[jillybean]

---

I'm putting the 4♣ card down, how about you?

[smerriman]

Seems tempting though if partner has enough to act - say, both aces - are we able to avoid two losers? Might be asking a bit much.

[akwoo]

I actually think 2♠ (or whatever game try showing something in spades and asking for more in spades is) is more likely to get you the right information - you really want to know if partner has a spade honor.

If partner bids 4♥ over 2♠, bidding 5♣ should tell them what's going on. If partner denies the game try, I'll make one more try with 4♣.

The point is that it's hard to get info about pointed queens after 4♣, but not as hard to show the club shortness after 2♠.

[mike777]

4H
Giving up on slam

[harikannan]

2S game try and 5C only if partner bids 4H. If he bids anything else, i bid 4H.

[Cyberyeti]

Need to be clear what sort of game try you're making, ♠ x, Qx, QJx, Ax (but not so much Axx(x)) are great holdings.

How often does partner have 4 hearts here in the OP's methods ?

[P_Marlowe]

smerriman, on 2025-May-04, 20:59, said:

Seems tempting though if partner has enough to act - say, both aces - are we able to avoid two losers? Might be asking a bit much.

I agree, even if he the two major suit Aces, which have to be dream cards, we still need to take care of the 3rd diamond
and the 3rd spade.

[Cyberyeti]

P_Marlowe, on 2025-May-05, 00:50, said:

I agree, even if he the two major suit Aces, which have to be dream cards, we still need to take care of the 3rd diamond
and the 3rd spade.

Well unless you're bidding a grand, you only need to take care of one of them, Ax, A10x, xxxx, xxxx is fine as most of the time is Axxx, Axx, xx, xxxx

The worst hand I could think of that would do the job (although not sure if you'd bid 3 first time) is Ax, 109xx, Qx, xxxxx, you don't need Q♦ unless they play ♥Ax and they may not be able to do so.
Also x, Axx, Qxxxx, xxxx will do the job most of the time

[jillybean]

---

Partner will accept a game try, showing clubs/control on the way.

[mikeh]

Late to the thread, which is pretty interesting. There are multiple issues, resolution of which is somewhat dependent upon methods.

Is the hand worth a slam try? Imo, yes. While I’d expect slam to be poor opposite the great majority of 2H responses, there are some where slam is very good, and even grand could make once in a while, although we’re not getting there while knowing what we’re doing. I expect some to claim that they’d get there using their pet methods. I don’t believe them (absent maybe a big club with relays).

What try should we make?

Methods matter but almost everybody plays some form of help suit game try. While our bid will initially sound like a try for game, competent partners know that it could be the start of a try for slam.

One issue is: what kind of holding does one have for a help suit game try (hsgt)? I’ve seen players do it with something like xxx. My own preference is that I always have an honour. Qxx(x) is about the worst I will have.

This matters because if one can have xxx than partner, looking at Qxx, doesn’t really have help. But if you might have Axx or K10x then Qxx is enough, provided that responder isn’t minimum.

Anyway, 2S is clear.

I had been planning on discussing how to move forward if partner rejected. In other words, are we worth another try after 1H 2H 2S 3H? After all xx Axx QJxxx xxx make slam.

But while I was dithering, jb posted that partner bid 3C.

Jb says that this was an acceptance of the hsgt. It wouldn’t be for me.

Say one has two suits where help in either one makes game worth bidding. One makes the cheapest try. That allows responder, with a good 2H bid but no help in your first trial bid, to show help elsewhere. So he might not help your first suit but his rebid, showing values in another suit, might be just what you need. Using the 3C bid here to promise help in spades and club values/cue is inefficient. Responder, in this method, has to bid 3H on all hands that reject the 2S try. This means that opener, needing help in one of two suits, can only ask about one.

For me, 3C denies spade help but says ‘had you tried via 3C, I would accept’.

Now, what if responder likes the 2S bid AND has a club control? There are two options.

One: bid 3C, acting as if one doesn’t help spades, then bid again if opener tries to sign off or, two, my preference if holding the club ace, jump to 4C over 2S. This says…I help spades and here’s the club ace.

As it is, once partner rejects spades, and shows something in clubs, I’m giving up on slam. Does that mean slam will never make? No….but it’s surely going to be thin at best. Try coming up with hands consistent with the auction and you’ll see that slams better than 50% are difficult to construct.

[DavidKok]

Initially I wasn't going to comment, but I'll share my thoughts briefly anyway.

First of all, game tries. They're worse than most people think. There are several options, including but not limited to:

• Help suit
  
• Long suit
  
• Short suit
  
• Generic/low information
  
• Good old college

Of these, I believe 'good old college' and 'low information' to be most effective on average, with 'short suit' in distant third, 'help suit' in close fourth and 'long suit' to be trailing this list. Let's take a slightly more abstract view of them for a second. We will get back to slam bidding later, so for now pretend with me we're only thinking about game.

Compared to blasting game, when does making a game try gain?

Firstly, if partner accepts the invite to game, we've gotten to the same contract but shared more information en route (and possibly allowed for some lead-directing double or lead-undirecting pass, or the likes). On average we will be slightly behind here.¹
Secondly, if partner rejects game, obviously game also needs to fail for us to gain. They don't hand out those game bonuses for free.
Thirdly, for the invite to be better than the blast, partner needs to be in a better position to diagnose this than us. There's no point asking partner to cut the tie on a 50/50 if partner can't do better than 50/50 either - numbers are figurative, of course. Put differently, we need partner to have an information advantage on the auction. Normally this is easily achieved by the invite itself - it describes our hand to a narrow shape and strength. However:
Fourthly, and this is where we get into the weeds of specific game tries, we need the information shared to not help the opponents a lot while at the same time helping partner as much as possible. If the game try announces the killing lead to the opponents we might correctly diagnose that after the invite we should be in 3M, but without the game try we'd have made either 2M+2 or 4M=. Of course usually this doesn't happen, but the margins for invites to be profitable are so razor-thin in the first place that this can be a deciding factor.

Putting this all together, I think that 'good old college' and 1M-2M; 2NT 'low information' (bid 3M with a min, 3NT or 4M with a max, and you are allowed to but discouraged from bidding 3-of-a-suit to show more information about your hand) are more effective than e.g. the traditional approaches such as Kokish. It is rare for all four conditions to hold simultaneously, and the possible downsides of the high information methods are significant downsides to methods that don't stand to gain much in the first place. I have game tries on my system card but in practice rarely use them. It takes a lot of conditions to be just right for a high information invite to be the best course of action, and in practice I tend to avoid making these bids.

I mentioned Kokish game tries earlier. To avoid confusion, with that I mean that direct game tries after 1M-2M are short suit (with 1♥-2♥; 2NT showing short spades) while 1M-2M; 2M+1 is an asking bid 'which help suit would you accept'. Some people invert the two.² Either way you get to play both help and short suit trials - the two main scientific options. Note that using the slow route, e.g. 1♠-2♠; 2NT*-3♦ ("accept in diamonds, reject in clubs, no comment on hearts"), still allows you to make a HSGT in hearts if that was your suit. This way the relay bid is a full replacement for one set of game tries, at the cost of leaking massive amounts of information. To me this is a baseline of game tries, the minimum we can do without taking camouflage and information leakage into account. Many more direct suites, e.g. directly playing help suit game tries, allow for sufficiently comparable functionality that I'm happy to take this as the basics and discuss from here - if your own game tries are some custom combination that is only similar to short suit game tries or help suit game tries in some ways I ask you to look to the similarities rather than stress the differences - I find that in general it's much of a muchness anyway. The questions then become:

• Asssuming this basic Kokish suite, how much do we gain/lose on all of the routes?
  
• Which game try options are we missing or would we like to add?
  
• How does our slam bidding work?

My answer to number 1 is that I'm largely indifferent to any of them - the routes don't gain much when they come up (and sometimes cost when they do), and they rarely come up in the first place. Number 2 is easy: I want a low information game try, just asking partner about range without describing further. This is also one of the design tenets of the Maas 2NT response to the 1M openings - 'bid game quickly, bid slam slowly'. We can repurpose the 2M+1 help suit try to be more generic, and by default this overload is already partially present (e.g. 1♠-2♠; 2NT*-3♠ means 'I reject all help suit trials', which is a generic minimum, and 1♠-2♠; 2NT*; 4♠ means 'I accept all help suit trials', i.e. a nice maximum. There is slight overlap here with the low information meanings, provided you use these bids often enough). On to number 3, slam bidding (I told you we'd get back to it).

In this suite, how does your slam bidding go? The default is that all game tries are assumed to be 'either game tries or preludes to slam investigation', and responder isn't supposed to take away too much of our bidding space with a maximum. Oops, so much for the low information aspect and 'bidding game quickly'. Opener gains access to the jump bids above 3M for slam tries, but we don't want to use all our bidding space on a possible slam auction, so this double purpose is generally desirable. But the probabilities on it are not in our favour: we'll have a genuine game try far more often than a slam try, so by demanding a high-information response as a game try accept we're losing a bit with high frequency, to gain something rarely. I don't know of a simple solution to this (maybe a low information rebid by responder saying 'I accept your game try and am leaving space in case you had a slam try, but otherwise do not describe my hand'?). Having played Kokish, short suit, help suit and even long suit for a few years each, my experience is that the information leakage on the game deals far outweighs our benefits on slam auctions. The conclusion, while unpleasant, is clear: I want low information methods for game and high information methods for slam, and this means most of the traditional options are insufficient. The slam bidding through game tries puts too much strain on the game bidding through game tries, and we're losing too much too often to justify it.
For that reason I played a radical set with one partner: on 1M-2M, 2NT is a low information game try, while changes of suit are short suit slam tries. Now we can bid cheaply and scientifically on slam deals, and the game tries go through good old college or low information (and, for completeness: jump bids show a slammish 5-5). The argument for this shortage choice is that on 1M-2M we 'never' have a slam try without shortage or a long second suit - such a hand would have opened 2♣.
However, I think it's conspicuously convenient to introduce a slam-searching method in this discussion after having seen the hand. I think the discussion on what to do with more standard methods is possibly more interesting, as it applies to a wider audience. In that case I would bid 3♣ (short suit game try, possibly a prelude into slam investigation), where 2♠ would be Kokish for either a low information game try or a help suit game try asking bid/prelude to a help suit game try. Just, by the brief introduction above, keep in mind that I think many experts are misapplying game tries on this and related auctions, and the classical dual-purpose 'game try or prelude to slam' bids are costing more than many realise.

¹There is another (in)famous way for the slow route to game to lose, by announcing to the world that the points are split approximately 25-15. Sometimes this allows a defender to double a contract that they otherwise wouldn't have dared to, with some famous examples (e.g. from Fredin) on 1NT-2NT; 3NT-(X) asking for a spade lead in a weak hand, confident that partner has enough missing honours that the extra trick from the lead will be enough to set the contract.
²Personally I believe short suit game tries are slightly superior to help suit game tries, for the main reason that if partner rejects the invite (or kicks the ball back with a descriptive bid) the opponents now aren't on high alert regarding a side suit. Spelling it out: if a help suit try is rejected it means both declarer and dummy are weak there and cannot get rid of the losers in that suit, so it's an attractive lead. I've seen 3M= after a HSGT with 4M= at the other table without a few times myself (and the spectactular 3M-1 versus 4M= twice, if I recall correctly). It's not that common, but it is one additional factor reducing the effectiveness of these help suit game tries.

[mikeh]

David did a very good job discussing game and slam tries.

I do think that there is a difference, or should be, between imp and mp strategies, although my implementation of those differences isn’t always consistent.

At imps, especially if vulnerable, I tend to avoid game tries, other than the old-fashioned ‘bid game and try to make it’.

Many years ago, before the invention of 2S as range ask/clubs in response to a strong notrump, I was on a team with Fred Gitelman. He advised never inviting 3 N if responder didn’t have a major suit. Just decide what to do…pass or….the more usual choice (especially if vulnerable) bid 3N. He said that in his experience blasting was worth roughly half a trick.

Firstly, using stayman when responder merely wanted to invite in notrump (virtually nobody then or now plays 1N 2 N as invitational), leaked info. Say you’re on lead with Q10xx in both majors. After 1N 3N, good luck guessing….you may well lead into declarer’s chunky 4 card major, blowing trick and a tempo. Secondly, you’re guessing on defence as to whether opener has a min, a max, or medium…placing honours is far more difficult than it is when opener tells you whether he’d accept or reject an invitation.

Similar reasoning applies to blasting 4M.

For those reasons, in my partnerships my game try, if help suit, doesn’t just ask for help…responder needs to hold a max for the raise and help in the suit. This is because we don’t make tries very often and can afford to be particular. The hand can’t be worth a blast but has to hold some prospect for game opposite the ‘right’ hand.

As to which game tries…in my previous post I suggested 2S as help suit and I’d still argue that’s slightly better than a short try on this hand. I generally don’t like tries or bids where partner will assume I have a stiff when I have a void….the void hand is a full trick better than the stiff hand, most of the time.

One can, as David points out, play two way tries.

For me, I like 2S over 2H to be natural…as in a hsgt. After 1M 2M, we use 2N to ask partner to bid the cheapest suit in which he would reject a short suit try.

This does leak some info but not as much as when opener specifically identifies his short suit.

We did look at whether we should eliminate or change our game try structure so as to make most ‘game try’ bids slam tries, but the frequency of a slam try after a simple raise is pretty low. The frequency of game tries is lower for us than for many pairs, but still higher, imo, than slam tries, plus we seem to cope pretty well with slam bidding in general.

[DavidKok]

Thank you for the response. I think we're in close agreement. I still want to respond to some points, and I don't know how to do that without it seeming like I disagree in general. I don't, it's just some minor details.

mikeh, on 2025-May-05, 12:20, said:

For those reasons, in my partnerships my game try, if help suit, doesn’t just ask for help…responder needs to hold a max for the raise and help in the suit. This is because we don’t make tries very often and can afford to be particular. The hand can’t be worth a blast but has to hold some prospect for game opposite the ‘right’ hand.

This comes with its own risks on a rejected invite. You've written before about your 'invite heavy, accept frequently' approach, which seems somewhat inconsistent with this description. If the game try asks for a somewhat specific hand I expect this route to end in 3M often.

mikeh, on 2025-May-05, 12:20, said:

For me, I like 2S over 2H to be natural…as in a hsgt. After 1M 2M, we use 2N to ask partner to bid the cheapest suit in which he would reject a short suit try.

This does leak some info but not as much as when opener specifically identifies his short suit.

This is, I think, our biggest disagreement regarding game tries - while also being a pretty minor disagreement. I think the short suit try has an advantage when it comes to information leakage, because while it does narrow down the hand shape it doesn't pinpoint any particular lead. If I make a short suit game try and partner rejects that doesn't motivate any particular lead - there's no merit in trying to establish the suit declarer is short in, and declarer likely has extras in the others. And if the try is accepted this also doesn't suggest much - responder could have had a generic maximum, or have little wastage opposite the shortage. My guess, but this is not backed by keeping score, archive analysis or simulations or the like, is that the help suit game try lead-directing information outweighs the shape information, when it comes to helping the defenders.
This connects to my point above - with help suit game tries I am actively afraid of ending in 3M, having told the opponents not just that we're lacking strength for game but also where we are weak. I think this is one of the worst sequences to have on a constructive auction.

mikeh, on 2025-May-05, 12:20, said:

We did look at whether we should eliminate or change our game try structure so as to make most ‘game try’ bids slam tries, but the frequency of a slam try after a simple raise is pretty low. The frequency of game tries is lower for us than for many pairs, but still higher, imo, than slam tries, plus we seem to cope pretty well with slam bidding in general.

This I fully agree with. I don't wish to suggest that my short suit slam try approach is good, just that it was one thing we tried among many to resolve the conflict between low information game sequences and high information slam sequences. I agree that game tries are more common than slam tries, even if you rarely use them. I'd go one step further and then lean into the bids being game tries, not 'game tries or advance slam tries', and let responder make low information responses by blasting or signing off. Worst case opener can bid again after the jump response, or jump past 3M on the second round.

[jillybean]

Here's the full hand, now I will take some time to comprehend the replies above.
This is another hand I did not play, auction as presented to me.

---

[akwoo]

That is not a 2♥ raise. (Well it is in my Precision partnership, barely, but we open many 9 counts at this vulnerability and 1♥ is at most 15.)

[jillybean]

akwoo, on 2025-May-05, 18:57, said:

That is not a 2♥ raise. (Well it is in my Precision partnership, barely, but we open many 9 counts at this vulnerability and 1♥ is at most 15.)

Agree. Absent other methods or hand evaluation skills, this is the auction you faced.

[mike777]

Unfortunately 2H is simply not an acceptable bid in a bidding quiz problem.

Discussing 1NT or 2C or 4h or 3H limit raise or what is bad about 2H is reasonable.

[akwoo]

jillybean, on 2025-May-05, 19:03, said:

Agree. Absent other methods or hand evaluation skills, this is the auction you faced.

Well, I'm going straight to 4♥ after the 3♣ bid (whatever it means), even if I know partner could have this. If partner had bid 1N, we might be in a not so great (but not terrible) slam.

[jillybean]

mike777, on 2025-May-05, 20:27, said:

Unfortunately 2H is simply not an acceptable bid in a bidding quiz problem.

Discussing 1NT or 2C or 4h or 3H limit raise or what is bad about 2H is reasonable.

I understand. This is not a "bidding quiz" but is how the auction progressed at the table.
I could have chosen to post this hand hand from South's perspective and we wouldn't allow the 2H bid.
The hand was sent to me by North, wanting to know how they could have bid to slam so let's first look at it from that angle.
Over 3C showing clubs or club control and accepting the game try, should North make a 3♦ cue hoping to hear a spade control?
Hearing 3S you still won't know how many KC you are missing without launching into exclusion. 4H was correct given this bidding.

It's unclear if South has the methods to show this hand, 1M 3M = 4 card support, invitational.

[smerriman]

jillybean, on 2025-May-05, 21:56, said:

Over 3C showing clubs or club control and accepting the game try, should North make a 3♦ cue hoping to hear a spade control?

I was (probably pessimistically) concerned about not being able to make slam even if every point in South's hand was working.. knowing that they have at least some of their limited values in clubs makes things worse, so must surely be a 4♥ signoff now.

jillybean, on 2025-May-05, 21:56, said:

It's unclear if South has the methods to show this hand, 1M 3M = 4 card support, invitational.

2♣ GF, maybe? Support, singleton in spades, and nearly 7 tricks in their own hand..