eepmeep (John Fernandez) writes:

Subject: Pairing Question
From: (SDchessTDL)
Date: 7/31/2003 1:52 AM Eastern Daylight Time
Message-id:

John,
You should make your first priority pairing players 1 & 2. 3 can
not play them. That leave 4 & 5 who are within 80 points of each
other. 3 has to play 6.
I do not understand your problem. I must admit to pairing without a
computer for more than 25 years. This is simple you are trying to make
it to hard.
Terry Likens

Terry,

I see your point, but why is pairing 1 vs. 2 so absurd? In a 5 person score
group, I don't see why it's unnatural, and I don't see what's wrong with it
from a rules standpoint.

John Fernandez

USCF pairing rules are a collection of "shoulds". the idea is to (try)
to apply these "shoulds" in order. Right at the top is "top half plays
bottom half". Rules near the top *tend* to overrule all of the rules
further down (when they conflict).



--
Kenneth Sloan
Computer and Information Sciences (205) 934-2213
University of Alabama at Birmingham FAX (205) 934-5473
Birmingham, AL 35294-1170 http://www.cis.uab.edu/info/faculty/sloan/

Alan Fifield wrote:

Your tournament was short one round.

I guess 4 rounds of Game/30 isn't enough in an evening. How sad.

If the justification was to force GM's to fight , would you have rejected a
short draw with a double forfeit of both GMs ala Jarecki?

That wasn't the justification, it's just that the players shouldn't be that
averse to it.

Moving players
around in the last round to force "fights" is a bad idea.

I didn't do this. It's just something that I thought was an added benefit after
the fact.

Also you denied two other
players a chance to play a GM.

Fortunately, they'd each already played one.


Better just to pair top down and let the cards
lie where they fall.

Indeed. I guess this is a difference between someone who learned how to pair
using cards (where top-down is the ONLY way to keep your sanity), versus me,
who grew up with the computer, and looks at the group as a whole.

No excuse for bad behavior but it was a bad pairing. If I was you, I'd think
about repairing this relationshp (if possible) rather than banning someone
(which is how I read "no longer welcomed").

I think it's irrelevant, really. This player had been fined already once for
violating the rules of the tournament. It's disrespectful to the sponsors (who
are putting up several thousands of dollars for this event), to the other
players, to the organizers and the arbiter.

Not the end of the world; that's why us TD's get "The Power Bucks".

Remember, I'm a volunteer, and as an organizer, I don't get a single penny from
the event, in fact, I lose a lot of money on it. I do it because I care to do
something good for chess, and because it's been wildly successful, so I don't
mind spending 20 hours a week and what little disposable income I have on this
project.

For a player to throw a hissy fit and storm out jeapordizes the event as a
whole. We can't allow that to happen.

John Fernandez

Ken Sloan wrote:

USCF pairing rules are a collection of "shoulds". the idea is to (try)
to apply these "shoulds" in order. Right at the top is "top half plays
bottom half". Rules near the top *tend* to overrule all of the rules
further down (when they conflict).

Now this is useful. Thank you.

John Fernandez

John,

I think both your points are valid. The rules are often short on the kind of general guidelines Bill states
(not the 400 point, but the others) and you are subject to criticism no matter which path you take. I think
Bill's pairings are great, but I am not sure I would get to them by following what it says in the rulebook.
Especially with regards to the flexibility and priority of deciding who to drop vs. making the pairings
within the score group first!

- Dan H

eepmeep (John Fernandez) wrote in message ...
John,
The piont of a swiss is to pair the top half againest the bottom
half.
You violated the basic swiss pairing rule. I know if I was involved as
one of the players. I would have protested and appealed your ruling. I
also would have helped anyone apealing the pairing.
Terry

I highly doubt it. The pairing is, according to the strict letter of the law,
legal. It's just not ideal, and there's a better way, even though it requires
common sense, and stuff which isn't really in the rulebook.

John Fernandez

John,
I would protest with rule 27A3 pg 80 Top half vs lower half. Then
29C1 on
pg 103 Top half vs lower half ( John this rule keeps repeating because
it is IMPORTANT ) Rule 29D1 pg 104 THE odd player. Then the one you
brought up 29 J1
or J2. In this case instead of pairing with players they should play
they were paired with players rated 400 points higher.
I could go on about the rules I would appeal on. The problem you
have is proving your pairing is legal. He would not be the first Grand
Master to have an organizer pay for their mistake. I know what it is
to volunteer service. You must be able to admit mistakes.
Terry Likens

--

"John Fernandez" wrote in message
...
John,
I would protest with rule 27A3 pg 80 Top half vs lower half. Then
29C1 on
pg 103 Top half vs lower half ( John this rule keeps repeating because
it is IMPORTANT ) Rule 29D1 pg 104 THE odd player. Then the one you
brought up 29 J1
or J2. In this case instead of pairing with players they should play
they were paired with players rated 400 points higher.
I could go on about the rules I would appeal on. The problem you
have is proving your pairing is legal. He would not be the first Grand
Master to have an organizer pay for their mistake. I know what it is
to volunteer service. You must be able to admit mistakes.
Terry Likens

I guess that's your choice. A player appealing pairings is a very lame
thing,
though.

I'm asking on this forum for the specific reason of trying to find out if
it
was a mistake. It seems that it is- but only because those of a decade or
more
ago would have done it a different way. Maybe that's good enough.

John Fernandez

John, it all goes back to what you said about how you learned to pair. I
was a TD in the 70s and 80s [FYI, and you probably don't recognize this name
(look it up if you need to), I was trained by John Osness], and would have
come up with the 'other' solution (dropping #3) as well (almost
immediately). There would have been no question from the players.

The computer can never get the intuition necessary to do this, as the rules
must be ordered for the computer to use them. Kind of like exchange sacs,
eh?

sandirhodes

Eric Mark wrote:

John,

maybe I missed it somewhere in this thread, but what did the computer
want to do in this situation?

The computer dropped the 4 and paired #1 and #2. Since it tends to
overcompensate for colors, I thought that was pretty damn good reason not to
change it. When the computer leaves colors bad when there's another seemingly
possible way to fix them, you have to know something's up. I was able to
logically make sense if it by the "determining the odd player" rule. This led
me to believe that it was the correct pairing.

I am absolutely convinced it is the LEGAL pairing, but it isn't the correct
pairing. This is due to the poorly worded and overly strict ruleset.

Also, you should know this: (I don't.) What do FIDE rules say about
this? As far as I know, there's no 200-point limit there. I know
European players are very sensitive to colors, sometimes even to the
point of trumping score, but that does not seem to be a major factor
here.

It seems pretty hard to follow (FIDE actually is strongly discouraging Swiss
system, it's simply a bad way of pairing), but it seems to me that either
pairing would be acceptable. It seems to select downfloat based on colors, so
it looks like the 1 vs. 4 pairing would happen instead of 1 vs. 2.

There's also a nice rule which I like:

"Once official pairings have been made public, they shall not be changed unless
[a) two players meet more than once; b) a player recieves a bye who has
recieved one; c) a player who will get a bye in the future has receieved one;
d) someone's color difference becomes more than 2 Whites or more than 2 Blacks
e) someone receives the same color 3 times in a row] are violated."

So, whatever the pairing I made, would have stood.

I know a Local TD who would have considered dropping the lowest-rated
2-pointer--or maybe the next-lowest, depending on colors--and treated
the 2.5 pointers and 2-pointers as one 6-player score group.

Yes. I've heard of this.

This guy has never had a GM play in any of his tournaments, though.
Imagine how your irate GM would have reacted to that....

Well, the "6 man group" would have produced the same pairings, after all, there
are only 2 legal ways of doing it.

My understanding is that in such a 6 man group, the player's ranking would
first be determined by score- so that the 1.5 would be #6. This way would get
the "GM/Goichberg Correct" pairings.

Are you sure this TD stuff is more fun than playing?

Cheers,

Eric M

No, but to be honest, it's either do this, or get out of chess completely. I'm
not going to play in this country under an absurd ruleset that aims to turn
chess into a competition of maniacs, judged by a schizophrenic. Nor am I going
to do it with conditions which border on inhumane, with costs which are higher
than any other tournaments in the world.

John Fernandez

Ernest W. Schlich wrote:
An interesting
aside, when you have to change a pairing due to players having met before,
and
then make further changes or transpose more than one spot for color, where
should you calculate the rating difference from, from the original pairing,
from the first spot where the pairing does not violate the not play the same
player twice rule, or elsewhere?
Regards, Ernie

Ernie,

Answer 3) Elsewhere.

Without showing my work, It would always be the spot you are at when evaluating
making switches.

As an example, if a player was originally going to be paired against a 2200,
and now he is temporarily paired against a 2450, I would be willing to make a
switch to a 2500 if it improved the colors. Even though this is a 300 switch
from the original pairing.

Best,
Grant

Ernie Schlich wrote:

I gave the problem to my wife (she was an experienced director in my first
rated event in '72) and she came up with dropping #3 as first try in less
than a minute and wanted to know what the problem was.

It seems to be a "I spent decades doing pairings by hand" vs. "I'm used to the
computer doing them."

I tend to look at the group as a whole, versus top-down, when possible. In this
case, we have a 5 (6 with the lower group) player group, and it's easy to look
at the group as a whole.

If you need to justify it, the comments concerning top half against bottom
half
being the justification carry the most weight with me.

Right. The problem is that Top half vs. Bottom Half (27A3) leads me to 29C1,
which leads me to 29C2 which leads me to 29D1. I was trying to see what
transpositions and interchanges I can make- and it seemed to me that 29D1 and
29J2 were telling me that I can't do 1 vs. 4 and 2 vs. 5, since 29D1 tells me
that 4 should drop, and 29J2 seems to tell me I can't switch 3 with 4 or 5.

It's kind of funny that
after all of John's complaints about the USCF's rules being too detailed, he
is now a little unhappy when the pairing rules are very general.

Well, this was a case where I had a rule which was not very detailed (top half
vs. bottom half) which led me to rules which were very detailed, specifically
the odd player. I didn't know where the odd player rule fit in (and to be
honest, neither did the computer).

When looking through the rulebook over and over, you tend to look for pretty
detailed things, and the odd player rule and 200 point rule stood out. Also,
the computer tends to be VERY insistent on fixing colors, so the fact that it
didn't go out of its way to fix colors was a very strong hint.

An interesting
aside, when you have to change a pairing due to players having met before,
and
then make further changes or transpose more than one spot for color, where
should you calculate the rating difference from, from the original pairing,
from the first spot where the pairing does not violate the not play the same
player twice rule, or elsewhere?

Regards, Ernie
Ernest W. Schlich

This is something I've struggled with, and the rulebook isn't very clear on
this.

It seemed to me that I simply go in rulebook order- and since 29D1 (Odd player)
comes before the first mention of "Natural pairings" (29J5), that I determine
the odd player first.

That's really what I had to go on- the first detailed rule which tells me
EXACTLY what to do and always fits was 29D1.

It seems to me, though, that this is not how pairings are usually done, and
therefore, there should be something between 29C and 29D which details what
exactly the "Natural pairing" is, especially since it becomes very important to
ascertain come 29J.

As far as my general dislike of detailed USCF pairings, this seems to be a case
where the rules are very very detailed, and then the major issue of determining
natural pairings seems to be completely overlooked. I guess it's supposed to be
a "common sense thing", but that doesn't jive with how the rest of the section
is written.

John Fernandez

--

"John Fernandez" wrote in message

It seems to be a "I spent decades doing pairings by hand" vs. "I'm used to
the
computer doing them."

I tend to look at the group as a whole, versus top-down, when possible. In
this
case, we have a 5 (6 with the lower group) player group, and it's easy to
look
at the group as a whole.

Well, maybe it is a 5 (*4* after the drop) player group instead.