Mel's Rule: Request for Simulation

[p***@infi.net]

Over at Bridgewinners there has been some discussion regarding a balanced limit raise hand. The auction begins 1H-1NT; 2H and several claim that responder must upgrade to a game force, in effect valuing the third trump in a 6-3 fit at about 2 hcp or half a trick. Someone commented that this represented the difference between a 6-3 and 5-3 fit, but since opener will have already counted something for his six-card suit, I this seems more of a claim about 6-3 vs. 6-2.

In contrast, Mel Colchamiro popularized a guideline for responding to a weak two in a major: add your hcp and trumps, if the total is 17, invite game. If the total is 19, bid game. This implies each extra trump opposite a six card suit is worth about 1 point, which has always been my assumption.

If anyone would card to run a simulation, I’d like to see the results of 6-1, 6-2 and 6-3 fits where the opener has exactly six trumps and no side four-card suit while responder has 1, 2 or 3 trumps and no side doubleton or five card suit. I would fix opener at 6 or 7 hcp (i.e., minimum weak two) while responder has 18 hcp + 1 trump, then 17 hcp + 2 trumps, then 16 hcp + 1 trump. If Mel’s rule is accurate, all three cases should on average have solid and approximately equal play for game. I tried searching here and on BW and did not find any previous results.

Thanks if anyone tackles this!

[t***@att.net]

I don't have a direct answer, but while Opener may know about the 6-card suit, Responder knows whether there are 8, 9, or possibly 10 Trumps. The LTC does suggest removing 1 loser for a known 9-card fit and 2 losers for a known 10-car fit.

If 1H-1NT, 2H shows a 6-card suit, Responder can drop 1 loser for 3-card support and 2 losers for 4-card support. This assumes Responder isn't 4333.

[p***@infi.net]

Post by t***@att.net
I don't have a direct answer, but while Opener may know about the 6-card suit, Responder knows whether there are 8, 9, or possibly 10 Trumps. The LTC does suggest removing 1 loser for a known 9-card fit and 2 losers for a known 10-car fit.
If 1H-1NT, 2H shows a 6-card suit, Responder can drop 1 loser for 3-card support and 2 losers for 4-card support. This assumes Responder isn't 4333.

If responder isn't 4333 then he must have a potential ruffing value, which is a different kettle of fish. But a loser = a trick = about 3 hcp, which strikes me as quite excessive opposite a hand with at least two doubletons and therefore a fair chance of matching responder's doubleton.

[Player]

Get the bean counter to do a simulation. He simulates everything.

[jogs]

Post by p***@infi.net
Over at Bridgewinners there has been some discussion regarding a balanced limit raise hand. The auction begins 1H-1NT; 2H and several claim that responder must upgrade to a game force, in effect valuing the third trump in a 6-3 fit at about 2 hcp or half a trick. Someone commented that this represented the difference between a 6-3 and 5-3 fit, but since opener will have already counted something for his six-card suit, I this seems more of a claim about 6-3 vs. 6-2.
In contrast, Mel Colchamiro popularized a guideline for responding to a weak two in a major: add your hcp and trumps, if the total is 17, invite game. If the total is 19, bid game. This implies each extra trump opposite a six card suit is worth about 1 point, which has always been my assumption.
If anyone would card to run a simulation, I’d like to see the results of 6-1, 6-2 and 6-3 fits where the opener has exactly six trumps and no side four-card suit while responder has 1, 2 or 3 trumps and no side doubleton or five card suit. I would fix opener at 6 or 7 hcp (i.e., minimum weak two) while responder has 18 hcp + 1 trump, then 17 hcp + 2 trumps, then 16 hcp + 1 trump. If Mel’s rule is accurate, all three cases should on average have solid and approximately equal play for game. I tried searching here and on BW and did not find any previous results.
Thanks if anyone tackles this!

Show the link.
Does Mel mention that in general 6331 is better than 6322?

[Fred.]

Post by p***@infi.net
Over at Bridgewinners there has been some discussion regarding a balanced limit raise hand. The auction begins 1H-1NT; 2H and several claim that responder must upgrade to a game force, in effect valuing the third trump in a 6-3 fit at about 2 hcp or half a trick. Someone commented that this represented the difference between a 6-3 and 5-3 fit, but since opener will have already counted something for his six-card suit, I this seems more of a claim about 6-3 vs. 6-2.
In contrast, Mel Colchamiro popularized a guideline for responding to a weak two in a major: add your hcp and trumps, if the total is 17, invite game. If the total is 19, bid game. This implies each extra trump opposite a six card suit is worth about 1 point, which has always been my assumption.
If anyone would card to run a simulation, I’d like to see the results of 6-1, 6-2 and 6-3 fits where the opener has exactly six trumps and no side four-card suit while responder has 1, 2 or 3 trumps and no side doubleton or five card suit. I would fix opener at 6 or 7 hcp (i.e., minimum weak two) while responder has 18 hcp + 1 trump, then 17 hcp + 2 trumps, then 16 hcp + 1 trump. If Mel’s rule is accurate, all three cases should on average have solid and approximately equal play for game. I tried searching here and on BW and did not find any previous results.
Thanks if anyone tackles this!

It might make sense with good controls and a weak
doubleton, but with most such raises I want opener
to be able to bid 3NT with the appropriate texture.

What alternative use are they suggesting for
1M 1NT
2M 3M ?

With doubleton support responder can choose between pass,
2NT, and 3M. With less than a limit raise and a tripleton
responder should have raised directly, or been prepared
to pass 2M.

Fred.

[Barry Margolin]

Post by p***@infi.net
Over at Bridgewinners there has been some discussion regarding a balanced
limit raise hand. The auction begins 1H-1NT; 2H and several claim that
responder must upgrade to a game force, in effect valuing the third trump in
a 6-3 fit at about 2 hcp or half a trick.

I can't find the BW thread you're talking about -- can you provide a
link?

My understanding is that this has little to do with hand valuation, and
is just a matter of practicality. Responder doesn't have any way to
invite game and show his length -- 3H is a 2-card invitation, and a new
suit is natural and weak (although I suppose an impossible 2S could be
used in the case where opener's suit is hearts). So the only game try is
to bid game and have opener try to make it.

[p***@infi.net]

Post by Barry Margolin

Post by p***@infi.net
Over at Bridgewinners there has been some discussion regarding a balanced
limit raise hand. The auction begins 1H-1NT; 2H and several claim that
responder must upgrade to a game force, in effect valuing the third trump in
a 6-3 fit at about 2 hcp or half a trick.

I can't find the BW thread you're talking about -- can you provide a
link?
My understanding is that this has little to do with hand valuation, and
is just a matter of practicality. Responder doesn't have any way to
invite game and show his length -- 3H is a 2-card invitation, and a new
suit is natural and weak (although I suppose an impossible 2S could be
used in the case where opener's suit is hearts). So the only game try is
to bid game and have opener try to make it.
--
Barry Margolin
Arlington, MA

http://bridgewinners.com/article/view/matchpoint-analysis-of-an-unbid-game/

[jogs]

Post by p***@infi.net
http://bridgewinners.com/article/view/matchpoint-analysis-of-an-unbid-game/

This link isn't an article by Mel Colchamiro.

Why would you need a simulation? Do you believe there is any validity to LoTT. It is actually reasonable reliable up to 18 trumps(tricks). Therefore every additional trump is worth a full(or nearly) trick.

[p***@infi.net]

Post by jogs

Post by p***@infi.net
http://bridgewinners.com/article/view/matchpoint-analysis-of-an-unbid-game/

This link isn't an article by Mel Colchamiro.
Why would you need a simulation? Do you believe there is any validity to LoTT. It is actually reasonable reliable up to 18 trumps(tricks). Therefore every additional trump is worth a full(or nearly) trick.

Mel used to write a column in the bulletin. There and elsewhere he popularized his Rule of 17. I was unable to fins any online reference specific to that rule. I'd like a simulation because there seems to be quite a discrepancy in opinions about the value of the third trump in support when partner has shown six. There is not necessarily any conflict with LoTT, where an extra trump adds one total trick, which is not the same as one trick for our side. The BW discussion and Mel's rule both deal with non-competitive sequences and involve bidding beyond our level of trumps, so it cannot be claimed that bidding four may be the best bid even when game rates to fail.

[rhm]

Post by p***@infi.net
Over at Bridgewinners there has been some discussion regarding a balanced limit raise hand. The auction begins 1H-1NT; 2H and several claim that responder must upgrade to a game force, in effect valuing the third trump in a 6-3 fit at about 2 hcp or half a trick. Someone commented that this represented the difference between a 6-3 and 5-3 fit, but since opener will have already counted something for his six-card suit, I this seems more of a claim about 6-3 vs. 6-2.
In contrast, Mel Colchamiro popularized a guideline for responding to a weak two in a major: add your hcp and trumps, if the total is 17, invite game. If the total is 19, bid game. This implies each extra trump opposite a six card suit is worth about 1 point, which has always been my assumption.
If anyone would card to run a simulation, I’d like to see the results of 6-1, 6-2 and 6-3 fits where the opener has exactly six trumps and no side four-card suit while responder has 1, 2 or 3 trumps and no side doubleton or five card suit. I would fix opener at 6 or 7 hcp (i.e., minimum weak two) while responder has 18 hcp + 1 trump, then 17 hcp + 2 trumps, then 16 hcp + 1 trump. If Mel’s rule is accurate, all three cases should on average have solid and approximately equal play for game. I tried searching here and on BW and did not find any previous results.
Thanks if anyone tackles this!

Your request is very specific on responder and not very specific on opener.
Jogs rightly observes that there is a significant difference between 6322 and 6331 distribution.
On the other side if you insist on no side suit doubleton, 3 card support invariably means 4333 and doubleton support and no 5 card side suit means 4432 distribution for responder.
In both these cases simulation has shown that on average you will have better chances to make 3NT than 4M.
I looked at a more common balanced case.
Assume opener is 6322 and responder is balanced (5 card side suit and/or side suit doubleton permissible), how many points do you need combined to make game in a major 50% or better. (Here also 3NT is a strong contender for game)
Distribution of HCP between both hands has not much of an impact, unless one hand is very devoid of HCP.
The result is you need 26 HCP if you got a 6-1 fit, 24 HCP if it is a 6-2 fit, 23 HCP if a 6-3 fit and at least 22 HCP if a 6-4 fit.

However, making both hands balanced defeats the value of surplus trumps to a large extent.
The value of surplus trumps are much more valuable if 2 conditions are met:

1) unbalanced distribution.
2) your meager HCP are made of aces not quacks

If these conditions are met the value of additional trumps increases the trick potential on average by at least 2 HCP each.

[p***@infi.net]

Post by rhm

Post by p***@infi.net
Over at Bridgewinners there has been some discussion regarding a balanced limit raise hand. The auction begins 1H-1NT; 2H and several claim that responder must upgrade to a game force, in effect valuing the third trump in a 6-3 fit at about 2 hcp or half a trick. Someone commented that this represented the difference between a 6-3 and 5-3 fit, but since opener will have already counted something for his six-card suit, I this seems more of a claim about 6-3 vs. 6-2.
In contrast, Mel Colchamiro popularized a guideline for responding to a weak two in a major: add your hcp and trumps, if the total is 17, invite game. If the total is 19, bid game. This implies each extra trump opposite a six card suit is worth about 1 point, which has always been my assumption.
If anyone would card to run a simulation, I’d like to see the results of 6-1, 6-2 and 6-3 fits where the opener has exactly six trumps and no side four-card suit while responder has 1, 2 or 3 trumps and no side doubleton or five card suit. I would fix opener at 6 or 7 hcp (i.e., minimum weak two) while responder has 18 hcp + 1 trump, then 17 hcp + 2 trumps, then 16 hcp + 1 trump. If Mel’s rule is accurate, all three cases should on average have solid and approximately equal play for game. I tried searching here and on BW and did not find any previous results.
Thanks if anyone tackles this!

Your request is very specific on responder and not very specific on opener.
Jogs rightly observes that there is a significant difference between 6322 and 6331 distribution.
On the other side if you insist on no side suit doubleton, 3 card support invariably means 4333 and doubleton support and no 5 card side suit means 4432 distribution for responder.
In both these cases simulation has shown that on average you will have better chances to make 3NT than 4M.
I looked at a more common balanced case.
Assume opener is 6322 and responder is balanced (5 card side suit and/or side suit doubleton permissible), how many points do you need combined to make game in a major 50% or better. (Here also 3NT is a strong contender for game)
Distribution of HCP between both hands has not much of an impact, unless one hand is very devoid of HCP.
The result is you need 26 HCP if you got a 6-1 fit, 24 HCP if it is a 6-2 fit, 23 HCP if a 6-3 fit and at least 22 HCP if a 6-4 fit.
However, making both hands balanced defeats the value of surplus trumps to a large extent.
1) unbalanced distribution.
2) your meager HCP are made of aces not quacks
If these conditions are met the value of additional trumps increases the trick potential on average by at least 2 HCP each.

Thanks! Your results confirm my instinct that 6-3 vs. 6-2 simply isn't worth as much as half a trick or 2 hcp or whatever absent a likely ruffing value. In the BW thread, North's hand was Axx Axxx Qx J108x and after 1S-1NT; 2S several posters claimed that responder "must" jump to 4S. I think that represents poor evaluation on this specific hand (which to my eye does not meet your two conditions.) I should also note that I considered game a poor prospect opposite opener's KJ9xxx Kx 9x KQ9, due to the risk of a club ruff.