Ockham's razor or K.I.S.S.
Stig Holmquist
that the number of variables or parameters should be as few as
possible in an elegant theory. "It is futile to do with more things
that which can be done with fewer".
In common parlance it means that the simplest method or explanation is
usually the best or correct one. It is often express with the acronym
KISS, which has many interpretations, such as Keep It Simple, Silly,
or Stupid or Shithead. Take your choice in any given situatiion.
On Nov. 8 I posted a discussion of Bergen's six step "Adjust-3"
method for better hand evaluation as part of a post by David
about "improving the point count", It has been deleted but can be
retrieved at Google. I also subitted the very same idea to the ACBL
for possible publication in the Bulletin.
In short, I proposed a much simpler method to obtain the same count.
It was documented with a statistical evaluation of 36 hands generated
with the DMPro. The N hand had 357 Work pts and 365.75 pts based on
the BUM-RAP count advocated by Bergen as the most accurate. By my
slightly modified method the N hands had 365.5 pts, and thus the same.
On every hand illustrated by Bergen I got the same values he did.
Logic would seem to dictate my simpler method is better than Bergen's.
Stig Holmquist
henry...@yahoo.com
In re: Stig's claim about 'deleted' posts:
The thread link is here:
http://groups.google.com/group/rec.games.bridge/browse_thread/thread/f31921a55d53a26e/cdfb31b514189a09?hl=en&lnk=gst&q=#cdfb31b514189a09
It works fine in my browser (Mozilla) and brings up the entire thread.
Henrysun909
Bertel Lund Hansen
> It works fine in my browser (Mozilla) and brings up the entire thread.
Stig himself wrote that the message can be found at GoogleGroups.
Google does not acknowledge cancels. It is therefore possible to
find messages at Google in spite of the writer (or someone else)
having cancelled it.
However I looked up the message-ID - I suppose this is the
central message:
<citlh45nu9n45a1pq...@4ax.com>
This message is available at my Danish newsprovider
(news.stofanet.dk), and my provider *does* acknowledge cancels.
So I doubt that the message is deleted.
--
Bertel
http://bertel.lundhansen.dk/ FIDUSO: http://fiduso.dk/
Will in New Haven
Even if you got very similar, rather than identical, values a simpler
method would be superior. I would think "Bridge World" would be a
better choice than the Bulletin but who knows.
--
Will in New Haven
Alan Malloy
> henry...@yahoo.com skrev:
>
>> It works fine in my browser (Mozilla) and brings up the entire thread.
>
> Stig himself wrote that the message can be found at GoogleGroups.
> Google does not acknowledge cancels. It is therefore possible to
> find messages at Google in spite of the writer (or someone else)
> having cancelled it.
>
> However I looked up the message-ID - I suppose this is the
> central message:
>
> <citlh45nu9n45a1pq...@4ax.com>
>
> This message is available at my Danish newsprovider
> (news.stofanet.dk), and my provider *does* acknowledge cancels.
> So I doubt that the message is deleted.
>
I think that in this context, what Stig means by "deleted" is "I can't
see it in my newsreader". Presumably his news provider keeps a shorter
archive than yours does. At any rate I don't think Stig was claiming
that his article was maliciously deleted (as he has done in the past);
he has simply misspoken due to a misunderstanding of how Usenet works.
--
Cheers,
Alan (San Jose, California, USA)
Bertel Lund Hansen
> archive than yours does. At any rate I don't think Stig was claiming
> that his article was maliciously deleted (as he has done in the past);
> he has simply misspoken due to a misunderstanding of how Usenet works.
I guess you're right. Thanks for the explanation.
Nick France
Oh how quickly one dumps his hero Goren to try and score a debating
point (too bad he didn't achieve that either). Goren's ability to
sell the 4321 point count system was based on making them whole
numbers. Its a simple fact, most people don't like fractions. There
is no way around it.
Sorry Stiq but your way is no simplier than Bergen's way and
personally I think both are more complicated then they have to be and
both have problems in estimating the point count that is more accurate
but has fractions. For Bergen 4 aces adds up to 17 points although
the more accurate point count system he is trying to get to without
fraction gives this as 18 points. As for you, 4 jacks add up to 4
points and not the 3 points given by the system you are trying to get
to.
And, of course, both make no reference to where the honors are (in
long or short suits) or if they are in combination with other honors,
just to mention 2 things. As for most accurate evaluation, neither
comes close to the K&R evaluation of opening hands (but this is for
from simple), which I suggest you look at to see what kinds of
adjusments are needed to evaluate an opening hand.
I think Bergen's attempt at simplification of adjusting the standard
point count doesn't work well this time (no where as simple or as good
as his rule of 20 even if that has flaws), and Stig's claim to have a
better way has the same flaws. I'm happy to see that, at least, he
understands that the Goren point count has flaws and he has come up
with a way to adjust them. That is a good first step in understanding
the value of a hand. Now maybe we won't see any more post for him
that give an evaluation in absolute terms.
Nick France
Stig Holmquist
In a recent thread you wrote: "You are a very bad bridge player and no
one who has played against you would say otherwise".
What makes you suggest Rubens would consider anything written by me
for publication any more. He once printed my short piece about 8-fit
in the Oct. 01 issue on p. 27, but ever since I critizised his old
writeup about the Roth-Stone system he never responds to anything I
submit.
Stig Holmquist
Derek Broughton
> On Nov. 8 I posted a discussion of Bergen's six step "Adjust-3"
> method for better hand evaluation as part of a post by David
> about "improving the point count", It has been deleted but can be
> retrieved at Google.
It has NOT been deleted. It is the essence of Usenet servers that - except
for archives like Google - posts _expire_. Different servers have
different retention periods, but very few would have a 3-month retention,
so that post will have expired on almost all servers.
> In short, I proposed a much simpler method to obtain the same count.
...
> On every hand illustrated by Bergen I got the same values he did.
> Logic would seem to dictate my simpler method is better than Bergen's.
As pointed out, Adjust-3 _is_ only six steps, and you should be doing 3 of
them anyway. At that point, Ockham's razor hardly enters into discussion,
as neither method is complicated. Some people will find Bergen's method
simpler, some yours.
Will in New Haven
I don't know that he would print something by someone who has never
won anything but I bet he WOULD if it were a real improvement and he
was approached pleasantly. A simple count that got more accurate
results would be an attractive concept.
brsri...@gmail.com
Sometimes it is all about judgment. No point count can tell you
accurately every time what to bid. Recently I happened to bid on BBO a
hand to 4H.
I was North. IMPs. It went:
West North East South
P P 3D 3H
P 4H All pass
I held as North:
S KJT4
H Q98
D 932
C AT2
The partner I have never played before held as South:
S Q98
H KJ763
D AT
C KQ4
And West held AT42 in the heart suit we have bid to game.
Just for the record...game made.
Boris
brsri...@gmail.com
And sometimes the opponents misdefend like in this case:
West: S 2 H K943 D J853 C A972 while East: S AKJ5 H T852 D K62 C J5
North(me): S QT9863 H 76 D AQT94 C -- while South: S 74 H AQJ D 7 C
KQT8643
West North East South
P 1S P 2C
P 2S P 3C
P 3D P 3N
P 4D P 4S
All pass
East on lead:
H8/HA/H9/H6
CK/CA/S3/C5
DA/D2/D7/D3
DT/D6/S7/D5
CQ/C2/H7/CJ
HJ/HK/S6/H2
D4/DK/S4/D8
HQ/H3/D9/H5
CT/C7/DQ/S5
SA/C3/S2/S8
SK/C4/H4/S9
HT/C6/C9/ST
SQ/SJ/C8/DJ
I was crazy enough to bid and make this game, probably the only one on
this deal on BBO LOL
Boris
Stig Holmquist
<de...@pointerstop.ca> wrote:
>Stig Holmquist wrote:
>
>> On Nov. 8 I posted a discussion of Bergen's six step "Adjust-3"
>> method for better hand evaluation as part of a post by David
>> about "improving the point count", It has been deleted but can be
>> retrieved at Google.
>
>It has NOT been deleted. It is the essence of Usenet servers that - except
>for archives like Google - posts _expire_. Different servers have
>different retention periods, but very few would have a 3-month retention,
>so that post will have expired on almost all servers.
My server keeps threads back to 4/15/08. So why was this one deleted
as were other posts with my name on them? It looks deliberate. Wake up
Stig
Will in New Haven
> On Sat, 21 Feb 2009 14:11:24 -0400, Derek Broughton
>
> <de...@pointerstop.ca> wrote:
> >Stig Holmquist wrote:
>
> >> On Nov. 8 I posted a discussion of Bergen's six step "Adjust-3"
> >> method for better hand evaluation as part of a post by David
> >> about "improving the point count", It has been deleted but can be
> >> retrieved at Google.
>
> >It has NOT been deleted. It is the essence of Usenet servers that - except
> >for archives like Google - posts _expire_. Different servers have
> >different retention periods, but very few would have a 3-month retention,
> >so that post will have expired on almost all servers.
>
> My server keeps threads back to 4/15/08. So why was this one deleted
> as were other posts with my name on them? It looks deliberate. Wake up
It would be wrong to do that but it sure would be tempting. But I
don't know anyone who would have the power to do it and I don't think
anyone should have.
Martin Ambuhl
> My server keeps threads back to 4/15/08. So why was this one deleted
> as were other posts with my name on them? It looks deliberate. Wake up
I had no idea you were so important. No doubt there is a special
building in Langley dedicated to censoring your writing. We have
already found that your thoughts are so deep and troublesome that the
_Bridge World_ and ACBL _BUlletin_ quake at the thought of publishing
them, and that your writing is so sublime that only a conspiracy of the
highest order would keep the world from seeing it published. Now we
find that your news server has folding before the enormous pressure of
this conspiracy and has hired a team with the sole purpose of deleting
material which might criticize an important bridge write. I think it's
time that _you_ wake up: your nightmare is your own creation.
Bertel Lund Hansen
> My server keeps threads back to 4/15/08. So why was this one deleted
> as were other posts with my name on them? It looks deliberate. Wake up
If it was deleted with a cancel, it ought to be deleted from my
newsserver as well.
Bertel Lund Hansen
> It would be wrong to do that but it sure would be tempting. But I
> don't know anyone who would have the power to do it
Most people have, but they do not have the technical knowledge to
set it up, and I won't provide the recipe. It's not too
difficult.
Bertel Lund Hansen
> I had no idea you were so important.
I will not take part in any exchange of nicities, but I will
point out that cancels sent by someone other than the writer do
happen.
Stig Holmquist
Your reference to the K&R method is useless to intermediate players,
who have never heard of it and much less know where to find it .I know
of only one player amongst 400 at my local club who can use it.
Stig
Nick France
>
> - Show quoted text -- Hide quoted text -
>
> - Show quoted text -
If you want to get better, you should become that one and try to
understand the concepts behind it. Most bidding is learning to adjust
the value of the hand. Let me strongly suggest "The Secrets of
Winning Bridge" by Jeff Ruben to help you along.
As for K&R you need to start running in a better bridge circle. Most
good players have heard of it. Its not something most can actually use
at the table but it is an evaluation to go back to afterwards to see
if your judgment on opening was good or not.
Now a good place to put some of your concepts to the test is to come
to the Sectional this weekend in Hamden. you only can improve by
playing up.
Will be looking for you there
Nick France
Stig Holmquist
As for K&R you need to be reminded of a few facts about it.
Rubens rejects the term K&R and asks you instead call it the 4C or K4C
metod. He does not wish to be identified with it. I once asked him for
an explanation to one specific part of it and he responded I should
ask Kaplan, who had been dead a few years then.
As for the 4C method itself, Kaplan wrote: "4C stands for Caution!
Complex Computer Count". "this computer count was devised for use by
machines; it is not meant for mere mortal Masters. The most practical
use for our count is for ...post mortems".
Unless you have very old issues of BW from Oct.82 you would not be
able to learn about 4C. A few websites offer the ability to apply the
4C to any hand, but they do not get the same values. Good luck.
Stig Holmquist
Derek Broughton
> On Sat, 21 Feb 2009 10:40:34 -0800 (PST), Nick France
> <gand...@att.net> wrote:
>
>>And, of course, both make no reference to where the honors are (in
>>long or short suits) or if they are in combination with other honors,
>>just to mention 2 things. As for most accurate evaluation, neither
>>comes close to the K&R evaluation of opening hands (but this is for
>>from simple), which I suggest you look at to see what kinds of
>>adjusments are needed to evaluate an opening hand.
>
> Your reference to the K&R method is useless to intermediate players,
> who have never heard of it and much less know where to find it .I know
> of only one player amongst 400 at my local club who can use it.
But Stig - none of them have heard of the Holmquist method either! Not
having heard of the method is no argument against it - that becomes just a
matter of education.
Derek Broughton
> My server keeps threads back to 4/15/08. So why was this one deleted
> as were other posts with my name on them? It looks deliberate. Wake up
and btw, servers can usually expire different groups after different
periods - does yours keep rec.games.bridge back to last April?
Derek Broughton
> Stig Holmquist skrev:
>
>> My server keeps threads back to 4/15/08. So why was this one deleted
>> as were other posts with my name on them? It looks deliberate. Wake up
>
> If it was deleted with a cancel, it ought to be deleted from my
> newsserver as well.
Actually not. Cancels are propagated from server to server, but most
servers ignore them.
Derek Broughton
> On Sat, 21 Feb 2009 14:11:24 -0400, Derek Broughton
> <de...@pointerstop.ca> wrote:
>
>>Stig Holmquist wrote:
>>
>>> On Nov. 8 I posted a discussion of Bergen's six step "Adjust-3"
>>> method for better hand evaluation as part of a post by David
>>> about "improving the point count", It has been deleted but can be
>>> retrieved at Google.
>>
>>It has NOT been deleted. It is the essence of Usenet servers that -
>>except
>>for archives like Google - posts _expire_. Different servers have
>>different retention periods, but very few would have a 3-month retention,
>>so that post will have expired on almost all servers.
>
> My server keeps threads back to 4/15/08. So why was this one deleted
> as were other posts with my name on them?
Then get a better server. Most servers don't permit "cancels" (because
there's no way to verify who cancelled the post). If your server actually
permits anybody to cancel anybody else's post it's pretty much pointless to
use it as you can't rely on anything (and if somebody really _were_
maliciously deleting your posts - and I'm the one you accuse of doing it -
consider yourself lucky. If your server permits "cancels", it almost
certainly permits "supercede" - which means a person could replace your
post with something else. Hmmm. Maybe that's why your posts so often make
no sense...).
> It looks deliberate. Wake up
Only to the paranoid. Nobody else has seen any evidence that your posts
disappear. More's the pity.
mors...@gmail.com
> On Sat, 21 Feb 2009 15:44:15 -0800 (PST), Nick France
>
>
>
4C) evaluation.:
You could download it, and input hands to it, and it would tell you
the count. If you know PERL you can mod the program to include your
own evaluation method and then compare the two.
His page seems to be blocked now, and I cant find my perl copy. I have
a C copy but that is harder to read.
If anyone has a copy of Jeff's PERL program, could they post it (or a
link to it) here?
pgmer6809
mors...@gmail.com
> Ockham's razor is a guiding tenet or maxim in science. It implies
> that the number of variables or parameters should be as few as
> possible in an elegant theory. "It is futile to do with more things
> that which can be done with fewer".
>
> In common parlance it means that the simplest method or explanation is
> usually the best or correct one. It is often express with the acronym
> KISS, which has many interpretations, such as Keep It Simple, Silly,
> or Stupid or Shithead. Take your choice in any given situatiion.
>
> On Nov. 8 I posted a discussion of Bergen's six step "Adjust-3"
> method for better hand evaluation as part of a post by David
> about "improving the point count", It has been deleted but can be
> retrieved at Google. I also subitted the very same idea to the ACBL
> for possible publication in the Bulletin.
>
> In short, I proposed a much simpler method to obtain the same count.
> It was documented with a statistical evaluation of 36 hands generated
> with the DMPro. The N hand had 357 Work pts and 365.75 pts based on
> the BUM-RAP count advocated by Bergen as the most accurate. By my
> slightly modified method the N hands had 365.5 pts, and thus the same.
> On every hand illustrated by Bergen I got the same values he did.
> Logic would seem to dictate my simpler method is better than Bergen's.
>
> Stig Holmquist
Danny Kleinman in his book "The No Trump Zone" describes an evaluation
method he says is good for balanced hands.
It is based on a Bridge World Sptember 2001 Article.
starting point: A=13, K=9, Q=5, J=2 (Total is 29)
.
Then you add some points for honors in combination:
Each honor in combination with an A gets an extra HCP. So AK=23,
AQ=19, AQJ=23 etc.
Each lower honor in combination with a King also gets an extra HCP (if
not already in combination with an Ace.)
So KQ= 9+5+1 = 15, KJ= 9+2+1=12, KQT= 9+5+1=15 etc
Tens get 1 point if accompanied by an honor or a 9. So AJT = 13 + 2 +1
= 16. QJT = 5 + 2 + 1 = 8. also QTx = 5 + 1 = 6. etc.
Kleinman says that the 'average hand' will include one honor in
combination with another, so the 'average' hand will have 30 Kleinman
points.
You now divide your Kleinman Points by 3 and you get your 'equivalent'
Work HCP. However there is a bonus. If your remainder is 0 you have a
'poor' hand for your point count.
If your remainder is 1 you have an average hand for your point count,
and if your remainder is 2 you have a good hand for your point count.
So AKxx Kxx xxxx xx gives 13 + 9 + 1 + 9 = 32 Kleinman Points or a
'good' 10 HCP.
but KQxx QJxx QJT xx is worth 9+5+1 + 5+2 + 5+2+1 = 30 Kleinman points
or a 'bad' 10 HCP.
As with any evaluation method you subtract 1 Kleinman point for honors
in short suits: stiff K=9-1=8; Q or Qx = 5-1=4; J, Jx or Jxx=2-1=1.
and also for 4333 shape. Which loses 2 Kleinman Points.
The PERL version of the KnR algorithm by Jeff Goldsmith can be found
at:
http://www.jeffgoldsmith.org/perl/knr.txt
and it includes the Kleinman count as well as the KnR count.
pgmer6809
pgmer6809
I found it. Jeff now has his own domain name. The program is at:
http://www.jeffgoldsmith.org/perl/knr.txt
Bertel Lund Hansen
> Actually not. Cancels are propagated from server to server, but most
> servers ignore them.
Quite. That is why I supplied the forum with the information that
my newsprovider does not ignore cancels.
Stig Holmquist
Now there is a simple method if I ever saw one. It will be very
popular with intermmediate players as an alternativeto Bergen's.
Stig Holmquist
brsri...@gmail.com
> On Sat, 21 Feb 2009 21:56:47 -0800 (PST), "morse1...@gmail.com"
>
> - Prikaži citirani tekst -- Sakrij citirani tekst -
>
> - Prikaži citirani tekst -
Any point count method that I know of is flawed. It does not work that
way. Counting your fingers won't improve your bidding and playing
skill.
It will only abate the true potential of a young player. Everyone has
to discover the game by himself. A basic point count method is as good
as any. There are too many variables.
Boris
brsri...@gmail.com
brsrich...@gmail.com je napisao/la:
> > - Prika¾i citirani tekst -- Sakrij citirani tekst -
> >
> > - Prika¾i citirani tekst -
>
> Every point count method that I know of is flawed. It does not work that
boblipton
wrote:
Wouldn't it be easier to position mirrors unobtrusively behind the
other three players and move them around as people move?
Bob
David Babcock
> Counting your fingers won't improve your bidding and playing
> skill
but at least it short-stops the discussion of whether to use
fractions...
> It will only abate the true potential of a young player.
...
> A basic point count method is as good
> as any
If the young player is taught what the point count is intended for,
and can "get" that, an improved point count may be of some marginal
value at the appropriate time in his progress. I don't think anyone
here is suggesting anything more than that.
If your point is that Bergen did not do *his audience* -- the ACBL
rank and file, including the newbies -- any favors by publishing
Adjust-3 in the Bulletin, I'm with you. The good news is that at my
club, the day-to-day C players found Bergen's article totally
incomprehensible. :-))) (For non-ACBL'ers: C is the lowest of the
three strata in the stratification system widely used at ACBL
clubs.)
David
Derek Broughton
> On Feb 22, 8:03 am, brsrich...@gmail.com wrote:
>
>> Counting your fingers won't improve your bidding and playing
>> skill
>
> but at least it short-stops the discussion of whether to use
> fractions...
I have a horrible vision of using Occam's razor to shorten one's fingers...
Stig Holmquist
Thank you for making my day and confirming my verdict of Adjust-3 as
convoluted. How could it have been otherwise when Bergen was assisted
by HS909, whose mastery of logic and English is doubtful at best.
I hope you have informed the ACBL Editor of your observations.
Stig
brsri...@gmail.com
I don't even know what this thing is, it surely sounds like a medieval
torture, the infamous "thumbscrew torture" comes to mind.
http://www.medievality.com/thumbscrew.html
Derek Broughton je napisao/la:
henry...@yahoo.com
Whew! That's a stinger of an insult there ol' Stiggy-poo.
But at least I'm in good company:
Larry Cohen, Eddie Kantar, Danny Kleinman, Richard Oshlag, and Jeff
Rubens are all mentioned, along with me, in the forward to Marty's new
book.
I do not know when Adjust-3 was published in the ACBL Bulletin, but
I'd be willing to bet Stiggy a fair amount of money that it was
accepted for publication BEFORE I started working with Marty on his
new book. That would, of course, imply that my contribution to his
idea is, well, zero.
How regrettable that realizing such a thing would render Stig's
ongoing ad hominem attack meaningless.
Henrysun909
David Babcock
> Thank you for making my day and confirming my verdict of Adjust-3 as
> convoluted.
You're welcome I'm sure, but any method with fractions is not too
likely to win a KISS award either, IMHO.
David
Will in New Haven
wrote:
Rubens insulted him, by telling him to ask someone who had been dead
for some time, but I don't Stig even _got_ it.
--
Will in New Haven
Derek Broughton
> Whew! That's a stinger of an insult there ol' Stiggy-poo.
>
> But at least I'm in good company:
>
> Larry Cohen, Eddie Kantar, Danny Kleinman, Richard Oshlag, and Jeff
> Rubens are all mentioned, along with me, in the forward to Marty's new
> book.
>
> I do not know when Adjust-3 was published in the ACBL Bulletin
Jan/Feb 2009
> , but
> I'd be willing to bet Stiggy a fair amount of money that it was
> accepted for publication BEFORE I started working with Marty on his
> new book.
I hope you're talking about about a new not-yet-published book. Since it
was in Bergen's regular monthly column, I'd be surprised if it's more than
6 months old.
> How regrettable that realizing such a thing would render Stig's
> ongoing ad hominem attack meaningless.
Henry, you're making Stig's point for him about your mastery of logic - "ad
hominem" attacks are _inherently_ meaningless :-)
henry...@yahoo.com
I started working on "Slam Bidding Made Easier" in October 2008 (I
think; perhaps it was September 2008). That book went to the
publisher in January 2009 and I received my copy in late January, I
think.
If the adjust-3 hit the Bulletin in Jan 2009, then it was probably
submitted for publication sometime that summer, if my knowledge of
publishing turn times is approximately correct. In any event, I no
longer receive the Bulletin, and I did not know that adjust-3 was
something Marty was already working on before he started sending my
pages of his manuscript.
Of course, you are correct that ad hominem arguments are inherently
meaningless. As I recall, they were prominently featured in my class
on the symbolic logic some decades ago as a fundamental fallacy of
argument.
Still, ad hominem arguments, flawed and fallacious though they may be,
somehow still seem to get substantial credence, especially from
trained chemists who cannot muster up convincing argumentation in
favor of his bridge points.
Henrysun909
Carl
wrote:
>
> I started working on "Slam Bidding Made Easier" in October 2008 (I
> think; perhaps it was September 2008). That book went to the
> publisher in January 2009 and I received my copy in late January, I
> think.
>
> If the adjust-3 hit the Bulletin in Jan 2009, then it was probably
> submitted for publication sometime that summer, if my knowledge of
> publishing turn times is approximately correct. In any event, I no
> longer receive the Bulletin, and I did not know that adjust-3 was
> something Marty was already working on before he started sending my
> pages of his manuscript.
>
>
Marty Sez:
Carl,
Adjust-3 was conceived by me in 2008.
It represents an easier method to count HCP
than the very accurate but cumbersome
computer-tested HCP:
4 1/2, 3, 1 1/2, 3/4, 1/4 HCP for ace thru ten.
You are obviously welcome to quote me.
Regards,
Marty
Bill Jacobs
>
> What makes you suggest Rubens would consider anything written by me
> for publication any more. He once printed my short piece about 8-fit
> in the Oct. 01 issue on p. 27, but ever since I critizised his old
> writeup about the Roth-Stone system he never responds to anything I
> submit.
>
> Stig Holmquist
There might be a lesson for you in that :)
Cheers ... Bill
Stig Holmquist
I hope you do not include half points as fractions even thought in
formal math they are. If players cannot handle half points they will
have problems with quick tricks such as AQ and Kx,
Stig
David Babcock
> On Sun, 22 Feb 2009 13:46:00 -0800 (PST), David Babcock
> >You're welcome I'm sure, but any method with fractions is not too
> >likely to win a KISS award either, IMHO.
>
> >David
>
> I hope you do not include half points as fractions even thought in
> formal math they are. If players cannot handle half points they will
> have problems with quick tricks such as AQ and Kx,
But are quarters really where the problem is? In counting a la
Martelli, you use a quarter *once*, after it dawns on you that Work,
plus half the excess of aces vs. queens, plus a quarter of the excess
of tens vs. jacks, gets you there. Can one really have trouble with
this unless one has trouble with fractions from square one? I am with
Bergen in thinking that *if* Martelli is to be usefully simplified,
going to whole numbers is the only thing that might get it done for
the average bear. I question only whether Marty achieved what he was
aiming at.
Perhaps of interest, though not proving anything one way or another in
the current discussion, Martelli himself discussed some
simplifications at
http://groups.google.com/group/rec.games.bridge/msg/e258d42534c24716
David
Stig Holmquist
They also say : "Subtract 1 for any suit whose lowest card is higher
than a 10". That looks backwards and could better have read :deduct 1
from any hand with no honor cards. Have I missunderstood them?
Stig
Will in New Haven
It _does_ look wierd but it may simply be an odd way of putting it.
Honor cards in suits like KQJ, KJ, AKJ, AQ, and even AKQ don't pull
quite the weight they do with a bunch of Xes behind them and that is
their way of phrasing it.
Douglas,
: Ockham's razor is a guiding tenet or maxim in science. It implies
: that the number of variables or parameters should be as few as
: possible in an elegant theory. "It is futile to do with more things
: that which can be done with fewer".
Unfortunately that statement is not really tenable any more.
Go to http://www.csse.monash.edu.au/~webb/
and look at the Occam's Razor papers.
Basically he constructs simple classifiers, chooses the one with
the best performance and then adds complexity to it and finds
that it produces improvement in accuracy.
OR is a nice bit of urban lore but that's all.
Douglas,
Tasmania
Derek Broughton
That's just plain silly. Even Occam recognized that the _right_ theory
might not be the simplest. Occam's razor is a way to choose between two
unprovable hypotheses, with a little more accuracy than flipping a coin.
The existence of life itself gives evidence that the simplest solution is
not always right. Nevertheless, Occam's razor is still a guiding tenet in
science. Without it scientists would just accept incredible kludges like
dark matter or epicycles, and then we'd be nowhere.
--
derek
jcri...@yahoo.com
> Douglas, wrote:
> > On Feb 22, 12:22 am, Stig Holmquist <stigfjor...@hotmail.com> wrote:
> > : Ockham's razor is a guiding tenet or maxim in science. It implies
> > : that the number of variables or parameters should be as few as
> > : possible in an elegant theory. "It is futile to do with more things
> > : that which can be done with fewer".
>
> > Unfortunately that statement is not really tenable any more.
> > and look at the Occam's Razor papers.
> > Basically he constructs simple classifiers, chooses the one with
> > the best performance and then adds complexity to it and finds
> > that it produces improvement in accuracy.
>
> > OR is a nice bit of urban lore but that's all.
>
> That's just plain silly. Even Occam recognized that the _right_ theory
> might not be the simplest. Occam's razor is a way to choose between two
> unprovable hypotheses, with a little more accuracy than flipping a coin.
> The existence of life itself gives evidence that the simplest solution is
> not always right. Nevertheless, Occam's razor is still a guiding tenet in
> science. Without it scientists would just accept incredible kludges like
> dark matter or epicycles, and then we'd be nowhere.
> --
> derek
I once came across the advice: "Do not shave yourself too closely with
Ockham's Razor".
Steve Willner
> OR is a nice bit of urban lore but that's all.
This is a good example of what Webb appears to be arguing against:
reaching a general conclusion on the basis of limited data. _At most_,
the Webb papers are concerned with the limited field of machine
learning, and Webb appears to be using a field-specific concept of what
"Occam's Razor" means.
Derek Broughton <de...@pointerstop.ca> wrote:
> Nevertheless, Occam's razor is still a guiding tenet in
> science. Without it scientists would just accept incredible kludges like
> dark matter or epicycles, and then we'd be nowhere.
Dark matter is a great example. When it was first suggested in the
1930's, nobody took it seriously. Now that there are no fewer than five
independent lines of evidence for it, it's the standard theory. That
doesn't mean a better theory won't come along some day, but if it does,
it will either be simpler or will explain more data.
Applied to bridge, the maxim probably says something like "don't make
your methods more complex unless doing so will improve your results."
Unfortunately, that leaves it to be determined whether any particular
complexity will improve your results or not.