The Effect of Age on Chess Improvement

  • Post author:
  • Post category:General
  • Reading time:2 min(s) read

Two months ago I wrote a post titled Descriptive Data which described the chess goals survey statistics. One of the points mentioned in that post was that younger players tend to improve more quickly. It seems discouraging for adults who are looking to quickly make gains.

I created a boxplot of the annual rating gains in each age category among survey takers. Here’s my two sentence summary on what a boxplot is:

  1. The middle horizontal line in each box is the median of the group
  2. The top and bottom of the box represent the top and bottom quartiles of the data

Average Rating Gain By Age

chess rating gain by age
Annual Rating Gain by Age Category

The data is sparse in the <10 and 60+ age groups, so let’s focus on the 10-59 age range. There is a downward trend that shows, on average, the older players improve less quickly than the younger players.

But…. there is more to it than that! An even stronger predictor of yearly rating gain is initial rating. Here’s a boxplot showing the initial ratings of each of the age categories.

Initial Rating by Age

initial chess rating by age
Initial Rating by Age Category

Again, ignoring the <10 and 60+ categories, there appears to be a very slight upward trend in start rating as age increases. To those of you who read the descriptive data post, I know what you’re thinking. “Matt, don’t we need to know how many hours these players spend on chess?”

Hours of Chess Study

average time spent on chess by age
Hours of Chess Study by Age Category

Here you go! There appears to be a U-shaped effect on hours spent by age category. I’ve noticed this at over-the-board tournaments as well. There is a lack of players showing up in the 20-40 age range. Chessgoals data confirms that these age groups are spending less time on chess.

Key Takeaway

When we combine age, initial rating, and time spent on chess in a linear regression model we can come up with an estimate on how age effects chess improvement. The data is showing that after adjusting for initial rating and time spent (both with 2nd-order terms), there is a linear effect of age on annual rating gain of approximately -3 points for every year one ages. Not terrible, but there is an uphill battle to learning chess as one ages. Stay tuned for a future post on specifically how adults can study for improvement.

This Post Has 6 Comments

  1. Dennis Mays

    Thanks Matt for the article on the impact of age on chess improvement. The U-shaped curve of time spent on chess with age is interesting. I’m looking forward to the future installment on how an older player can improve.

  2. Matt Jensen

    Thanks Dennis! Will be coming soon. We have a long list of post ideas 😉

  3. Matt Jensen

    Yes I think so too Laura!

  4. John Lee

    What time controls are the rating gains based on? I think that makes a huge difference. I assume that the older you get, the harder it is to improve on shorter time controls. Is that true?

  5. Matt Jensen

    We asked the survey takers which rating they consider the best measure for your own chess improvement. The next question asks what their rating was in that category 12 months ago. I should provide a breakdown of those results, but I believe it was mostly OTB ratings or a mix of blitz/classical online.

    Good question on the time control vs age. I’ll think about the best way to look at that. Maybe going forward track changes?

Comments are closed.