This repo will be reimplemented by storing the data scraped from HLTV in a database. This should make performing analysis on information besides the world rankings much easier once a local database has been built.
python3 src/scrape.py
will scrape the world rankings data from HLTV and store the output in both a csv and pickled pandas data frame in the data directory. They will overwrite the data files currently in the data directory.
python3 src/plot.py plot_type
plot_type can be either 'points' or 'ranks'. 'points' graphs each team's rank from the points system HLTV uses to derive the ranking. 'ranks' will graph the top 30 teams.
Because of the amount of data and the fact that many teams have come into the HLTV top rankings for only brief periods at a time, the current plotting method culls a lot of the data. For the points plot, this means that only teams who have appeared in at least 1/2 of HLTVs world rankings will be plotted. For the ranks plot, only teams who have at least made the top 5 will be displayed.
Often is the case in Counter Strike where the core (or entirety) of a team changes their team name or moves to a different organization. It would be reasonable, then, for the lines from these two teams to be connected, however currently this is not the case. A mapping from a previous team name to the current team name could be constructed manually, as would be the case for TSM -> ? -> Astralis, however since in many cases it is important to know the date the team members changed (LG -> SK -> MiBR), this would also have to be taken into account.
Currently the colors are random, and only the teams who are in the most recent HLTV top rankings (and who pass the culling parameters) will be labeled. Ideally the lines should be the same as the team's primary color (or the line could have a stroke with the team's secondary color), but that will come with either a manual mapping or another scraping tool that gets information from the team's logos stored on HLTV.
Ideally this kind of plot would be somewhat interactive so that it could be more easy to follow particular lines when exploring with the mouse. This would make it much easier to read exact values and follow the path of a particular team over time. The likely solution to this is to use plotly instead of Matplotlib, which currently is slated for future development.
Further, an option to simply update the current data files rather than re-scrape the entirety of HLTVs rankings should be added. This, in addition to being able to scrape regional rankings as well should not be too difficult to add.