Re: Point Count vs. LTC (long, lecture/rant in favor of point count)
- From: frisbieinstein@xxxxxxxxx
- Date: 18 Mar 2006 02:50:54 -0800
paulhigh@xxxxxxxx wrote:
Point Count vs. Losing Trick Count or Loser / Cover Card methods:
In a recent thread, the poster discussed overloading Bergen raises (3C
and 3D over one of a major) to portray various "light" and
"heavy" splinters. He defined these in terms of "honor cards"
(A,K,Q, I assume); a light splinter would have three honor cards, a
heavy four, with more Aces than Queens in each case.
Presumably he is thinking in terms of Losers and Cover Cards, ala
George Rosenkranz. A normal minimum opening 1H or 1S bid is assumed to
have seven losers (count each missing A, K or Q in a suit as a loser,
but no more losers than you have cards in the suit), so the light
splinter hopes to cover three losers via high cards and one more for
the singleton; four cover cards opposite seven losers equals ten tricks
(game.)
I'm quite grateful to M. Harrison Gray (Grey?) and Rosenkranz for
introducing me to the concept of valuing a hand in terms other than
high card points, but once you've learned that, isn't it obvious
that a point count APPLIED WITH SENSE is more accurate than trying to
count losers and cover cards?
In its most naïve form, Losing Trick Count values Aces, Kings, Queens,
and any card over three in a suit as being equal in value. One soon
learns that two four-card suits on the side aren't as useful as a
five-bagger; opposite three cards the fourth card in a four-card suit
normally has only a 1/3 chance of becoming a winner while the fourth
card in a five card suit has a 2/3 chance and the fifth card is a near
certain winner. The Italians, I believe, count the fourth card in a
four-card suit as half a loser; the Bissel count counts the fourth card
in a suit as 1/3 of a trick and add 4/3 for each additional card, so
4-3 and 5-3 suits have their proper values of 1/3 and 5/3 tricks (aside
from top cards and ruffing values.)
Then, one learns that high cards perform a triple function -- they
provide winners directly, but they also combine to promote the value of
partner's cards, and they provide control (pulling trumps to prevent
defensive ruffs, stopping a suit long enough to set up a discard,
stopping the defense from pulling trumps while you ruff losers, etc.)
Short suits may provide control and/or ruffing tricks, but they have no
combining value, while long suits provide neither combining value nor
control. Does anyone seriously believe a five-card side suit will
typically be as valuable as an extra King and Queen?
Next, one learns that Queens aren't really as useful as Aces, hence
the usual requirement to "balance" Aces and Queens. But the 4321
count does that automatically: when our side has at least 20 points,
every additional three points adds an average of one trick, so a King
is worth 1 trick, an Ace somewhat more, a Queen somewhat less, and out
of several Jacks we can hope/expect one to prove useful. Note
especially that AKQ in any suit where at least one partner has at least
three cards will normally be worth three tricks. The secret to applying
point count is to recognize those situations where you can expect your
cards to be worth more or less than these average values.
In the specific case of raising with singletons, I have found the old
Goren rules to work very well: add three points for the singleton, add
one for having Jxxx, Qxxx or Kxxx of trumps, subtract one for having
only three small trumps (with Qxx, you would add one and then subtract
one for no net change) , deduct a point for an aceless hand or a high
card in the the short suit. Sometimes 5431 shape will be better than
4441, especially if the five card suit is opposite partner's
doubleton and the three card suit can benefit from a discard, but this
is uncertain; perhaps add one point for this shape if the suit is at
least fair and you can anticipate being able to use the discard..
If you can portray the singleton as with a splinter, partner counts the
Ace of your short suit as only three points, ignores any other honors
in the suit, and add two points if he has at least three cards in the
suit. Add nothing for only two cards and deduct three points for a
singleton or void. (These precise adjustments were never spelled out by
Goren since Dorothy Truscott had not yet invented splinters, nor Oswald
Jacoby his 2NT raise, so it would be rare to have exact knowledge of
partner's short suit.) Per Goren, opener also counts both his short
and long suits after the raise.
An example: AQxxx Kxx Kx Jxx deals opposite Kxxx Axxx Axxx x . One
spade, four clubs (splinter, say 11-14 hcp, four trumps and a
singleton.) Responder values his hand at 15 (11 hcp, 3 for the stiff
club, +1 for the trumps). Opener ignores the club Jack, and counts 12
hcp + 1 for the five card suit + 1 for the doubleton + 2 for having
three or more clubs, 16 total. He figures partner for 14-18 dummy
points and, with 33 possible but not certain, cue-bids 4D. Responder
bids 4H and now opener expresses doubt by bidding 4S. Having only 15,
responder passes, and the near-certain result is 11 tricks, as we would
expect from the combined total of 31 points.
Add either red Jack to either hand, and slam may make on a finesse, but
probably should not be bid given the possibility of a 4-0 trump break
or defensive ruff. Add any two Jacks and the slam is probably worth
bidding. Make responder's hand Kxxx AQxx Axxx x or Kxxx Axx AJxxx x
and slam is excellent (and is likely to have reasonable play opposite
any hand which makes a slam try), but KJxx Axxxx Axx x would be poor.
The first case would surely reach slam on any valuation; the second
counts to 17 if we add for the good shape; the last values only 15 or
16 if responder does not add for trump promotion (trumps are possibly
too good) and/or does not add for the Ace-empty side suit. Under LTC,
case 1 has 6 losers, cases 2 and 3 have 7. Of course an expert might
deduct a loser from case 2, and not from case 3, but how obvious is
that to an average bidder? I personally would rather play with a
partner who might overbid case 3 than one who would not recognize the
potential of case 2 or indeed of KJxx AJxx AJxx x . A valuation system
which purports to ignore Jacks is absurd, and estimating to the nearest
third of a trick is in my (arrogant) opinion clearly superior to
estimating full or half tricks. Schmoints, Points!
I used to put considerable effort in evaluating hands for opening bids.
Then I put my methods to the test and found out they were but
superstition. In my current state of advanced sagacity I use a simple
rule that the random pickup partner will expect. The opening bid is
like the drive in golf -- beginners worry about the drive a lot, good
players just get the thing in play safely. By the time your second bid
comes around you will have some idea where the hand is going, and if
you have a third bid then one can split hairs.
.
- References:
- Prev by Date: Re: Interesting disaster
- Next by Date: Increased penalties?
- Previous by thread: Re: Point Count vs. LTC (long, lecture/rant in favor of point count)
- Next by thread: Bill Hughes Simultaneous Pairs 2006 dom
- Index(es):
Relevant Pages
|
