Click the Original, Code and Reconstruction tabs to read about the issues and how they were fixed.

Original


Source: AFL Record (2020).


Objective

The objective of this data visualisation is to communicate how the scoring accuracy of AFL teams has changed over time.

Audience

The audience for this visualisation is readers of the AFL Record. These readers would be AFL fans of varying demographic backgrounds, who should not necesarilly be expected to have any particular statistical knowledge.

Issues

The visualisation chosen had the following three main issues:

Limited use of data. The visualisation only depicts scoring accuracy from seven years - 1956, 1966, 1976, 1986, 1996, 2006, 2016. This is despite the fact that this data is available for every year from 1897 to the present. The visualisation derives no benefit from what appears to be an arbitrary decision to omit the vast majority of relevant data.

Choice of graph type. The visualisation is essentially a bar graph which has repaced bars with triangles. A bar graph is not a great choice for time-series data to begin with, and the decision to use triangles instead makes it more visually difficult to judge the relative size of objects. There’s no apparent rationale for this design decision.

Poor use of colour. The visualisation splits the triangles into two different shades of blue-gray. It would be natural to assume this means the data belongs to two different categories - but, instead, this appears to be a purely aesthetic decision. It unnecessarilly obfuscates what should be a simple and straightforward visualisation.

Reference

AFL Record: Week 3, page 20. (2020). “AFL goalkicking accuracy over time”. Retrieved May 10, 2020, from issuu.com: https://issuu.com/lifestyle1-media/docs/afl_record_-_week_3__2020/1?ff&camefrom=EMCL_3267537_156470235

Code

The following code was used to fix the issues identified in the original.

library(readr)
library(ggplot2)
library(magrittr)
library(plotly)

data <- read_csv('chart.csv')

chart <- ggplot(data=data, aes(x = Year, Accuracy)) +
  geom_line(colour="firebrick3", size=1.25) +
  labs(title = "VFL/AFL scoring accuracy, 1897-2019", x = "Season") +
  scale_x_continuous(breaks = c(1900,1910,1920,1930,1940,1950,1960,1970,1980,1990,2000,2010,2020)) +
  scale_y_continuous(breaks = c(40, 45, 50, 55), labels = c('40%','45%','50%','55%'))

chart <- chart %>% ggplotly(width= 800, height = 400)

Data Reference

AFL Tables (2020). “Match Scores 1897-2020”. Retrieved May 10, 2020, from afltables.com: https://afltables.com/afl/seas/season_idx.html

Note: this data was scraped from AFL Tables using the fitzRoy R package, developed by James Day, Robert Nguyen, Oscar Lane, Matthew Erbs and Jason Zikovic.

Reconstruction

The below visualisation fixes the previously identified issues. It is also interactive! Hover over any point in the line to see scoring accuracy for that individual season.