October 2017 - 7

➺ InBox
Wanted: effective rating system

When I was a relative newcomer to
duplicate bridge, I wrote a letter to the
editor that was published in the April
2008 issue. The main point of my letter
was that the number of masterpoints
a player had might not be representative of how good they are, and I wanted
to try to measure my improvement (if
any) as I tried to progress.
But factors such as how often someone plays, the strength of the field in
the game, the competency of one's
partner (playing with a "pro"?), single/
double/triple points from special
games, number of pairs in a section,
etc., confused the issue.
Stratified games helped newcomers like me, and I even received points
scoring less than 50%, which I suppose
is OK, but the real questions were how
good of a player was I, and was I improving? The number of masterpoints
I accumulated did not represent my
strength, and people even joked that it
was more like an attendance record.
Chess players have a rating which
goes up and down based upon their
performance, taking into account the
strength of their opponents. Beat a
stronger opponent and gain a lot of
points. Beat a weaker opponent and
gain fewer points. Lose to a stronger
opponent and lose a few points, etc.
The editor responded to my letter
by saying that a long-term goal was to
basically develop a rating system that
reflected current playing strength,
as well as to retain the cumulative
masterpoints earned. This seemed like
a great idea to me, sort of like a baseball player having lifetime statistics
recognizing the achievements of his
career, and current season statistics. I
just figured I'd continue playing, trying
to improve, and see when the ACBL
would act on this, although longtime
players told me not to hold my breath!

Letters to the Editor

Well, it's been over nine years and
nothing has been done. It seems that
now is the time to actually do something about this. With the advent of
electronic scoring in most clubs, all the
data about a player and his partner and
opponents is available. It should be reasonably easy to develop an algorithm to
determine the current playing strength
of a player. In my opinion, the current
playing strength would not be an exact
number to hang your hat on, but rather
a good indication of the skill the player
has. I'll leave the details on how to do
this to the ACBL to figure out.
Just to be clear, I am not saying to
eliminate masterpoint totals earned
over a lifetime of playing, just to add a
new measure: current playing strength.
BARRY ELBERG
Westbury NY
It has long been a topic of discussion
in the North American bridge community that the current system of ranks,
based on lifetime accumulation of
masterpoints, is not a good reflection
of current ability. New, talented stars
are often forced to play in team events
well below their ability level, and
older, long-standing players are often
forced to play in events that far exceed
their skill level based on a long bridge
career's accumulation.
I would like to propose we supplement the current rank system with
a rating system. Rank would be unchanged, based on lifetime points. Rating would adjust up or down based on
recent (perhaps last five years?) performance. And stratification/bracketing/event eligibility would be based
on rating instead of rank. I believe this
solution would improve the situation
for everyone and has little downside.
MARTY DENEROFF
New York NY

Respect for Goren

I was reminded, by Harris J. Levey's
letter about the value of counting short
suits (July 2017), of something which
has puzzled me for years: the widespread misunderstanding and neglect
of the Anderson-Goren distributional
point count.
For opening bids, Goren made it
clear that this method valued long
suits and merely counted short suits as
a convenience. Indeed, his point count
initially attributed 1 point to every
card beyond four in the longest suit
and 1 point to every card beyond three
in other suits. If there are long suits,
there must be shortage somewhere,
and exactly the same result is attained
by the shortcut of counting for short
suits. Either way amounts to counting
1 point for each card by which a hand
differs from flat distribution.
It is remarkable that points for
length can be assigned on the same
scale, but the 3 points counted for a
void are not really different from the
points accorded to a king. The 3 points
represent the value of holding suits
which could make a good trump suit
or side suit, just as the 3 points for the
king represent the possibility of making good use of that card in the play.
They are both probabilities only. The
void may not indicate length worth
3 points if there is a misfit; the king
may not be worth 3 points if the ace is
unfavorably located.
It is easy to see that a hand which
just satisfies the Rule of 20 counts
either 13 or 14 in Goren's method, so
the difference would appear to be in
a range where players might disagree
about whether a hand should be
Letters to the editor are welcome by regular mail and email.
Brevity is considered a virtue, as is subject matter with relevance
to a majority of ACBL members. Unsigned letters are not
considered for publication, so please include your name and
hometown. Letters may be edited. All letters will receive a
response. Send emails to editor@acbl.org.

October 2017 Bridge Bulletin

Table of Contents for the Digital Edition of October 2017