG is going to lead to both better results and more improvement.

Having said that, I listen to fred (and basically everybody else who's written about it) when he says "after 1st/2nd cuebidding, some sort of Keycard in case too many of them are 2nd". But, you know, "after".

(*) Okay, in my main "Gerber is baby food" partnership, we can keycard in any suit, below game. Assuming we are judicious about the suit, I can absolutely see us "asking for aces" and stopping in 4NT. I've never tried that, though.
(**) Or, in that same partnership, "play 4♥ from the correct side" with 12 opposite 12. I'm sure just that has won more than the slams we've missed from Baby Food.

[mycroft]

pescetom, on 2026-June-14, 06:37, said:

In my experience many people in real life get excited by Ace asking (and/or confused between Blackwood and Gerber replies) and screw up more frequently than they would like to admit, certainly more often than being off two Aces.

Heh, that's even one of my "training situations" for directors.

• I have had ...4NT-p-4x, when alternatives explained to them, bid 5x "to be comparable" - only to find that 4x was the Gerber response, the RKC response was 5y. Players actually find that very difficult to understand because of "make it sufficient" disease. Question for directors - do you look at the offender's hand to find out whether they were "responding to 4NT at the 4 level" or "responding to Gerber"? Do you ask them what they were trying to show? Do you do nothing but explain the "has the same meaning" bit of L23 and expect the offender to get it right (and for partner, who knows nothing, to know to pass if he got it wrong)? Do you force them to play in 5x if it isn't comparable, if you somehow know that it isn't (during the auction or as an adjusted score after)? Note: I don't know the right answer. I do know that *having an answer* is righter than "huh?".
  
• I have had ...4♣-p-5x, (as a player, in fact).(*) Are you prepared for the L25 discussion (and potentially the UI discussion after) when they clue in?
  
• I have had ...4NT-p-5x, "Oh I meant 5y". Again, are you prepared to work out whether "gave the Gerber Response, to Blackwood" or "always wanted to bid 5y, pulled the wrong card"?

(*) In that partnership, in approximately 15k hands, we had 3 Gerber auctions.

• Once, I gave the Blackwood response.
  
• Once, partner bid the Gerber response - at the 5 level.
  
• And once, we were playing EHAA, so "Gerber after Stayman only" and after 1NT-p-4♣ partner didn't pass.

We stopped playing Gerber.