ECF Arbiter Regulations

Debate directly related to English Chess Federation matters.
Roger de Coverly
Posts: 21301
Joined: Tue Apr 15, 2008 2:51 pm

Re: ECF Arbiter Regulations

Post by Roger de Coverly » Tue Jan 10, 2017 11:33 pm

E Michael White wrote: he knew of at least some of the other areas of senior arbiter inconsistency. I'm not convinced we get value for lunches out of the senior arbiter title.
Possibly not relevant, but the attitude of some of them to computer usage is completely wrong.

Some of us have been using computer programs since at least the 1970s. If "the computer" doesn't produce the result you expect there are several possibilities:-

The data is incorrectly input. Programs usually use ratings indirectly, by setting a pre-tournament sequence of PIN numbers.
The program has incorrectly implemented the underlying deterministic method of pairing
The logic of the deterministic method is itself incorrect.

Actually I'm not entirely even sure that the concept that pairings should be deterministic is completely accepted. As recently as the 2015 British, it could be suspected that the pairings had been tilted to avoid pairing twin brothers. Their very late appearance on the evening in question being a give-away.

With pseudo-random numbers available, I have a pet theory that pairing methods could be simplified by employing a certain amount of random generation. Such methods fall into disrepute when done manually since you could not tell whether there was an intentional or unintentional bias.

E Michael White
Posts: 1420
Joined: Fri Jun 01, 2007 6:31 pm

Re: ECF Arbiter Regulations

Post by E Michael White » Wed Jan 11, 2017 12:20 am

You've just reminded me I meant to post something about random numbers in a recent thread. To stop giving Mr Zyxel too many byes in too many tournaments it would be possible, at the start of an event for a computer program to generate an additional PIN by pseudo random number generation, for each player and use that as the criterion of last resort rather than alphabetical order. The seed for the pseudo random number generator could be derived from the internal code representation of the tournament name combined with a tournament number issued by FIDE. In that way it would be reproducible as well as relatively random and obviate the need for inserting a bogus zzz player to float to the median. It might even persuade senior ECF arbiters to use programs.

Post Reply