Descriptive Data

Chessgoals.com contains cleaned data from November 2018 – present on 384 chess players. This post will summarized the most important variables. The majority of the players were ages 10-50, but we still had quite a few players over age 50.

Frequencies of age group

The dependent variable for most of our analyses was annual rating gain. The median in our data set was 110 points, with the 25th percentile at 25 points and the 75th percentile 218 points. The median weekly hours spent was 7 hours, with 25/75th percentiles at 4 and 14 hours respectively.

Summary of Annual Rating Gain
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
 -340.0    25.0   110.0   144.2   218.0   640.0 
Summary of Weekly Hours Spent on Chess
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  0.000   4.000   7.000   9.572  14.000  42.000 

We will get into modeling in a future post, but there are three very strong predictors of annual rating gain:

  • Hours Spent (more hours tend to see more gains – good!)
  • Starting Rating (lower starting rating tends to gain more points annually)
  • Age (younger players tend to improve more quickly)

Here’s a look at the annual rating gain by hours spent, stratified by starting rating. There does appear to be a positive association between hours spent and rating gain up until about 25 hours spent. After that the data is a sparser, but it’s interesting to ponder if there is a burnout point above 25-30 hours of chess per week.

Annual Rating Gain vs. Chess Hours/Week, Stratified by Start Rating

Please let me know if there is any other data that’s particularly interesting in the comments.

This Post Has 3 Comments

  1. Dennis Mays

    It is interesting and perhaps expected that older players may have smaller rating gains compared to younger players. Old brain versus young brain. I would like to see more about the affect of age on chess improvement sometime. Many years ago Donald Trump asked GM Pal Benko what it would take for him (Trump) to become a chess grandmaster. GM Benko’s response was to Trump was “To be born again!”

  2. Dennis Mays

    Make that *effect

    1. Matt Jensen

      Thank you Dennis. I’ll work on adding a specific post on improving as players age. The next few learner series posts are specifically aimed at adults over 1700.

Comments are closed.