Aggressive vs Conservative Invites

[mikestar]

In one of Bob's lessons, he advocates aggressive game invitations and conservative acceptances. For example, in response to 1NT, if responder is somewhere between pass and 2NT, Bob advocates bidding 2NT. Then if opener is somewhere between pass and 3NT, he should pass.

 

I think that this is flawed. You will be playing in 2NT/3M a lot of the time. My preference is for sound invitations and aggressive acceptances. In the above example, if responder is somewhere between pass and 2NT, he passes. If he does bid 2NT and opener is somewhere between pass and 3NT, he bids game. This style has you playing in game as often as the other way, but less often in the dangerous and unrewarding one short of game contracts.

 

Comments?

[mikeh]

Aggressive invites and conservative accepts are exactly the opposite of what I play.

 

His style will reach or miss the same number of games as the converse, so there is no edge there one way or the other.

 

But there will be a set if hands on which the invite will be declined and the cards lie badly and the one-below game contract fails while the 2 below game makes (or avoids a double or goes down one less trick).

 

Accordingly, it will pay, in the long run, to play those hands one level lower. This is what happens when the invite is heavy and the accept light.

[Limey_p]

His style will reach or miss the same number of games as the converse, so there is no edge there one way or the other.

 

I disagree. You will miss some games if your invitation is conservative.

 

As is common at the table, it is better to have any agreement rather than none at all. That said, Mike & Mike, IMO you are in the minority on this one.

 

AP

[keylime]

I'm not much of a fan of invites. If my pard makes a noise it's hard to not let go of the bone. :ph34r:

[Guest Jlall]

I'm a fan of agressive invites and agressive accepts.

[mikeh]

His style will reach or miss the same number of games as the converse, so there is no edge there one way or the other.

 

I disagree. You will miss some games if your invitation is conservative.

 

As is common at the table, it is better to have any agreement rather than none at all. That said, Mike & Mike, IMO you are in the minority on this one.

 

AP

I think maybe you misunderstood.

 

Both schools will frequently bid in identical fashion: there is an overlap between light and heavy inviters: a considerable percentage of hands on which they both invite, and a considerable body on which opener will either always reject (a bare minimum) or always accept (significant extras).

 

we have essentially 4 cases to consider:

 

1. inviter is heavy/ acceptor is light. Both heavy and light inviters invite, but by 'light' for acceptor, I mean that narrow set of hands which are above dead minimum but not by enough for the heavy acceptor to take to game. The light acceptor bids game, the heavy passes.

 

On this layout, heavy invite gets to game, light plays two levels lower.

 

2. inviter is heavy/acceptor is heavy: both schools get to game here.

 

3. inviter is light/acceptor is light: neither side gets to game, the light plays a level lower

 

4. inviter is light/acceptor is heavy. By heavy, I again mean that group of hands on which a heavy acceptor would go and a light would pass

 

Now the light inviter gets to game and the heavy inviter plays two levels lower.

 

We see that in 2 of the 4 situations (categories 2 and 3), both schools reach or avoid game. The difference within those two categories is that the light invite crew is playing type 3 hands one level higher than the other school. It is here that the problem arises: since there is no bonus for almost reaching game, and sometimes, when the cards lie poorly, the light invite crew will lose to the heavy invite.

 

We can reasonably assume that the relative frequency of categories 1 and 4 are the same. If we do so, then it is apparent that, across all 4 categories, the two approaches are equivalent except for the problem with category 3 hands.

 

Thus, while the difference is not huge, it is clear. And since there is NO advantage accruing from light invites/heavy accepts to offset this cost, no logical player would choose that approach once they think about it.

[joshs]

His style will reach or miss the same number of games as the converse, so there is no edge there one way or the other.

 

I disagree. You will miss some games if your invitation is conservative.

 

As is common at the table, it is better to have any agreement rather than none at all. That said, Mike & Mike, IMO you are in the minority on this one.

 

AP

I think maybe you misunderstood.

 

Both schools will frequently bid in identical fashion: there is an overlap between light and heavy inviters: a considerable percentage of hands on which they both invite, and a considerable body on which opener will either always reject (a bare minimum) or always accept (significant extras).

 

we have essentially 4 cases to consider:

 

1. inviter is heavy/ acceptor is light. Both heavy and light inviters invite, but by 'light' for acceptor, I mean that narrow set of hands which are above dead minimum but not by enough for the heavy acceptor to take to game. The light acceptor bids game, the heavy passes.

 

On this layout, heavy invite gets to game, light plays two levels lower.

 

2. inviter is heavy/acceptor is heavy: both schools get to game here.

 

3. inviter is light/acceptor is light: neither side gets to game, the light plays a level lower

 

4. inviter is light/acceptor is heavy. By heavy, I again mean that group of hands on which a heavy acceptor would go and a light would pass

 

Now the light inviter gets to game and the heavy inviter plays two levels lower.

 

We see that in 2 of the 4 situations (categories 2 and 3), both schools reach or avoid game. The difference within those two categories is that the light invite crew is playing type 3 hands one level higher than the other school. It is here that the problem arises: since there is no bonus for almost reaching game, and sometimes, when the cards lie poorly, the light invite crew will lose to the heavy invite.

 

We can reasonably assume that the relative frequency of categories 1 and 4 are the same. If we do so, then it is apparent that, across all 4 categories, the two approaches are equivalent except for the problem with category 3 hands.

 

Thus, while the difference is not huge, it is clear. And since there is NO advantage accruing from light invites/heavy accepts to offset this cost, no logical player would choose that approach once they think about it.

Its actually slightly more complicated than this, and does depend on the auction.

For simplicity, lets assume you played 1H-2H as 8-10. the exact range doesn't matter, I just chose this range since 3 points is divisible by 3. Also lets just suppose you want to be in game with 25 points and not really with 24. Again this is all simplified, but the main idea is the the is some threshold of combined playing strength (of which HCP is only a part) at which you want to be in game.

 

First, lets suppose your ONLY game try was 3H. Then your argument was 100% correct.

 

If you play 1H-2H-3H as Agressive Style then you would bid 3H on 15 and only accept with 10. Thus with 15 opposite 8 you are slightly too high. With 16-8 you bid game and are also too high. With 16-9 you bid game, and its correct.

 

if you played 1H-2H-3H as conservative style you would bid 3H on 16, partner would bid game on 9 and pass with 8. You would get too high but might miss game on 15 opposite 10.

 

If that was all there was too it, this style trades 1 good game for twice getting too high.

 

But now lets introduce game tries OTHER than 3H.

1H-2H-3C:

Playing the agressive style you bid this on 15 and partner bids game on 10 Or if loves his hand after the club bid, bids 3H on an 8 that doesn't like clubs that much, but can bid something else with the hands in between (8's with a club fit or 9's with a marginal club fit).

 

Its this ability to make a counter game try where agressive game tries shine.

 

Lets forget about the exact hand and suppose after the 3C bid responder's hand is then worth 8,9,10 (but may be different than before).

 

Now in the agressive style, opener bids 3C with 15 or 16.

responder bids game with 10, bids something in the middle with 9, and signs off with 8. The 16-8 no longer gets too high.

 

In the conservative stlye, opener bids 3C with 16. with 17 he just bids game since responder is always good enough. Since opener is so tightly defined for the game try, you get less benefit from the counter-game try.

 

So now, in the agressive style you get too high on 15-8. But in the conservative style you are too low on 15-10. Since you are more likely to make 3H on 15-8 than go down in 4H on 15-10, this trade is a bad one, even at mps, and is very bad trade at imps.

 

Anyway, thats why the usual style is agressive invites. When inviting in NT, there are not counter game tries available, so its a different matter.

[hrothgar]

We can reasonably assume that the relative frequency of categories 1 and 4 are the same. If we do so, then it is apparent that, across all 4 categories, the two approaches are equivalent except for the problem with category 3 hands.

 

...

 

And since there is NO advantage accruing from light invites/heavy accepts to offset this cost, no logical player would choose that approach once they think about it.

Folks are posting some very definitive statements without bothering to back anything up with actual analysis. Par for the course, but still rather annoying. I understand that actually doing some work and seeing what the numbers look like is much more difficult that pulling assertions out of you ass. None-the-less, it does, on occasion, add something to the conversation.

 

In this case, the assumption that category 1 and category 4 are equal in frequency is decidedly NOT true. In actually, if you look at hands whose combined strength is sufficient for game the (Light Invite/ Heavy Accept) hands are significantly more frequent than (Heavy Invite/Light Accept). Here are the results of quick simulation that I ran. I recognize that HCPs are a pretty poor method to approximate playing strength. None-the-less, this is a commonly accepted metric and the same principle holds true for other metrics.

 

(I apologize in advance for the formatting. I'm not sure how to create a table...)

 

.......6.....7......8......9.....10.....11

12 3.12 3.74 4.06 4.21 4.09 3.74

13 2.87 3.42 3.64 3.65 3.50 3.10

14 2.54 2.97 3.09 3.04 2.83 2.46

15 2.09 2.39 2.46 2.37 2.15 1.82

16 1.67 1.87 1.86 1.75 1.55 1.27

17 1.28 1.37 1.34 1.22 1.05 0.82

18 0.90 0.96 0.90 0.81 0.68 0.52

19 0.60 0.63 0.58 0.50 0.40 0.30

20 0.39 0.39 0.35 0.29 0.23 0.16

 

The columns show the number of HCP held by opener.

The rows show the number of HCP held by responder.

 

If you compare different combinations of hands that total 24 HCP, its pretty clear that combinations like (14 / 10) or (15, 9) are MANY times more frequent than (17 / 7).

 

This is a very natural and intuitive outgrowth of the expected distribution of HCP. Opener has a hand that is stronger than average. Responder (who made a single raise) has a hand that rates to be weaker than average.

 

Invite light is closer to the mean than invite conservative

Accept heavy is closer to the mean than invite aggressive

 

throw in vaguely normal distribution and away we go

[mikeh]

Hrothgar

 

Your problem is that you have made an assumption, that I believe is in error..

 

Your critique is based on the erroneous asumption that it is the stronger hand that makes the invite. Thus, the light invite crew makes more invites than does the strong invite crew, and, because of the expected distribution of hcp, will reach more games.

 

But I made my simplifying assumption that the two groups reached and missed the same number of games because in many auctions it is the weaker hand that invites.

 

Thus, a simple example may be in order:

 

1♠ 2♠ game try

 

Now the big hand is inviting.

 

Assume the go -no go point is 25: I know, it is artificial, but bear with me.

 

Assume the dividing line between the heavy and light invite group is that the light invites with 15 and the heavy with 16.

 

Now the light invite gets to game on 15/10 and misses on 16/9: on 16/9 the responder passes (again, this is a SIMPLIFYING assumption: the 'numbers' don't matter, the concept does)

 

Clearly 15/10 is more common than 16/9 so that the light invite gets to more 'good' games (assuming, as we do here that '25' means a good game and '24' not).

 

But flip this to 1♠ 3♠, where 3♠ is limit.

 

Responder, with the weaker hand, is making the move.

 

Assume that (again, I stress, I am not putting these numbers out as high card points, just as a metric of some kind): light inviters need 10 to make this limit raise and heavy need 11.

 

The light inviters need opener to have 15 to go, while the heavy need only 14.

 

Thus the light inviters get to game on 10/15 and miss on 11/14 while the heavy inviters get there on 11/14.

 

11/14 is more common than 10/15 (according to your table, and my intuition), so now the heavy inviters get to more games than their counterparts. A more graphic example is 1N (strong) and the decision to invite game: now responder is significantly weaker, and the heavy invite crew will reach far more games than the light invite: 16 9 is far more prevalent that 17 8, as an example in hcp.

 

One can easily come up with many invitational sequences in which either the stronger or the weaker hand makes the invite.

 

This recognition that the inviter may be either opener or responder, may be weaker than the norm or stronger, underlies my assumption.

 

it may well be that proper analysis would show that the weak hand invites family is more numerous than the strong hand invites family, but that analysis is far beyond my abilities and time.

 

Now, if I have made my own stupid assumption in this analysis (which is far from inconcievable), I am sure that I will hear from you. but if I have correctly identified a flaw in yours, maybe you could do two things: one would be to admit it and the other would be to think a little more carefully before writing more of your needlessly abusive comments. Contrary to what you appear to believe, there is no direct correlation between being an insulting boor and being perceived as insightful.

[cherdano]

To add s.th. in the direction Josh was arguing: Mike's discussion of course ignored the fact that we can make specific game tries (long suit, short suit, etc.).

If we invite more aggressively, then the number of auctions where we get to make a more informed decision about game increases, which should be a good thing.

 

I doubt this is really a disagreement with Mike. We all like to make more aggressive invite if we have a descriptive bid to do so (so that partner will accept on the right set of hands). Also, we like to make more aggressive invites when we know we have a 9-card fit, when the 3-level is rather safe.

 

The only case where I really hate making invites when it is not descriptive AND will help opponents to lead/defend... (2♣ Stayman followed by 2N)

 

Arend

[joshs]

This might be another one of those " bad bridge vocabulary" problems like calling the lead convention where you lead 3'rd from even and low from odd 3'rd and 5'th leads. (The normal lead from 6 cards without an honor sequence playing this lead convention is 3'rd best not 5'th best).

 

Over 1N, no one plays agressive invites and conservtive acceptences. Its a very bad treatment, for all the reasons layed out in this thread.

 

Lets suppose you want to be in 3N on 25 HCP. If you want to be in game on 24 HCP, shift the range for the invites down by 1 point in the argument. If you think you need 26 shift them up by 1 point, that fact does effect the argument.

 

Lets suppose you want to invite 3N and have the nice bid of 2N available to invite 3N over partner's 15-17 NT.

 

A. Playing "Agressive invites, and conservative acceptances" you bid 2N on 8 and opener accepts only with 17.

 

B. Playing "conservative Invites and agressive acceptances" (this is normal) you bid 2N on 9 and opener accepts with 16-17.

 

 

In A, you bid 3N on 9+ and get to game on 15-9 (bad), 16-9 (good), 17-8 (good), and 17-9 (great).

You play 2N on 15-8 (very bad) and 16-8 (acceptible).

 

In B you bid 3N on 10+ and get to game on 16-9 (good), 17-9(great)

You play 2N on 15-9 (acceptible)

You play 1N on 15-8 (good) and 16-8 (good) and 17-8 (bad)

 

For strategy A you get a better result on 17-8 (as richard points out, that is the least likely combination)

 

For strategy B you get a better result on 15-8 and 15-9 and 16-8 which are all more common. So its a better than a 3-1 gain for strategy B.

 

If my previous post was clear, the 1M-2M auctions are different because the invite can be wider ranging. That is because of the presence of counter game tries. And yes, if you throw in the fact that when partner is 8-10 or 7-10 the upper end of the range is more common than the botton end of the range, it makes inviting on 15-16 all that more useful, as long as there is room to sort out the difference between a maximum 2M bid and a good medium 2M bid.

[pbleighton]

Josh' point is a good one - heavy vs. light depends on the context.

 

Another example is how light you open, including how light relative to the field in the minimum strength of your opening 1x bids, and how it affects your invitational range - minimum for a 2/1 vs. GF.

 

When I played light openings (5-KQ854-K105-Q874 wasn't a stretch at all) in a relatively sound-opening ACBL field, where the average minimum strength is one to two points higher than this, it was imperative to invite on 11 counts and good 10s, or we would miss a bunch of games the field was routinely bidding. I wasn't going to GF on 12 counts opposite these openings, so my invitational range was 10+ to 13-, a wide range out of necessity.

 

I'm currently playing very sound one bids (13+ unbalanced, 14+ balanced), and we invite with 9-10, GF with 11, and I see no reason to go to 8-10 as an invitational range.

 

Another example is NT range relative to your minimum openers. With strong NT, you have a high percentage of your hands at the bottom of the range (especially in the minors with 5 card majors), which maximizes the number of hands which won't accept an invite, and where you languish in 2NT, the worst contract in bridge. I won't venture numbers here, or Richard will beat me upside the head with a wad of simulation code :lol:

 

OTOH, if you play 12-14 NT and open unbalanced 12 counts (or lower), the percentage of acceptor hands is substantially higher, and you have two incentives for a wider range - fewer 2NT contracts as a percentage of invites, and the risk of missing game opposite balanced 15 counts (again, especially if you are playing in a strong NT field).

 

Peter

[hrothgar]

Your problem is that you have made an assumption, that I believe is in error..

 

Your critique is based on the erroneous asumption that it is the stronger hand that makes the invite. Thus, the light invite crew makes more invites than does the strong invite crew, and, because of the expected distribution of hcp, will reach more games.

You're correct:

 

If you have a situation where a weak hand is making some sort of quantitative invite, the frequencies do reverse themselves.

 

>It may well be that proper analysis would show that the weak hand invites

>family is more numerous than the strong hand invites family, but that analysis

>is far beyond my abilities and time.

 

For what its worth, I think that the weak invite family is substantially more frequent that the strong invite family playing "traditional" system. Following a 1NT opening, the NT opener usually isn't able to make a quantitative invite. This eliminates an awful lot of hands right off the bat. Equally significant, if we look at 1M and 1m opening, single raises are normally more common than game inviational raises. Score another "win" for the frequency of the weak family invites.

 

With this said and done, a partnership could play some games with the way in which they designed their game invitational sequences: In theory, you could try to identify the expected median strength for a singe raise, conditional on opener having sufficient strength for a game invite. This would actually permit you to equalize the frequency of category 1 and category 4. I played arround with this idea when I was looking at the MOSCITO raise structures, but I could never decide if this was a good idea or not.

[Winstonm]

This is an interesting debate but a lot depends on whether opener or responder is making the invite and whether there is/are room/tools for fine tuning. NT auctions seem to be poor examples because they are based primarily on simple addition - do we have 25 HCPs in the combined hands. With 15-17 as a NT range, it would seem the only problem hand would be the responding hand that evaluates to worth 9. As there is still a bonus to bid and make game, the emphasis should be on bidding on to game rather that finding a nifty stop - unless you regularly open 14 point hands within your 15-17 range.

 

From my experience, it seems that those who favor the "heavy invite/light accept" style also tend to open many borderline to weakish hands and respond with weak hands and therefore are more prone to get overboard at the 3-level; those who like "light invite/heavy acceptance" tend to be of more sound opening minds and feel that the opening bid and responses themselves have somewhat narrowed the range to "not too awful" so the only reason to bid on is with extras - either extra shape or extra high cards.

 

It would seem the dividing line might be logic: which is worth more in the long run, a style designed to stop a level lower when 3 doesn't make, which will be worth a 5 or 6 imp pickup when it is right or the try for game style which will gain 7 or 10 when it is right?

 

Maybe I'm being oversimplistic - I'm sure someone will point out the errors of my ways. :lol:

 

Winston

[Free]

I'm a fan of agressive invites and agressive accepts.

exactly my thoughts B)

[Trinidad]

Over 1N, no one plays agressive invites and conservtive acceptences. Its a very bad treatment, for all the reasons layed out in this thread.

This is a simplification again (which isn't necessarily bad, since it is a complicated subject).

 

Give yourself:

♠ Kxx

♥ KJxx

♦ x

♣ Jxxxx

 

Is there anybody who wouldn't bid 2♣ in response to a 15-17 1NT? But what if partner rebids 2♦? Then there is nothing left to do but to make an aggressive invite with 2NT. Fortunately, partner knows that you can have this stretch hand and will accept the invitation only with a real maximum.

 

Now assume that you play the quite popular treatment that 2NT is a transfer to diamonds and that a natural invitation to 3NT goes through Stayman (not promising a four card major). Again, partner will only accept with a real maximum which means that you will have to invite aggressively.

 

So, also after a 1NT opening, lots of players play aggressive invites and conservative acceptances. Not because that is the optimal strategy, but simply because it is necessary to fit the rest of the system.

 

Rik

[whereagles]

I'm a fan of agressive invites and agressive accepts.

That's funny. I'm a fan of no invites. Just bid game :rolleyes: