Men’s World Championship Figure Skating Scores

from 2005-2024

Jasmine Sun

The Data

This data set contains scores from the Men’s World Figure Skating Championships from 2005 to 2024 by skater. It was created by Mitchell Friess, an American figure skater and data scientist, using score data from skatingscores.com.

# A tibble: 6 × 12
   year skater     sp_total sp_tes sp_pc sp_quads fs_total fs_tes fs_pc fs_quads
  <dbl> <chr>         <dbl>  <dbl> <dbl>    <dbl>    <dbl>  <dbl> <dbl>    <dbl>
1  2005 Lambiel        80.3   42.6  37.7        1     144.   67.3  76.9        2
2  2005 Buttle         77.4   40.5  36.9        0     136.   63.4  74.9        0
3  2005 Sandhu         67.3   33.7  33.6        1     134.   62.2  72.9        1
4  2005 Lysacek        73.4   38.7  34.7        0     134.   63.2  70.6        0
5  2005 Van Der P…     66.5   34.7  31.8        0     134.   68.7  64.8        0
6  2005 Weir           70.5   35.9  35.6        0     133.   60.7  72.6        0
# ℹ 2 more variables: total <dbl>, total_quads <dbl>

The original data set did not include total scores (sp_total + fs_total = total) or total quads (sp_quads + fs_quads = total_quads), so I have added those two variables.

The data begins in 2005, as that is when the International Skating Union implemented the current scoring system. In 2018, they modified the way that Grade of Execution (GOE: how well the jump is performed) is calculated.

The data set does not contain data for the year 2020, as the World Championships were cancelled due to the COVID-19 pandemic.

Background Information

Figure skating is a sport that combines athletic prowess with artistic expression, and examining scoring trends over time can reveal how the sport has evolved in terms of technical difficulty and artistic presentation.

The data set shows the Short Program Technical Element Score (technical difficulty), Program Component Score (artistic presentation), Total Score, and Number of Quads and the Free Skate Technical Element Score, Program Component Score, Total Score, and Number of Quads.

Quadrupal-rotation (quad) jumps are worth the most points. In base value alone, which does not include GOE, they range from 9.5 to 12.5 points, while triple-rotation jumps range from just 4.2 to 8 points.

Why this data set

I chose this data set because as a former competitive figure skater, I have an interest in analyzing the trends and patterns in competitive figure skating.

Additionally, The 2025 World Championships will be held during Spring Break in Boston, so I am excited to see how the upcoming scores will align with past trends.

Evolution of Average Scores in Men’s Figure Skating

Density of Quad Jumps Attempted (Even Years Only)

Relationship Between Number of Quads and Total Score

Core Concepts and Insights (scrollable):

  1. Evolution of Average Scores in Men’s Figure Skating:
    1. The line graphs demonstrate that there is a clear upward trend and substantial improvement in scores.
    2. The line graphs demonstrate steady upward growth in scores, except for a dip in 2018, which could be due to the change in GOE calculations. This could be further researched upon by finding data regarding GOE scores.
  2. Density of Quad Jumps Attempted:
    1. The density curves show that there’s a clear shift in the distribution of quad jumps attempted over time, moving from left to right on the x-axis.
    2. In earlier years (lighter colors), the density was concentrated around 0-1 quads. In recent years (darker colors), the density has shifted towards 3-5 quads, indicating a significant increase in technical difficulty.
    3. The range of quad jumps attempted has widened over time, suggesting greater variability in skaters’ technical abilities. In earlier years, it was unfathomable to compete more than 3 quads, but more recently, the range of number of quads goes from 0 to 8.
  3. Relationship Between Number of Quads and Total Score:
    1. The jittered scatterplots and fit line demonstrate a clear positive relationship between the number of quads attempted and the total score achieved.
    2. The regression equation (y = 13.44x + 198.79) suggests that each additional quad jump is associated with an average increase of 13.44 points in the total score. This is higher than the base values for quad jumps (9.5 to 12.5) points because when a jump is performed well, it also increases GOE and the Program Component Score. This demonstrates that increasing the technical difficulty of a program also increases its artistic presentation.
    3. Despite the strong correlation, there’s considerable scatter around the regression line, indicating that factors other than quad jumps also significantly influence the total score, the most important being artistic presentation.