Suggestion for FIDE: K-factor ‘holiday’

by Edwin Lam
1/9/2023 – The Covid pandemic in the last two years have led to a chess boom online, and the young guns have benefited the most from having ‘extra time’ on hand due to being at home a lot throughout the years 2020 and 2021. They seem to have done well to widen their learnings of the game, making large gains on the rating scale. Edwin Lam Choong Wai, chess enthusiast and coach, has a suggestion how FIDE should handle this. We would be interested to hear you opinion on the subject.

ChessBase 17 - Mega package ChessBase 17 - Mega package

ChessBase is a personal, stand-alone chess database that has become the standard throughout the world. Everyone uses ChessBase, from the World Champion to the amateur next door. It is the program of choice for anyone who loves the game and wants to know more about it.

More...

With increased focus of staying at home and going online, many young players in Asia went online to learn chess from GMs and IMs around the world! They also took part in various online simuls globally! Parents are even willing to pay up to 100 pounds Sterling per session of online coaching with top GMs from Europe!  

14 year old FM Poh Yu Tian from Malaysia, at the prize-giving ceremony of the Penang Open.

Poh Yu Tian famously defeated GM Ftacnik at the 2022 Olympiad besides also beating GM Jan Gustafsson at the 2022 Bangkok Open.

These kids and teens – collectively, young woodpushers – have truly gained a huge leap of strength and this was evident in the final rankings of this Penang classical chess event – with outstanding performances by FM Poh Yu Tian, CM Xie Kaifan, Dinh Nho Kiet, CM Anderson Ang Ern Jian, Yeoh Yuan Hui, Darshan Revinthiran, Ng Sheng Feng, Nguyen Vuong Tung Lam, Uzair Shahar, Ernest Yek Zu Yang, Lin Jie Xun, Kavin Mohan and Genivan Genkeswaran amongst others! 

Under-rated Young Guns!

Many of these young guns are severely under-rated – their playing strengths have increased considerably during the years when OTB tournaments were not allowed – and they added huge amounts of rating points by drawing or beating higher rated opponents at the tournament! 

Qualitatively-speaking, Darshan Revinthran, Uzair Shahar and Ernest Yek added over 100 rating points each, with Ernest gaining the most at 151 points! CM Xie added 89 points while Dinh added 80 points! Nguyen and Lin Jie Xun added over 60 points each! And, Genivan also added 53 points.

Zooming in using quantitative analysis of the full results spectrum of the Open category at the Penang event, we built the following table to better strengthen our argument here:

Round Number (Open category) Total Games played in that round Total Wins by Lower-rated Players Total Number of Draws
1

40

1

5

2

40

7 7

3

40

5

10

4

40

4

9

5

39

5

8

6

39

3

15

7

39

8

11

8

39

11

12

9

39

12

7

Out of a total of 355 games played across the 9 rounds, 56 of them were won by the weaker player, while another 84 of them saw the stronger player only drawing against their lesser counterpart! What this means is that on average, roughly 15 games (out of ~40 games per round) in every round saw the lower-rated player (with the exception of 4 unrated players in that field) gaining some amount of rating points off their more illustrious opponents!

Different Tournament, Same Story

This phenomenon is not unique to Penang Open alone, but have been seen in recent months from other events, as well! To illustrate this point, we built a similar table as above from another recently-concluded event: the 7th Johor International Chess Open 2022. 

In the table below, from the 349 games played over 9 rounds, 50 of them were won by the weaker player! Another 101 were drawn by the lower-rated exponent. On average, 16 games per round saw the lower-rated player (with the exception of 6 unrated players) gaining rating points from their higher-rated opponents.

Round Number (Open category)

Total Games played in that round

Total Wins by Lower-rated Players

Total Number of Draws

1

39

6

10

2

39

3

13

3

39

3

8

4

39

4

13

5

39

5

8

6

39

8

13

7

39

7

10

8

38

6

16

9

38

8

10

By the end of the event in Johor, Vietnamese talent, Huynh Le Minh Hoang, added over 140 points. Fellow Vietnamese, Nguyen Manh Duc, added over 120 rating points while his compatriot, Dinh Nho Kiet, gained over 80 rating points. Teens, Genivan Genkeswaran and Yee Hao Loong of Malaysia, shone in the event with the additions of around 100 points each. Not to be outdone is Filipino, Ivan Travis Cu, and his addition of close to 80 points and Marcus Chen of Singapore with his increase of roughly 50 points. 

GM-slayer Genivan

While these young woodpushers gained in strength, the more matured players dropped points as their strength in the game decreased. Thus, a young gun’s 1500 rating is deceiving, as is an older player’s 2300 rating might probably not be reflective of his actual strength. Interestingly, the then 1576-rated Genivan Genkeswaran was probably under-estimated by some of his titled opponents in Johor! During that event, he scalped GM Nguyen Van Huy while drawing GM Alexei Barsov and IM Pavel Shkapenko. The games whereby he succeeded in slaying or drawing these titled players are given in the games section below in this article. 

 

 

Genivan Genkeswaran is the boy on the left-most in this photo, with his thumb up, in this picture taken of the entire Selangor state team at the 2022 Malaysian Schools Chess Championship (MSSM) in Kota Kinabalu. Photo shared by David Yek

A Proposal for FIDE

With such a huge discrepancy in ratings of players and their actual strengths, I would like to suggest to FIDE to implement a kind of K-factor ‘holiday’ over a period of 1 year from now with a larger factor applicable for all rated players throughout this period! As a recap, the K-factor is the development coefficient used to calculate the ratings of the players worldwide.

The current K-factor rules by FIDE are as follow:

    • K = 40 for a player new to the rating list until he has completed events with at least 30 games
    • K = 20 as long as a player's rating remains under 2400
    • K = 10 once a player's published rating has reached 2400 and remains at that level subsequently, even if the rating drops below 2400
    • K = 40 for all players until their 18th birthday, as long as their rating remains under 2300
    • K = 20 for RAPID and BLITZ ratings all players

The suggested K-factor ‘holiday’* I am proposing here is as follows:

    • K = 40 as long as a player's rating remains under 2400
    • K = 20 once a player's published rating has reached 2400 and remains at that level subsequently, even if the rating drops below 2400
    • K = 50 for all players until their 18th birthday, as long as their rating remains under 2300

*This proposal does not cover the Rapid and Blitz ratings due to the recent updated ratings’ regulations announced by FIDE on 1st Oct 2022. 

This proposal will enable under-rated players to much more quickly gain points during this period and rise in their rankings soonest, while fixing the problem with inflated or over-rated players! Additionally, this will also spur greater participation of chess players back to OTB play as the chances for a rise in rating is high and this will be a welcome for classical chess tournament organisers worldwide who are looking to jumpstart their activities again after years of hiatus! 

Perhaps this is something that Arkady Dvorkovich and his team at FIDE might want to consider for the year 2023? I welcome comments to this article and debate, besides any effort by anyone worldwide to start an online petition to FIDE to look into this option! 


Edwin Lam Choong Wai is a Malaysian chess player and author. He was previously attached to Procter & Gamble doing local, regional and global marketing roles, before joining Pfizer, Essilor and Yeo’s in both Malaysia and Singapore. He had also previously been attached to The Purpose Group, a creative and digital marketing agency in Ho Chi Minh City. He is now based in Malaysia having started an education venture known as My SKOLA+ (http://myskolaplus.com) since end-2017.
Discussion and Feedback Submit your feedback to the editors


Discuss

Rules for reader comments

 
 

Not registered yet? Register

edwinlamchoongwai edwinlamchoongwai 1/12/2023 12:34
Thank you all for the comments! There are some good, constructive suggestions below, which can be added to modify the original proposal above, which is only a temporary measure to correct the under-rated and over-rated present post the pandemic gap! I plan to write a follow-up article on this topic, with some additional data from recent chess events to further present this case! Stay tuned!
adbennet adbennet 1/10/2023 11:20
All ideas of "fine-tuning" the rating system are completely unnecessary. Over-rated and under-rated automatically corrects itself -- just play. When I play a youngster and lose, my Elo goes down, theirs goes up. Simples.
Michael Jones Michael Jones 1/10/2023 02:26
Any player, regardless of their rating or that of their opponent, may take a quick draw if doing so would benefit their position in the tournament. The fact that a 2400 and a 2300 did so against a likely under-rated junior says very little. Neither does 40% of games resulting in the lower-rated player getting at least a draw; particularly in the later rounds when the Swiss system ensures that players will be more evenly matched - it's likely that in many cases the rating difference was quite small so the higher-rated player failing to win is not much of a surprise. There doesn't seem to be a major problem here; if any player initially has a rating which is not reflective of their true strength, it will gradually become so as they play more games.
Aighearach Aighearach 1/9/2023 06:43
I'm not sure that a 20 move agreed draw after a sloppy opening makes the case.

It seems to just be the typical, age-old case of an older player making a couple dumb aggressive moves in response to mistakes by their younger opponent, and here, it merely ended in an equal position. Given the quality of the opening, if they had played it out the game would have likely had a decisive result.

I think a simple solution is that if your opponent has less than some minimum number of total games, or less than another number of games in some specified period of time, then their rating change would be reduced. We already have things like rating floors, ratings are not zero-sum.
littlefish littlefish 1/9/2023 05:44
Wouldn't this suggestion actually make things worse? At K=40 for all players under 2400 they'd lose twice as many points against underrated juniors. An idea to prevent this could be: Count games against players under 18 and under 2300 Elo only with K=10 for their opponents - at least until the end of 2023, when the pandemic effects should be mostly over.
JNorri JNorri 1/9/2023 05:26
ELO ratings are not a tool of reward and punishment. At least they should not be, to keep the mathematical integrity as intact as possible (it never is completely).
calvinamari calvinamari 1/9/2023 02:13
Get over it. This ain’t exactly the rain of fire and end of days. Just play more games.
brabo_hf brabo_hf 1/9/2023 02:09
The suggestion of K = 40 as long as a player's rating remains under 2400 is riduculous. I played 60 rated games in a year as 46 year old FM. With such K factor my rating would be fluctuating with hundreds of points each month. That is nonsense.

I also don't get it why you are talking about over-rated players or more mature players decreasing in strength. This makes no sense. Maybe if you are very old (+80) that somebody loses strength quickly but otherwise you keep your level for many years. I definitely don't feel any weaker than 20 years ago.

I do understand the problem but your proposal is pure nonsense. There is one very easy fix for the problem and that is just to reset the ratings of young players. You consider them unrated and after x games you give them the rating of their performance.
KnightOnTheRim KnightOnTheRim 1/9/2023 12:23
I think this proposal doesn’t show how a seasoned 2400 can avoid huge losses of points against a youngster who is ( a lot ) underated .
It would be complicated but I think a re-calculation every 25 games ( for example ) using the last rating and recalculating vs every opponent could a be solution .
cross cross 1/9/2023 12:17
I think there should be some sort of 'youth' rule to protect against dramatic rating losses when being 'upset' by a young person who is actually far stronger than their rating. I lost against 13 year old who was rated 1710 (I was around 2000) and I lost a bunch of points, but the kid went undefeated through nine rounds and won clear first place. There should be some rule that if a youth has a performance above a certain amount (not sure, perhaps four hundred points above their rating), people who are upset by them only lose one rating point.
1