Ozymandias wrote: ↑Mon Sep 16, 2019 5:23 pm
lkaufman wrote: ↑Mon Sep 16, 2019 4:27 pm
Does anyone run correspondence or freestyle events with rules designed to avoid or minimize draws, such as specified one-sided openings or startpositions, or modifications of the normal draw rules, or anything else? It seems that attempts to do something about draws should start with correspondence and freestyle events.
Well, you have thematic tournaments, although I don't think they're devised to tackle this particular problem.
At Freestyle, I seem to recall they tried to give a score of 07/0.3 for a stalemate, with very limited effect.
It's true that correspondence and freestyle events should be first to address the issue, but they mostly seem to dance around it.
The Grand Poobahs at ICCF are about as accustomed to innovation as dogs poop .. a rule in place currently (to be changed in January) restricts results of games within the ICCF quarter to no more than 50 to be included in their ELO calculation. If you happen to have completed >50 games your games are NOT included in their dumbass algorithm. Players have lost over 100 ELO points - proof available.
2019-034 Calculation for Fixed Ratings
Proposed by Uwe Staroske, QC - in co-operation with Gerhard Binder, RC
Abstract
Increase the number of 50 to 80 games to change the calculation method of fixed ratings.
Proposal
The proposal refers to item 16 of appendix 1 of the rules (The working rules of the rating system) Change the number of games from 50 to 80:
…except for those players who finished more than 80 games in the current period
Rationale
Players who exceed the threshold of 80 games are rated according to the working formula no. 5, although their rating is fixed.
Assessment
The number of 50 stems from “old times”, in which correspondence chess was not that fast as it is today. Furthermore the number of tournaments has increased considerably. Consequently more players than before surpass this threshold, without doing so deliberately.Effort
Update of the ICCF Rules document + update of the server software.
Modification to the webserver, cost to be estimated (probably low) and reported in the Services Director’s report to Congress. No big risk of errors.
Considerations
Less players than before are going to be rated according to item 5. Therefore in principle it is easier for the players to accumulate positive elo gains. Nevertheless it is necessary to determine a threshold to change the calculation method. If a player exceeds the number of 80 rated games in one elo period, the results of these games describe the player’s playing strength much better than the results from the past.
Since the current rule (threshold of 50 games) came into effect in 2016, about 20 players are pertained by this rule. The increase of the threshold decreases the number of concerned players to the very rare cases (2 players as an average).
Documentation
see above
Comments
7/5/2019 Mariusz Wojnar
The analysis is very superficial, and the conclusions are erroneous. I do not see rational justification for this change.
The rating system cannot be blamed and changed often when it works.
The duration of the tournament has nothing to do with it. Former postal tournaments (from "old times") lasted a similarly long time as now about 2 years (post office was more efficient then!), rarely it was about 3 years.
The real player's strength is demonstrated by performance rating (Rp) and not the small parts of points scored in massive number of weak tournaments! This was the reason for the limitation (the very rare cases, but they did happen) a few years ago (in 2016, it's just 2 years and few months) to 50 parties in the evaluated period, so that the players would not abuse the rating system!
From:
https://www.iccf.com/Proposals.aspx?id=72