North Carolina Chess Association
Charles Roberson – President
John Timmel – First Vice President
Grant Oen – Scholastic Vice President
Wayne H. Spon – Secretary/Treasurer
North Carolina Gambit – August 2022
This month’s Contributors
Charles Roberson – Editor
Web support, edit, and distribution
In this issue, you will find:
Charles Roberson – Interview with Asha Kumar
Walter High – Coverage of the US Open
Charles Roberson – Elo Rating System: The Myth of Playing Up
John Timmel – To Move and Win!
We’re looking for contributors to the Gambit!
Submit your games, articles, etc. to the editor Charles Roberson at email@example.com
NC Chess Interview
Interview with Asha Kumar and her parents by Charles Roberson
Asha Kumar is a NC chess player in the tenth grade. This year she won the NC Girls Scholastic Chess Championships making her NC’s invitee to the National Haring Invitational.
Asha performed well finishing tied for 2nd. She tied for first in the 2022 All Girls National Championship (U16). NC Chess interviewed Asha and her parents for this month’s newsletter.
My parents are supportive of me and my sister playing in chess tournaments. Although tournaments can be a stressful environment, they are also really enjoyable for the whole family.
My parents and I have always enjoyed going to the US Open every year.
My favorite tournament is NC states. Not only is states an amazing opportunity to improve your chess, it is also a place to meet other members of the chess community.
It is really exciting winning girls states. I’ve been playing in this tournament for the past four years, and it feels really great to finally win
I’ve been playing chess seriously and attending this tournament for four years.
As always, chess is full of uncertainties so I was unsure I would achieve it. All I could do was stay concentrated, and it ended up working out.
I usually celebrate a victory by going out to dinner with my family. Chess tournaments are often long and stressful so I think it is really important to reward yourself for the hard work.
I think when you treat a loss as a learning experience it becomes much more bearable. For example, it is beneficial to prioritize learning and growth over things like rating.
I have won the U-16 section in the girls national championship.
I got into chess after attending a summer camp at Durham Academy taught by Craig Jones.
I have been playing chess for about 6 years.
I was 9 years old when I started playing chess.
The longest break I’ve taken from chess was for 6 months about a year ago. I started getting serious around the age of 12.
My first chess rating was 635.
My peak chess rating is my current rating, which is 1948.
My favorite thing about Chess is the community. I have met so many incredible people through chess.
I like that in chess, each player has a unique style. While it is a very analytical game, there is still lots of room for creativity.
I like being able to learn and be around great players in chess tournaments. There is so much to learn from others, and chess tournaments are a great place to do that.
While I enjoy competing in chess, ultimately it is an activity that I play to enjoy. As I move through high school and the future, I intend to continue playing chess at least recreationally as it is a huge part of my life.
Chess is so much less stressful when you don’t worry about rating.
My favorite player is Rameshbabu Praggnanandhaa.
I would like to meet Magnus Carlsen.
My favorite online chess site is lichess.com.
My favorite chess book author is Jacob Aagaard.
I would like to visit the Marshall chess club in New York City.
Elo Rating System
The Myth of Playing Up by Charles Roberson
This month I will discuss the Elo rating system, some history, the fundamentals of how it works and the myth of playing up.
Arpad Elo invented the system which was adopted by the USCF in 1960 and it was adopted by FIDE 10 years later in 1970. He used real tournament data from the USCF to develop his system. This would have been quite the effort considering that the hand held calculator hadn’t been invented and he was possibly limited to pencil, paper and a slide rule. It is still in use by the USCF and has been modified on several occasions. One includes a big change in 2015 which deals with the rapid rating increase of new people relative to established players which caused established players to lose lots of points thus deflating the upper middle group of the players.
The mathematical system is based on expectation. If you perform to expectation, your rating doesn’t change. Perform above or below expectation and your rating will increase or drop accordingly. The system doesn’t allow you to increase your rating just by competing in more tournaments than anybody else. You must perform above expectation to increase your rating. So what is expectation and how is it calculated? How can the rating system know how you should perform?
The system expects you to perform exactly to your rating. It expects people rated higher than you to perform higher than you and the opposite for people rated lower. If you play a group of people with an average rating equal to yours the system expects you to turn in a 50% performance. If you play 5 opponents with an average rating of 1500 and you are rated 1500 the system expects you to score 2.5 out of 5. If you score 3 or more out of 5 tournament points then your rating will go up. Score 2 or less points out of 5 and your rating drops. The further away from expectation you perform, the more your rating will change. Also, your rating will change more in bigger tournaments than smaller ones. The base system will give you more points for performing above expectation in a 5 game tournament than in a 3 gamer. On top of that, there are bonus points for tournaments with 3 or more games and and more bonus points for the same level of performance for even bigger tournaments.
The system is based on a standard deviation of 400 points. A person rated 200 points higher than his/her average opponent is expected to win 76% of the tournament points. A person rated 500 points higher than his/her average opponent is expected to win 95% of the tournament points. Conversely, a person rated 200 points below his average opponent is expected to win 24% of the points and 5% of the points if he is 500 rating points below his average opponent. So, you can be one of the lower rated players in the field and raise your rating by scoring only 1 win out of 5 or get the same rating improvement by 2 draws out of 5. If you want all the mathematical details you may read this document
Rating System Algorithm.
Now for the myth of playing up. Many think that if you want to raise your rating as fast as possible you should play up a section and the rating system will boost you more. This is not true. If you perform 200 points above your rating, then your rating will go up a given amount no matter how you did it. Let’s say you are rated 1600. There are lots of ways to perform at the 1800 level. Just to identify a few:
- You play a field of 1400 rated players and win every game.
- You play a field of 1600 rated players and score 76% of the games.
- You play a field of 1800 rated players and score 50% of the games – in 5 games you get 5 draws or 2 wins and a draw.
- You play a field of 2000 rated players and score 24% of the games – in 8 games you score 4 draws or 2 wins or 1 win and 2 draws.
If you have used the USCF ratings estimator to get a quick estimate of your post tournament rating change ( at USCF rating estimator ), you’ll see that it doesn’t ask how you performed against each opponent; it asks each opponent’s rating and how many total tournament points you scored. So, it wants to know your average opponent, how many games and your overall score. It doesn’t care which opponent you won, drew or lost to. Lets look at some real numbers.
The above table came from entering data into the USCF rating estimator and we see multiple perfect scenarios for 2, 3, 4 and 5 game tournaments. I say perfect scenarios because I am assuming that you’ll play against exactly the correct opponent’s to achieve the intended rating gap, but more on this later. Before we get into the potential of playing up, lets observe a general trend that holds true: you get more points for performing at a given level by playing in tournaments that have more rounds if you play all the rounds. Notice, that a 1600 player performing at 1800 and playing an average field of 1400 gains 20 points in a 2 game tournament vs 52 points in a 5 game tournament against the same strength opponents.
We see that in 3 scenarios for 2 game tournaments playing up doesn’t make any real difference, all scenarios resulted in a new rating of 1620 except for the last scenario which resulted in 1621 (only 1 extra point). In a 3 game tournament, we see that playing up gains only 3 extra points. In a 4 game tournament, playing up yields an extra 6 points: a gain of 48 instead of 42. Finally, in a 5 game tournament, playing up yields an extra 7 points: a gain of 59 instead of 52. The extra yield is small and only if you can perform that well, if not playing up will not help.
So, there is nothing magical about playing up or any real advantage in the math: you have to perform well. The issue is that when you play higher rated players, you might not perform so well and thus lose rating points. In such cases, playing up is self defeating. Also, the probability of you scoring against each opponent is reduced, thus decreasing your odds of gaining rating points. To make things worse, you are making the tournament worse for those that have to play you: in effect, they are playing down when playing you. If you play too far up, you can hurt the integrity of the tournament. In such cases, everybody that plays effectively has a bye. If enough people play up, the tournament integrity can be sufficiently damaged.
On July 23, I ran a tournament in Burlington. There were two kids in the middle section that wanted to play up. I moved one of them up. The kid that didn’t play up won all of his games while the kid that played up score 1.5 of 3. The kid that didn’t play up gained 25 rating points and the other gained 11. So, the “magic” of playing up didn’t help. You might say that it gives you more experience playing stronger players which will help you learn faster. Well, what about the stronger player that had to play you the weaker player? Wasn’t that a diminished experience for him? Think of it this way: for each person playing up in a 5 game tournament, there are 5 people that have to play down for one of their games. This suggests there are more negatives than positives.
Additionally, playing up has some potential negatives for the person playing up. You will be the lowest rated player or one of the lowest rated players if several play up which can lead to several issues. All of these issues will partially defeat the purpose of playing up.
- If there is an odd number of people in your section and you are the lowest rated player, you will get a bye the first round.
- If multiple people play up and the section has an odd number of people, you will likely get a bye in one of the rounds if not the first.
- If multiple people play up and you perform similarly, you’ll end up playing each other in the late rounds.
- You could end up playing somebody else that played up and get a bye in a round.
Here are two of the tournament pairing rules that can cause these events. The first is the lowest rated player with the worst score in each round gets the bye. If you play up, you could be the lowest rated player. The second is that you pair people together that have the same score or nearly the same score in each round. After a few rounds, you may be paired with others that played up. So, there is zero guarantee that playing up will go the way you want nor is there any guarantee that you’ll get to play multiple much higher rated opponents.
For these reasons, there will be a new rule in the chess tournaments that I run: no playing up. You will be in the section that your rating puts you in.
To Move and Win!
By John Timmel
Last Month’s Spassky – Fischer
White to move and win:
Black to Move and win: hxg2!
White to move and win: Kd6
Black to move and win: Rxe3!