https://dacss.github.io/DACSS_601_Summer2023_Sec1/posts/HW3_Prasann_Desai.html

[giscus[bot]]

https://dacss.github.io/DACSS_601_Summer2023_Sec1/posts/HW3_Prasann_Desai.html

Reading in Data

https://dacss.github.io/DACSS_601_Summer2023_Sec1/posts/HW3_Prasann_Desai.html

[sp-conway]

Hi Prasann,

Nice job. You have done an excellent job of analyzing your data and making visualizations. I took notes as I read, and pasted them below. My biggest comment is to explain things a bit more to the reader. I don't watch much soccer, so there were a number of measures that I was confused about. I would assume little to no outside knowledge from the reader, and explicitly explain every measure you calculate and/or plot.

• Nice dataset.
• So to be clear, this is from an R package, so calling the fb_season_team_stats() function is what grabs the data?
• Can you explain what expected goals are? Is this because the season is not finished?
• Can’t you identify team rankings based on record?
• Comparison of average shots taken per 90 minutes
  • Can you clarify what you mean by for/against a particular team?
• Comparison of average shots on target per 90 minutes
  • Explain the difference between shots and shots on target
• Comparison of average shots on target per 90 minutes (% accuracy)
  • Can you explain what this percentage is?
• I think I’m still a little unclear about what expected goals are.
• Visualizing the linear relation between xG and goals scored
  • Nice plot.
• Volume of good quality chances created vs. league rank
  • Nice plot, but you should explain to the reader what “Normalized npxG to a scale of 0-20” is, how it is calculated, and why this measure is being used.
  • Did you repeat this plot?
  • It looks like the correlation is negative, not positive
• Unlucky
  • Good, but I would define exactly what you mean by “unlucky”
  • What is the measure of over performance? How is it calculated?
  • Please make the text larger for the team names plotted inside here.