Skip to content

tylerkerr/wig-heat

Repository files navigation

wig-heat

a set of Python3 tools to scrape, process, and visualize the data on Blizzard's war3 battle.net ladders listings.

this is now live at arranged.team!

todo:

✅ sequential scraping with error handling

✅ parsing game info from html

✅ writing gamedata to sql db

✅ scraping many games (currently have june 20 thru june 25 on useast)

✅ discarding non-TFT games (mappool, find roc/tft ratio, extrapolate on overlap games)

✅ figuring out which data to visualize

✅ writing queries to pass this data to d3.js

✅ creating visualizations with d3.js

✅ creating interface for the visualizations

✅ automating daily scraping and processing

data presentations:

  • total games per day by realm, all realms (line)

  • total games per day by gametype, each realm (line)

  • separated by gametype: average games per hour over week (week/hour heatmap)

  • internal statistics (table)

some useful gameIDs:

azeroth 21190538 = jan 1 2015

(728k games in 2015)

azeroth 21919343 = jan 1 2016 (1451624400)

(518k games in 2016)

azeroth 22437878 = jan 1 2017

azeroth 22671000 = june 20 2017 9pm

northrend 71407047 = jan 1 2015

(3.484m games in 2015)

northrend 73552502 = jan 1 2016 (games are severely out of order?) northrend 74891032 = jan 1 2016 (1451606400)

(1.791m games in 2016)

northrend 76682523 = jan 1 2017

northrend 77664000 = june 26 2017 2pm

lordaeron 14277875 = jan 1 2015

(366k games in 2015)

lordaeron 14644695 = jan 1 2016 (1451635200)

(397k games in 2016)

lordaeron 15042407 = jan 1 2017

kalimdor 30143853 = jan 1 2015

(1.280m games in 2015)

kalimdor 31424145 = jan 1 2016 (1451606400)

(1.127m games in 2016)

kalimdor 32551595 = jan 1 2017