About

History

This site exists due to an interesting sequence of events. As you can imagine, I live in Melbourne. Although I did follow AFL lightly when I was younger, it was only when I went to my first game in the middle of 2021 (in between lockdowns) that I began to properly follow it. Then I remembered that back in 2019, a friend and colleague of mine had mentioned he was trying to build a Brownlow medallist predictor (surprise, surprise) and was using a site to get relevant data (the site, of course, was www.afltables.com). Now during this time (back in 2021), I happened to be working on a work-related project that required stronger fluency in Python. So what better way to practise my Python than by practising web scraping that very same site. I was well aware of fitzRoy (the popular R package) but I wanted to code from scratch. Of course as time went on, I discovered other sources of information that enriched (and sometimes challenged) the original set of data I extracted, and of course I began extracting those too. It would be safe to say that I had fallen down the rabbit hole, but the act of collecting truly is both therapeutic and satisfying, and is what I imagine allowed hobbies such as stamp collecting to become not uncommon. In my profession, a lot of the work I do involves writing SQL queries, but it so happened that the project I was now working on required expertise in data modelling. What fortunate timing since this was exactly what I needed to do with this mountain of data that I now had. So the next step then became integrating all the extract data into a local database through a robust data model. The only shame there was that the project leaned towards DFM (dimensional fact model) modelling whereas I preferred 3NF (third normal form) modelling, which lends itself to being more flexible and error-avoidant with this domain of data. After all the effort (and further effort after a couple of missteps here and there), it happened that the very same project at work would soon require me to be become more familiar with front-end and API development. Well, at this point, why not just dive off the deep end and try make a site, right? (The API-related stuff also somehow became related to this site, but of course it was to help quench that thirst for more data.)

Sources

Although this site will include links to provide credit to (some of) the relevant sources used, it can be difficult for me to maintain them. As such, to explicitly give credit to where information typically came from (roughly in chronological sequence):

  • AFLTables
    • Baseline source of information regarding matches, players, teams, and field umpires
  • AustralianFootball
    • Detailed player details (e.g. dates of death)
    • Proper round names (e.g. sectional rounds, round robin finals, and finals numbering)
    • Sponsorship names for venues (e.g. Optus Oval)
  • Wikipedia
    • Detailed player details (e.g. full names, titles)
    • Captain information (at a season level and for grand finals, but not a match level)
    • Home grounds
    • Historical finals systems
    • Cross-referencing any other source
  • Draftguru
    • Draft information
  • AFLUA
    • Umpire information
  • AFL
    • Detailed match information post-2001 (where available)
There are times where it may seem like the information on this site does not reflect its sources. This is often because it might not be clear why this difference exists. An example of this would be which team is the home team and which is the away team for some of the finals matches. This is because I have changed it to align with the description of the finals systems provided by Wikipedia. It would seem as though that many sites might be interpreting that if a finals match was played at a particular team's home ground (for that season), then that team must be the home team, but this logic is flawed for a multitude of reasons (including teams that share a home ground, the grand final is typically played at the MCG by design, and matches that moved in the 2020-2021 season due to COVID). That is not to say that home ground advantage should not be considered, but it should not determine the status of home/away. Another example of what looks like discrepancy is the actual figures themselves. There will be matches where this site will report different statistics than many of the sources, but often this is because the figure was updated with values from the AFL Match Centre. It should be noted that Champion Data performs reviews of its statistics and can update them. For a specific example of this, there is a match (although I am struggling to remember the specific match), where a Geelong player was originally reported to have made one kick. Of course, this is what many fan sites then reported. However, the advanced statistics then stated that the same player made two effective kicks, and updated the kick tally to two. Although some of these fan sites added the effective kick tally, many (if not all) did not update the kick tally. What I would give for the team I barrack for personally to be able to have two effective kicks per kick.