THE WITHDRAWAL EFFECT


By Morten Lille°ren

 

THE WITHDRAWAL EFFECT

Some years ago I played against an elderly english gentleman by the
name of Bernhard Hanison. When playing, we exchanged a lot of of thoughts,
told a little about ourselves, and in general, enjoyed what we were doing.
One thing he told me, amazed me: He was playing more than 100 games simultaneously.


Me myself, I have a rule (that I break now and then...) never to have more than
30 games going at the same time. The reason is simple: If I exceed that number,
both my playing-strength and standard of life deteriorate. But not so with Hanison.
He enjoyed it. We finished our games,waved each other goodbye and went on
with our games and lifes.


Now, I'm a curious little bastard, so every now and then I take a look at
tournament tables and ratinglists to see how my former opponents are doing.
Some time after our games, I therefore noticed that his rating had dropped.
And not only dropped - he'd gone from 2299 down to 1855. He couldn't have
lost that much playing-strength within a couple of years. This is what I think
had happened: For some reason he'd withdrawn from play for a time (illness?).
Since he had many games going simultaneously, he lost them all.


Even if he had a legit reason for withdrawing, his ratingdrop couldn't
be prevented. And the more games, the larger drop. My "theory" is confirmed
by the fact that today Hanison is above 2200 again.


Why do I tell this story? Well, in 17.OL preliminaries,section 1, bd. 2
I was up against Khaled Chorfi. He's rated 1941. Nothing wrong with that.
But the fact that he was a SIM and ranked at bd.2 in such a team competition,
rouse my suspicions: And yes, some years ago(2002) he too had to withdraw from
all his games. He had, during the 6 years before the withdrawal, been stable above
2400 (hence the SIM-title) and then dropped around 500 elo-points in one go.
 

Being a curious nature (see above), I started to look into the eloquery to see if I
could find some other pecularities. And indeed: The lowest rated GM is
Ruben Berdichevsky from Argentine. He fell from 2633(2000) to 1843.
As far as I can see, he had to withdraw from at least 5 tournaments.
The lowest rated SIM is Abraham Raul Ramirez, also from Argentine.
He fell from 2525 (1996) to 1826 today (I'm not sure why, but it is an oddity)
And last, but not least: IM Jose Ortiz Elias from Peru, who fell from 2461
all the way down to 1539, more than 900 pts. If that'd been his elo today,
he would have been among the last 25 on the entire ICCF-list. But he has
started his trip up again, and is rated 1611 today.


There are other titled players below 2000, but I think I've made my point
(and this is NOT written to "out" players).
If an otb-player gets sick during a tournament, he withdraws from that game,
loses 5-10 elo and stays away from chess until he's well again. Not so with
us: We may have to withdraw from all our games, and end up with a huge ratingdrop.
So what? Isn't it only a little side-step, and after some time the rating's back approx.
where it started? (See Hanison above).


Yes - and no. My minor objection is that our low elo "ruins" the tournaments
we take part in after the drop: I may have been mistaken, but when I checked,
the 17.OL preliminaries, section 1, bd 2 dropped several title-categories down
because of Khorfi's participation in that group. This means that to achieve a norm,
one has to score much more points than what would have been necessary if he'd not participated. I didn't look too closely, but I think the other players have to score
1,5 point more in 10 games than they would have had to in 9 games if he hadn't played.
 

And, as you know, it's not possible to score more than one point in a game.
This means that his rating forces the other players to make a better score against
each others, if they want to achieve the same result as without him.


And this has to take place, amongst other things, against a player who
is grossly underrated... Which brings me to my main point:
Ratings are there to reflect a player's strength - nothing less and nothing more.
Of course we're all underrated, but the examples above are in a category of their own.
And surely, it contradicts the point of the whole rating system. It's simply no connection between the rating and the strength anymore. Which means that the system has failed.


What to do then?

I don't have any bright solutions to the problem. As long as it is a minor
problem (for the system), the best thing is maybe to leave it as it is.
But on the other hand, I don't think it's fair against the players who have a
perfectly legitimate reason for withdrawal. If they start playing again, their
starting-point is way below what they'd probably have if they enter as newcomers.
And this fact points against a possible remedy:


1.They can start up again, accordig to the class-initiated rating that ICCF
give new players (Masters have a preliminary rating of 2200 aso.), and

2. Even if they've played several hundreds of games in the past (as they all have),
their rating for the next 30 games can be set as unfixed,as is done with newcomers
today, and thereby speed up their "recovery". As I'm a novice in this field, I don't know
what it'll do with the "pool" of all the players' elo. Hopefully it will not push it off
balance and start a process like the greenhouse effect.
Taking a step aside, this can't be said to be the biggest injustice in the world today.
But there IS a small discrepancy, possibly one that can be dealt with! :o)

  

 
    

 
 

Main Page | Issue 01 | Issue 02 | Issue 03 | Issue 04 | Issue 05 | Issue 06 | Issue 07 | Issue 08
 
 

 

Copyright 2004,  International  Correspondence  Chess  Federation

    This page is maintained by     Raymond Boger