[zheddh]

To win in the long run, what probability of making a slam do you need to bid a small slam? a grand slam?

There are 8 different cases.

slam vul, imps
slam non - vul, imps
slam vul, mps
slam non-vul, mps

grand vul, imps
grand non-vul, imps
grand vul, mps
grand non-vul, mps

In MPs, it would depend a lot on the field i guess. Your opinion?

[Fluffy]

grand slam also depends on the hand in question as well as the field/other table.

If you have 35 balanced vs balanced, you can be pretty sure the other table is playing at least 6. But the problem with grands is that you might go 1 off when the other table is palying game. (I have seen part score vs grand slam at the 2 tables for +8 IMPs)

The books I've read said gran slam needs 70%, this is a gross aproximation IMO, but might work for you.

[Phil]

I could give you the percentages, however, it would be a very useful exercise if one of our intermediate players took it upon themselves to work out the math.

[mikeh]

The exercise suggested by phil is relatively easy for small slams, but not for grands, because, as Fluffy observed, sometimes the comparison is against game, not small.

When playing a weaker team, it is probably best to avoid grands that are less than about 75-80%, and, if a much weaker team, maybe 90%.

[P_Marlowe]

Fluffy, on Jul 23 2009, 03:05 AM, said:

<snip>
The books I've read said gran slam needs 70%, this is a gross aproximation IMO, but might work for you.

Just to put it into perspective and to give an example, what the pure
number means: 70% means roughly, that you only need trumps
breaking 3-2, which has a probability of 68%, if I remember it
correctly.
Of course besides the pure a priori percentages, tactical considerations
are also important, and quite often are more relevant.
And a approach never to bid grand slams (unless they are 100%, which
basically means 7NT) has some apeal.
From this followes, that it makes sense to give up on Grand Slam to have
better ways to investigate Small Slams.
My experience is based on Bridge (far) below the level of Fluffy / MikeH.

With kind regards
Marlowe

[nigel_k]

It depends on what assumptions you make about what will happen at the other table(s).

For small slams at IMPS, assuming they bid either game or small slam in the same denomination at the other table, you need 50%. This is because you stand to gain 11 imps non vul or 13 imps vul and lose the same number. For example if you are -100 and they are +650 you lose 13, while if you are +1430 and they are +680 you gain 13.

For grand slams at IMPs, assuming they play in either small slam or grand slam, you need about 56%. Non vul you gain 11 or lose 14 so need 14/25. Vul you gain 17 or lose 13 so need 17/30.

Obviously the assumption about the other table is more likely to hold in the small slam case than the grand slam case. You don't want to go one down in 7 when opponents play in game. If the assumptions are unreliable you need a better chance of success. Some of the things mentioned below about matchpoints also apply at IMPs, especially if opponents are not strong.

At pairs there are more factors to consider.

Any time there is a possibility of others playing in the wrong strain you want to be conservative. If the opponent do any bidding or you used any sort of convention such as fourth suit forcing then usually some tables will find a way to screw up and you can beat them just by making game.

If there is a chance of gaining a trick in the play you also want to be conservative.

With a 5-3 major fit and significant extra values often 3NT and 4 of a major make the same number of tricks. If the slam is 50% and half the field bids it you can play 3NT and get half the matchpoints when the slam makes (because you beat everyone in 4 of a major) and all the matchpoints when the slam fails.

If it's a minor suit slam, plenty of tables will play 3NT and you need to consider how that will go. If 3NT is likely to make an overtrick there's not much point in stopping in 5 of a minor. If 3NT is unlikely to make an overtrick there's no need to bid slam.

If there are no such special considerations, a slam needing a finesse at worst is ok at matchpoints.

You certainly should be much more conservative with grand slams at MPs compared to IMPs though it's hard to put a figure on it. To some extent the percentage required is academic because the likelihood of making will seldom fall between 68% and 90%. You want to bid the 90% ones and the occasional 84% ones.

The main problem hands are where you need a 3-2 break, i.e. 68%. As above, if your fit is 5-3, 6NT may make the same number and you can get most of the matchpoints without risk by doing that. If the fit is 4-4 in a minor you need to consider whether 6NT may make 6 when the minor makes 7. You want to bid 7 if it does and stop in 6 if it doesn't. If the fit is 4-4 in a major and you expect the field to find that fit, then you probably should bid a 68% grand. There is always a chance partner will turn up with the jack or ten and you are much better than 68%.

Hope this helps.

Nigel.

[mike777]

nigel_k, on Jul 23 2009, 07:17 PM, said:

It depends on what assumptions you make about what will happen at the other table(s).

For small slams at IMPS, assuming they bid either game or small slam in the same denomination at the other table, you need 50%. This is because you stand to gain 11 imps non vul or 13 imps vul and lose the same number. For example if you are -100 and they are +650 you lose 13, while if you are +1430 and they are +680 you gain 13.

For grand slams at IMPs, assuming they play in either small slam or grand slam, you need about 56%. Non vul you gain 11 or lose 14 so need 14/25. Vul you gain 17 or lose 13 so need 17/30.

Obviously the assumption about the other table is more likely to hold in the small slam case than the grand slam case. You don't want to go one down in 7 when opponents play in game. If the assumptions are unreliable you need a better chance of success. Some of the things mentioned below about matchpoints also apply at IMPs, especially if opponents are not strong.

At pairs there are more factors to consider.

Any time there is a possibility of others playing in the wrong strain you want to be conservative. If the opponent do any bidding or you used any sort of convention such as fourth suit forcing then usually some tables will find a way to screw up and you can beat them just by making game.

If there is a chance of gaining a trick in the play you also want to be conservative.

With a 5-3 major fit and significant extra values often 3NT and 4 of a major make the same number of tricks. If the slam is 50% and half the field bids it you can play 3NT and get half the matchpoints when the slam makes (because you beat everyone in 4 of a major) and all the matchpoints when the slam fails.

If it's a minor suit slam, plenty of tables will play 3NT and you need to consider how that will go. If 3NT is likely to make an overtrick there's not much point in stopping in 5 of a minor. If 3NT is unlikely to make an overtrick there's no need to bid slam.

If there are no such special considerations, a slam needing a finesse at worst is ok at matchpoints.

You certainly should be much more conservative with grand slams at MPs compared to IMPs though it's hard to put a figure on it. To some extent the percentage required is academic because the likelihood of making will seldom fall between 68% and 90%. You want to bid the 90% ones and the occasional 84% ones.

The main problem hands are where you need a 3-2 break, i.e. 68%. As above, if your fit is 5-3, 6NT may make the same number and you can get most of the matchpoints without risk by doing that. If the fit is 4-4 in a minor you need to consider whether 6NT may make 6 when the minor makes 7. You want to bid 7 if it does and stop in 6 if it doesn't. If the fit is 4-4 in a major and you expect the field to find that fit, then you probably should bid a 68% grand. There is always a chance partner will turn up with the jack or ten and you are much better than 68%.

Hope this helps.

Nigel.

This sounds great, but wow very complicated.