I don't see the advantage if the have voting after Olympiad. Recently, I got a supporting e-mail from a National Team player. I believe, majority of his teammates also support this proposal.
I want to explain more about using the last rating (FIDE or average) and not the last year average or maximum rating.
I guess, maximum rating (we use it now) is the worst possible solution, while any other (last or 12-month average) is much better. The difference between the last and the average is not that big; if governors like average, we can change it. However, after spending some time on this issue, I believe that using the "last" one is the best solution.
1. 2 possible scenarios there last rating is much better than the average.
a). Player was inactive for very long time. 2 months before the deadline he starts playing, plays 20 games and thus meets the minimum game requirement. He performs much lower than his initial rating and loses 20 points a month. Looks very possible. Let's say his rating was 2500. His "last" would be 2460. His average would be (2500*10 + 2480 + 2460)/12= 2495. Sure, "last" is more useful in this example.
b). Player gains a lot of points for very short period of time. In the past E.Hansen, A.Hambleton and others gained 100 points or more in less than 12 months. Eric gained 123 points (from May 2012 to May 2013; from 2454 to 2577). Aman gained 242 points (from May 2012 to May 2013; from 2226 to 2468). Which rating reflects better the real strength of a player? Sure, the "last".
2. Indeed, the "last" sometimes is too volatile, but with k-factor of 10, the volatility is low. In many cases, it's not just random fluctuation, but a part of long-term trend.
3. I guess, many people highly overestimate the scenario, there a certain player stops playing. Olympiad is definitely an interesting and important tournament, but the life must go on. I do not see how professional chess player would miss many tournaments because of Olympiad. Also player, who does not play last few months, may lose his spot as a result of the progress of other players.
I think the final rating at the deadline is fine. As Victor says players who are afraid of losing rating points should also be afraid of players catching them.
In general we have better participation at an AGM rather than a quarterly meeting. I also do not see any advantage to a delay. Do we like the system as it is currently arranged? Yes or no. If we like the system then there should be no change. If we don't like the current system here are two possible ways to change it. The problem with the current system is that it introduces a very subjective element into the Olympiad selection process. Having to find people willing to serve on the selection committee can be challenging at times. When the decision comes out there will be calls for the directors to intervene if the decision does not go the way some would have expected. At some point we will have a president and board of directors which might be inclined to intervene. Problems will ensue. I see these two proposals as a good way to head off that scenario.
We are not changing a bylaw with this change if we make a change. There is no place in our bylaws where such a bylaw would fit. We are changing our policy for choosing the olympiad teams. One of the things that we need to accomplish in the next few months is to codify a new set of policies to replace the handbook which only currently holds force because the directors have accepted the idea that we should hold to the handbook until we can rebuild an acceptable alternative way of codifying the policies that we have put in place over many years. This will likely be contained in a series of web pages. Peter McKillop and others are working on getting the handbook up to date. Once that happens we can then go through it and remove the things which no longer make sense or are inconsistent with the NFP act.
I agree with this. There is no necessity to rush a decision. Better take some time to arrive at the best formula.
Let's see what would have happened, had we used this system. I checked last 4 campaigns.
2010. Champion: Hebert (declined). Spraggett declined 2010, 2012, 2014. The Team: Bluvshtein - Roussel-Roozmon - Gerzhoy - Noritsyn - Samsonkin.
Bluvshtein: 2583 + 20 (2nd place) - 1 (previous Olympiad) + 10 (age) = 2612.
Roussel-Roozmon: 2488 - 10 + 3 + 10 = 2491.
Gerzhoy: 2469 + 5 = 2474.
Noritsyn: 2403 + 9 + 25 = 2437.
Hansen: 2415 - 10 + 30 = 2435.
Porper: 2431.
Panjwani: 2401 + 20 = 2421.
Samsonkin: 2406 + 15 = 2421.
So we have Hansen ahead of Samsonkin, while Porper and Panjwani (tie-break!) also above Samsonkin. According to proposed system, the Team would have been: Bluvshtein - Roussel-Roozmon - Gerzhoy - Noritsyn - Hansen.
2012. Champion: Sambuev. Bluvshtein and Spraggett declined. The Team: Sambuev - Gerzhoy - Noritsyn - Hansen - Porper.
Gershoy: 2489 + 10 (3rd place) - 1 = 2498.
Hansen: 2454 + 20 (2nd place) + 20 (age) = 2494.
Noritsyn: 2475 + 3 + 15 = 2493.
Porper: 2438 - 10 = 2428.
Panjwani: 2416 + 10 = 2426.
Porper still has 2 points advantage over Panjwani. In this case we have exactly the same Team.
2014. Champion: Sambuev. Spraggett declined. The Team: Kovalyov - Hansen - Sambuev - Gerzhoy - Hambleton.
Kovalyov: 2636 + 20 (2nd place) + 10 (age) = 2666.
Hansen: 2587 - 10 + 25 (last Olympiad) + 10 = 2612.
Gerzhoy: 2468 + 8 = 2476.
Noritsyn: 2449 + 10 (3rd place) + 9 + 5 = 2473.
Hambleton: 2455 - 10 + 10 = 2455.
We have Noritsyn above Hambleton. The Team would have been: Kovalyov - Hansen - Sambuev - Gerzhoy - Noritsyn.
2016. Champion: Krnan. The Team: Bareev - Kovalyov - Hansen - LeSiege - Krnan.
Bareev: 2675 - 10 = 2665.
Kovalyov: 2608 -10 + 9 = 2607.
Hansen: 2580 + 20 (2nd place) = 2600.
Sambuev: 2544 + 6 = 2550.
Preotu: 2462 + 35 = 2497.
LeSiege: 2497 - 10 = 2487.
Noritsyn: 2476.
We have both Sambuev and Preotu above LeSiege. The Team would have been: Bareev - Kovalyov - Hansen - Sambuev - Krnan.
Last edited by Victor Plotkin; 08-24-2016 at 10:01 AM.
This is good stuff Victor, thanks for your efforts, and it implies that we are closing in on a complete solution. I am pleasantly surprised at most of the effects so far. My only cognitive struggle is how to numerically give more "credit" for obviously shooting-up junior stars who are within acceptable "striking distance" of the 5th person on the team.
Aris, but the juniors are already given preference, with the extra points they get for being younger than 23. The number could be changed from 5 to 6 or 7 maybe, but I personally like it the way it is suggested.
I see no point in waiting for the olympiad to finish to vote on this. There is enough data already. Both "last rating" and "average rating" are much more fair and work better than the current "top rating", even with the mentioned downsides.
One possible issue of waiting until after the Olympiad is that the Olympic + and - points will be known and could affect people's voting - but perhaps no more so than people's bias towards giving credit for young players
We could vote now on
1. Change to rating system for Open (CFC/FIDE to FIDE)
2. Removal of selection committee (4 players by selection rating)
3. Change from highest rating
3a. Final rating
3b. Average of 12 monthly ratings.
and leave the three components of the Victor's adjustment to continued discussion, although generally people seem positive.
As for the age bonus, I could see an even more radical shift. 5 points for every year 20-24, 10 points for every year under 20. In other words that would be 75 points for a 15 year old. I think this might be even more pertinent on the womens side, where we have seen cases of ladies with realative inactivity getting in ahead of active youth.