10B. New Business - For Discussion Only - Conversion of Ratings between Rating System

New Agenda Item for discussion only. (Received from Paul Leblanc, CFC Ratings Auditor)

Part 1

CFC/FIDE/USCF/FQE Rating Gaps. Many CFC players have more than one rating. It is not uncommon for a CFC member's other rating to be more current and hence more accurate than their CFC rating. This issue was brought forward to me by John Coleman who observed that many of the juniors in the Windsor area play more USCF games than CFC games and because they are improving rapidly, their CFC rating is significantly lower than their USCF rating.

I would like to know if there is any support for an amendment to the Handbook to allow a CFC member to request that their rating be adjusted to match their FIDE/USCF/FQE rating? There would be, I imagine a specified differential (200 points?), a need to establish that the other rating is more "current" and a fee ($25?). The process would have to be worked out but before I go any further on this I'd like to hear if there is general support for the concept before I consider whether to table a motion at a future meeting.

--------------------

Part 2

The current rule is that if you’re under 1400 and away for ? (I’ve forgotten the time period) you get to start over. I believe it’s 5 years – I’d support reducing that period. I give again the example of the junior who stopped playing in tournaments, played 2000+ games on ICC over 2-3 years then came back at Expert strength but with a 1400 CFC rating...and promptly killed the local rating pool. Mathematically the Elo system works better when you have a single integrated pool rather than a horde of little pools. Another case was some 20-25 years ago when Manitoba was fairly deflated and a ratings boon was considered. This idea was discarded after a Canadian Open was held there and Manitoba players of all strengths dined on the competition!

I might be persuaded to say yes to FIDE but would say no to USCF and FQE and would resign all my chess positions before we even discussed C&M / AEM since I believe Larry’s ratings are more based on vanity than sound mathematics.

I would be sympathetic to a US national moving to Canada but would not approve of Canadian players amping their ratings by play in foreign events – and for chess purposes both are ‘foreign’ systems both in terms of their rating pool and the fine points of their mathematics.

The example of how the CFC handled the re-affiliation of the BCCF back into the CFC in the early 1970s including integration of Northwest Ratings into the CFC system comes to mind. I know it’s 40 years ago but with 20-20 hindsight I do believe the CFC did the right thing at the time. Something similar would be needed if Quebec ever chose to FULLY re-affiliate and I would support that.

Part 1.

I would suggest the following parameters:
Maximum of once per year (calendar year or interval of one year)
Cost of $50
No minimum difference between ratings

Part 2.
Yes the current rule is 5 years. The rule itself is ancient. I think I've applied it maybe 4-5 times in the 30+ years I've run tournaments.
I would suggest the Player / TD / Organizer would agree to the rating resetting, if their new rating is higher than their old rating (or even optionally if it isn't (?)) and send a note along with the tournament, as most players wouldn't even be aware they can do this.
As an opening thought, I would suggest U1600 - 4 or 5 years; U1200 - 2 or 3 years.

In case anyone is unclear on "Part 1" vs "Part 2", Part 1 is Leblanc's original proposal, Part 2 is his discussion in follow-up e-mail.

In general, for pairing purposes the Tournament Director is allowed to use any input on ratings he/she deems fit. I was TD of Yan Teplitsky's first Canadian tournament and as TD was presented with a Chess Informant and a friend of Yan's saying 'he has a FIDE rating' then turned to the lists and was told 'here it is'. He was listed as "J Teplicki 2420" and I have to say that an hour later I was deeply suspicious as he was an exchange down vs a B player - but he came back to win and won his other 4 games in the event as well. After 3 events he was 14.5-0.5 and only had a 2100-level performance rating presumably based primarily on his average rating of opponents. Had he been playing in a pool of 2000-2200 rated he likely would have had similar results and reached his present CFC rating much further.

What Paul says about single rating pools vs multiple smaller rating pools is quite true though not something I expect to see ever fully resolved in countries the size of Canada or the USA. It used to be the case that CFC events were the ONLY place in Canada to play competitive chess. That's no longer the case. As such rating distortions of the sort he cites are pretty much inevitable and ratings are at best an estimate of a player's strength and one never knows when you sit down to the board whether your opponent is playing well above or below his rating on any given day. Mathematically that's inevitable and is at the core of the entire rating system.

Just to clarify: only Part 1 of the first posting are my comments.
I have since then been approached by another TD asking for this to be considered.

I don't mind seeing some exploration of the concept, but I would have to be presented with some hard evidence of the following:
1) This is a significant issue, affecting >1% of CFC players.
2) Some definite framework describing how it would be applied.
3) Some examples of how ratings would be affected if any rule changes were applied, as opposed to the current rules, and how the results would be more fair.

Every time we tinker with the rating system, there is the possibility of it blowing up in our face. Furthermore, the CFC, USCF, FQE and FIDE all have their own rating systems. They are not identical. As long as there are different rating systems, or isolated pools within the rating system, there will always be cries of inequality. Given this we must approach this with care. There have not been many complaints recently about the rating system. I recommend not to fix something that isn't broke.

Quote:

---

Originally Posted by Garland Best

I recommend not to fix something that isn't broke.

---

Out of my head, there is a rule XYZ allowing players to rate their games played in international events.

Quote:

---

Originally Posted by Egidijus Zeromskis

Out of my head, there is a rule XYZ allowing players to rate their games played in international events.

---

It's from 1978

A Canadian may have a foreign FIDE-rated tournament count for their CFC rating by sending a letter to that effect in advance to the CFC. (And include $25 in advance with the letter sent to the CFC office. The letter should be received by the CFC office not less than 1 month before the start of the event, and include the $25 flat rating fee. - CFC Office policy effective December 1st, 2003) The CFC will rate the event only to the extent that FIDE does. Results must be reported within one month after the tournament is completed. Failure to submit a crosstable after registering an event bars the player from registering a tournament participation for three years.

In other words dating from when FIDE ratings were in their infancy.

Some good points from Garland.
1% of membership would be about 20 players. I doubt that we would get 20 applications unless we made the service free.
I was hoping to get some feedback on the framework aspect. I postulated a differential of 200 points as a starting point for the discussion.
From a rating point of view I think it would be a good thing. The typical applicant would be a junior player who's CFC rating is lagging his/her actual playing strength because they played in lots of USCF events. I have learned during my time that the rating system has a tremendous ability to resist inflation due to the huge number of juniors that start low, go up like rockets then quit chess after grade 12, taking all those rating points with them.
From a customer satisfaction point of view I also think it would show that the CFC is responsive to its members' reasonable requests.
I would be willing to donate my time to vet applications.
If I get a warm feeling that this initiative has a chance of succeeding, I'll come to the next quarterly meeting with a motion.

Quote:

---

Originally Posted by Lyle Craver

before we even discussed C&M / AEM since I believe Larry’s ratings are more based on vanity than sound mathematics.

---

The formulas look very similar for <2200 players. CMA is just one line, CFC in tabulated.