Help
Is there any statistic for the percentage of game invitations accepted?
Page 1 of 1
Percent of Invites Accepted
#2
- Group: Full Members
- Posts: 80
- Joined: 2011-November-07
Posted 2023-March-04, 14:15
Percentage? For me, all invites seem to be forcing. ;-) At least at IMPs.
Seriously, a good rule (certainly at IMPs) is to invite game slightly heavy, but accept often. The idea is you don't want to be in 3 of a major or 2NT, and then go down 1 when partner has a min. But if you are already in 3 of a major or 2NT, then going down 2 is not much worse than down 1. That changes a little at matchpoints, when down 2 could be the kiss of death. Of course vulnerability factors in too.
All of that is also subject to your skill at declarer play too, and the strength of your opponents at defense. State of the match should be a factor, at IMPs. At matchpoints, your current estimate of your game should be a factor. You will rarely win with a 53% game, so you might decide to push more, or less.
The point is, there are many factors worth considering, not just fraction of game tries accepted.
Seriously, a good rule (certainly at IMPs) is to invite game slightly heavy, but accept often. The idea is you don't want to be in 3 of a major or 2NT, and then go down 1 when partner has a min. But if you are already in 3 of a major or 2NT, then going down 2 is not much worse than down 1. That changes a little at matchpoints, when down 2 could be the kiss of death. Of course vulnerability factors in too.
All of that is also subject to your skill at declarer play too, and the strength of your opponents at defense. State of the match should be a factor, at IMPs. At matchpoints, your current estimate of your game should be a factor. You will rarely win with a 53% game, so you might decide to push more, or less.
The point is, there are many factors worth considering, not just fraction of game tries accepted.
#3
- Group: Advanced Members
- Posts: 12,867
- Joined: 2005-June-15
- Gender:Male
- Location:Canada
- Interests:Bridge, golf, wine (red), cooking, reading eclectically but insatiably, travelling, making bad posts.
Posted 2023-March-04, 14:59
I agree (mostly) with wuud
However, at imps whether one accepts an invitation should be extremely dependent upon vulnerability.
The key is imps won if you accept and are correct, assuming they reject at the other table, compared to imps lost if you accept and go down one, with their making a partscore at the other table
The rough analysis done by experts assumes (not because it’s true but because it simplifies the analysis without rendering it worthless) that neither game nor partscore are doubled and that the play and defence are much the same at both tables, and that the partscore makes and the game is down one.
When non vulnerable, if you bid game and make, you will pick up, say, 6 imps for +420 v 170 and lose 5 imps for -50 v 140.
This makes bidding game at imps, nv, pretty much a 50-50 proposition, all the more so when we consider two real life factors ignored in the simple approach. One is that most opps won’t double a partscore for a one trick set, but will happily double a game for a two trick set. So on a fraction of close decision hands, an opponent can figure out that (a) it’s close and (b) things are sitting very badly for you. Then the comparison is -300 v - 50 when even the partscore fails…and there’s no compensating upside.
The second reason is that it’s quite common to score an overtrick in a partscore due to the opps, at imps, trying to set the contract, often seeing the risk of an overtrick as worth assuming. So sometimes the comparison is between -50 and +170.
Finally, if you think, for example, that your possible 4S contract is simply on a finesse, and thus 50-50, you’re almost certainly wrong!
Say you have AKJxxx opposite xx, no spots. If you need 6 tricks from that suit, it’s not ‘on a finesse’. You need a 3-2 or 2-3 break with the queen onside…roughly 34.25%, not 50.
Say you have AKJ10x opposite xxx. Now it’s much closer to 50% but there will often be very low frequency, but non-zero, chances of a ruff in a side suit, etc.
So a good rule is to accept nv invites only when you think that the odds are at least 50-50.
Vulnerable, the same analysis leads to a much different result. I’m not going into the math here, but it is generally accepted that one should bid all vul games with at least a 40% chance of making, and some are a little more aggressive than that.
At mps, because the size of gain or loss on a hand is irrelevant to the scoring (if everyone else is +140, your -50 scores the same as -1100, and your +420 the same as +2210), the go-no go is 50%. Since most apparent 50% contracts are actually not that good, due to low frequency occurrences, in a very strong field, avoid 50% games
However, defence is the toughest part of the game and, in most mp fields, the opps will be highly variable in their defensive skills, I’d suggest being a little more aggressive…if…big if…whoever is declarer is above average in skill.
As for who stretches to invite or to accept, I’m 100% in agreement with wuud. Invite heavy (have full values, don’t stretch) and accept unless you have a good reason to reject. The so-called invite heavy, accept light approach. It’s mathematically demonstrably the best approach for the reasons articulated by wuud.
One final note. All of the above ignores ‘how one invites’
It is a reality, especially against good players, that descriptive game invitations can lead to failure when blasting would work. Indeed, some very good pairs either never or rarely make game tries…they just bid game. It’s not unknown to see a partscore fail at a table where an informative try was rejected and game make where it was blasted.
So especially at imps when vulnerable a lot of good players don’t use game tries. Say 1S 2S….do you bid a help suit, letting the opps learn of your worries about the suit and borderline values, or do you leap to game?
Personally, I blast a lot at imps red, and bid a little more descriptively nv or at mps.
I’ve given, as I’m wont to do, a long answer that probably doesn’t really address the question! But I’ve never seen statistics on the topic, every partnership’s approach will be slightly different, and I don’t even have a subjective feeling for how my partnerships fare in terms of frequency
However, at imps whether one accepts an invitation should be extremely dependent upon vulnerability.
The key is imps won if you accept and are correct, assuming they reject at the other table, compared to imps lost if you accept and go down one, with their making a partscore at the other table
The rough analysis done by experts assumes (not because it’s true but because it simplifies the analysis without rendering it worthless) that neither game nor partscore are doubled and that the play and defence are much the same at both tables, and that the partscore makes and the game is down one.
When non vulnerable, if you bid game and make, you will pick up, say, 6 imps for +420 v 170 and lose 5 imps for -50 v 140.
This makes bidding game at imps, nv, pretty much a 50-50 proposition, all the more so when we consider two real life factors ignored in the simple approach. One is that most opps won’t double a partscore for a one trick set, but will happily double a game for a two trick set. So on a fraction of close decision hands, an opponent can figure out that (a) it’s close and (b) things are sitting very badly for you. Then the comparison is -300 v - 50 when even the partscore fails…and there’s no compensating upside.
The second reason is that it’s quite common to score an overtrick in a partscore due to the opps, at imps, trying to set the contract, often seeing the risk of an overtrick as worth assuming. So sometimes the comparison is between -50 and +170.
Finally, if you think, for example, that your possible 4S contract is simply on a finesse, and thus 50-50, you’re almost certainly wrong!
Say you have AKJxxx opposite xx, no spots. If you need 6 tricks from that suit, it’s not ‘on a finesse’. You need a 3-2 or 2-3 break with the queen onside…roughly 34.25%, not 50.
Say you have AKJ10x opposite xxx. Now it’s much closer to 50% but there will often be very low frequency, but non-zero, chances of a ruff in a side suit, etc.
So a good rule is to accept nv invites only when you think that the odds are at least 50-50.
Vulnerable, the same analysis leads to a much different result. I’m not going into the math here, but it is generally accepted that one should bid all vul games with at least a 40% chance of making, and some are a little more aggressive than that.
At mps, because the size of gain or loss on a hand is irrelevant to the scoring (if everyone else is +140, your -50 scores the same as -1100, and your +420 the same as +2210), the go-no go is 50%. Since most apparent 50% contracts are actually not that good, due to low frequency occurrences, in a very strong field, avoid 50% games
However, defence is the toughest part of the game and, in most mp fields, the opps will be highly variable in their defensive skills, I’d suggest being a little more aggressive…if…big if…whoever is declarer is above average in skill.
As for who stretches to invite or to accept, I’m 100% in agreement with wuud. Invite heavy (have full values, don’t stretch) and accept unless you have a good reason to reject. The so-called invite heavy, accept light approach. It’s mathematically demonstrably the best approach for the reasons articulated by wuud.
One final note. All of the above ignores ‘how one invites’
It is a reality, especially against good players, that descriptive game invitations can lead to failure when blasting would work. Indeed, some very good pairs either never or rarely make game tries…they just bid game. It’s not unknown to see a partscore fail at a table where an informative try was rejected and game make where it was blasted.
So especially at imps when vulnerable a lot of good players don’t use game tries. Say 1S 2S….do you bid a help suit, letting the opps learn of your worries about the suit and borderline values, or do you leap to game?
Personally, I blast a lot at imps red, and bid a little more descriptively nv or at mps.
I’ve given, as I’m wont to do, a long answer that probably doesn’t really address the question! But I’ve never seen statistics on the topic, every partnership’s approach will be slightly different, and I don’t even have a subjective feeling for how my partnerships fare in terms of frequency
'one of the great markers of the advance of human kindness is the howls you will hear from the Men of God' Johann Hari
#4
- Group: Full Members
- Posts: 281
- Joined: 2017-May-30
- Gender:Male
- Location:Germany
-
Interests:Bidding System Design
Walking my dogs
2 player Hanabi
Posted 2023-March-05, 02:53
I would think that some stats, such as
- percentage of invites accepted
- percentage of games made
- percentage of partscores making enough tricks for game
- percentage of enemy partscores going down
and the like might help to analyze your general approach.
- percentage of invites accepted
- percentage of games made
- percentage of partscores making enough tricks for game
- percentage of enemy partscores going down
and the like might help to analyze your general approach.
#5
- Group: Advanced Members
- Posts: 7,367
- Joined: 2014-February-18
- Gender:Male
- Location:Italy
Posted 2023-March-05, 16:23
mikeh, on 2023-March-04, 14:59, said:
I agree (mostly) with wuud
However, at imps whether one accepts an invitation should be extremely dependent upon vulnerability.
The key is imps won if you accept and are correct, assuming they reject at the other table, compared to imps lost if you accept and go down one, with their making a partscore at the other table
The rough analysis done by experts assumes (not because it’s true but because it simplifies the analysis without rendering it worthless) that neither game nor partscore are doubled and that the play and defence are much the same at both tables, and that the partscore makes and the game is down one.
When non vulnerable, if you bid game and make, you will pick up, say, 6 imps for +420 v 170 and lose 5 imps for -50 v 140.
This makes bidding game at imps, nv, pretty much a 50-50 proposition, all the more so when we consider two real life factors ignored in the simple approach. One is that most opps won’t double a partscore for a one trick set, but will happily double a game for a two trick set. So on a fraction of close decision hands, an opponent can figure out that (a) it’s close and (b) things are sitting very badly for you. Then the comparison is -300 v - 50 when even the partscore fails…and there’s no compensating upside.
The second reason is that it’s quite common to score an overtrick in a partscore due to the opps, at imps, trying to set the contract, often seeing the risk of an overtrick as worth assuming. So sometimes the comparison is between -50 and +170.
Finally, if you think, for example, that your possible 4S contract is simply on a finesse, and thus 50-50, you’re almost certainly wrong!
Say you have AKJxxx opposite xx, no spots. If you need 6 tricks from that suit, it’s not ‘on a finesse’. You need a 3-2 or 2-3 break with the queen onside…roughly 34.25%, not 50.
Say you have AKJ10x opposite xxx. Now it’s much closer to 50% but there will often be very low frequency, but non-zero, chances of a ruff in a side suit, etc.
So a good rule is to accept nv invites only when you think that the odds are at least 50-50.
Vulnerable, the same analysis leads to a much different result. I’m not going into the math here, but it is generally accepted that one should bid all vul games with at least a 40% chance of making, and some are a little more aggressive than that.
At mps, because the size of gain or loss on a hand is irrelevant to the scoring (if everyone else is +140, your -50 scores the same as -1100, and your +420 the same as +2210), the go-no go is 50%. Since most apparent 50% contracts are actually not that good, due to low frequency occurrences, in a very strong field, avoid 50% games
However, defence is the toughest part of the game and, in most mp fields, the opps will be highly variable in their defensive skills, I’d suggest being a little more aggressive…if…big if…whoever is declarer is above average in skill.
As for who stretches to invite or to accept, I’m 100% in agreement with wuud. Invite heavy (have full values, don’t stretch) and accept unless you have a good reason to reject. The so-called invite heavy, accept light approach. It’s mathematically demonstrably the best approach for the reasons articulated by wuud.
One final note. All of the above ignores ‘how one invites’
It is a reality, especially against good players, that descriptive game invitations can lead to failure when blasting would work. Indeed, some very good pairs either never or rarely make game tries…they just bid game. It’s not unknown to see a partscore fail at a table where an informative try was rejected and game make where it was blasted.
So especially at imps when vulnerable a lot of good players don’t use game tries. Say 1S 2S….do you bid a help suit, letting the opps learn of your worries about the suit and borderline values, or do you leap to game?
Personally, I blast a lot at imps red, and bid a little more descriptively nv or at mps.
I’ve given, as I’m wont to do, a long answer that probably doesn’t really address the question! But I’ve never seen statistics on the topic, every partnership’s approach will be slightly different, and I don’t even have a subjective feeling for how my partnerships fare in terms of frequency
However, at imps whether one accepts an invitation should be extremely dependent upon vulnerability.
The key is imps won if you accept and are correct, assuming they reject at the other table, compared to imps lost if you accept and go down one, with their making a partscore at the other table
The rough analysis done by experts assumes (not because it’s true but because it simplifies the analysis without rendering it worthless) that neither game nor partscore are doubled and that the play and defence are much the same at both tables, and that the partscore makes and the game is down one.
When non vulnerable, if you bid game and make, you will pick up, say, 6 imps for +420 v 170 and lose 5 imps for -50 v 140.
This makes bidding game at imps, nv, pretty much a 50-50 proposition, all the more so when we consider two real life factors ignored in the simple approach. One is that most opps won’t double a partscore for a one trick set, but will happily double a game for a two trick set. So on a fraction of close decision hands, an opponent can figure out that (a) it’s close and (b) things are sitting very badly for you. Then the comparison is -300 v - 50 when even the partscore fails…and there’s no compensating upside.
The second reason is that it’s quite common to score an overtrick in a partscore due to the opps, at imps, trying to set the contract, often seeing the risk of an overtrick as worth assuming. So sometimes the comparison is between -50 and +170.
Finally, if you think, for example, that your possible 4S contract is simply on a finesse, and thus 50-50, you’re almost certainly wrong!
Say you have AKJxxx opposite xx, no spots. If you need 6 tricks from that suit, it’s not ‘on a finesse’. You need a 3-2 or 2-3 break with the queen onside…roughly 34.25%, not 50.
Say you have AKJ10x opposite xxx. Now it’s much closer to 50% but there will often be very low frequency, but non-zero, chances of a ruff in a side suit, etc.
So a good rule is to accept nv invites only when you think that the odds are at least 50-50.
Vulnerable, the same analysis leads to a much different result. I’m not going into the math here, but it is generally accepted that one should bid all vul games with at least a 40% chance of making, and some are a little more aggressive than that.
At mps, because the size of gain or loss on a hand is irrelevant to the scoring (if everyone else is +140, your -50 scores the same as -1100, and your +420 the same as +2210), the go-no go is 50%. Since most apparent 50% contracts are actually not that good, due to low frequency occurrences, in a very strong field, avoid 50% games
However, defence is the toughest part of the game and, in most mp fields, the opps will be highly variable in their defensive skills, I’d suggest being a little more aggressive…if…big if…whoever is declarer is above average in skill.
As for who stretches to invite or to accept, I’m 100% in agreement with wuud. Invite heavy (have full values, don’t stretch) and accept unless you have a good reason to reject. The so-called invite heavy, accept light approach. It’s mathematically demonstrably the best approach for the reasons articulated by wuud.
One final note. All of the above ignores ‘how one invites’
It is a reality, especially against good players, that descriptive game invitations can lead to failure when blasting would work. Indeed, some very good pairs either never or rarely make game tries…they just bid game. It’s not unknown to see a partscore fail at a table where an informative try was rejected and game make where it was blasted.
So especially at imps when vulnerable a lot of good players don’t use game tries. Say 1S 2S….do you bid a help suit, letting the opps learn of your worries about the suit and borderline values, or do you leap to game?
Personally, I blast a lot at imps red, and bid a little more descriptively nv or at mps.
I’ve given, as I’m wont to do, a long answer that probably doesn’t really address the question! But I’ve never seen statistics on the topic, every partnership’s approach will be slightly different, and I don’t even have a subjective feeling for how my partnerships fare in terms of frequency
Stuck on a phone and unable to upvote, but kudos
Page 1 of 1