Elo-Project

Custom Elo Class that was inspired by another github user:
https://github.com/ddm7018/Elo/blob/master/elosports/elo.py

ELO FORMULAS used were found on wiki:
https://en.wikipedia.org/wiki/Elo_rating_system

Based on an Elo formula from wikipedia, I have developed a class that incorporates the pandas library to create custom Elo ratings for any contest. Elo takes in a pandas DataFrame of scores, compares the scores of multiple people, then calculates the elo. It then stores the Elo object in pickle (I have a simple pickle_read & pickle_write functions, that you can use, or just write one yourself).

To start using call the generic_league function found in eloLeauge.py. This can work for multiple events at once, or a singular event. At the moment, it evaluates on date, so only one event per date for now, but I hope to change that to have a unique key for each activity so you can incorporate multiple events in the same day.

generic_league(df,
                score_column,
                file_path = './pickled-elo.p,
                lsw = False,
                save = True
                )

Parameters

• df
  • type: pandas DataFrame
  • In your df you must have 2 columns (case insensitive):
    • Date
    • Player
    • A Score Column (any name)
• score_column
  • type: string
  • Name of the score column (str)

Optional Parameters

• file_path
  • type: string
  • Defaults to ./pickled-elo.p
• lsw (Low Score Wins)
  • type: bool
  • Defaults to False
  • Set to True if you are playing a game where the lower score would win
• save
  • type: bool
  • Defaults to True
  • This is if you would like to save your Elo Object to pickle
  • Can be set to False if you would like to see a temporary result, or for testing

Returns elo object

Notes

• Elo starts at 1500 and will fluctuate on results
• Expect to see drastic changes at the start, but as you get more results the elo will become consistent
• Default file path (if not passed in) will be in the cwd and titled pickled-elo.p
• Class will add new players as their names show up on the list
• Drop_inactive removes from output (but still saves) players that haven't participated in the last 3 events

I have integrated a custom league for weekly work activity with Google Sheets using the Pygsheets method:

.set_dataframe()

I use that to save the following to different tabs in a Google Sheet. This was the easiest way for everyone to see the scores.

• To save/see the current elos

self.show_elo()

Example Output

	Rank	Player	ELO	    Games Played
0	1	John	1702.0	    5
1	2	Sue	1685.0	    4
2	3	Guy	1578.0	    3

• To save/see the historical elos

self.get_df()

Example Output

date	        John	Sue	Guy
2021-04-16	1618.0	1430.0	1511.0
2021-04-23	1594.0	1503.0	1511.0
2021-04-30	1653.0	1503.0	1511.0
2021-05-07	1690.0	1609.0	1540.0
2021-05-14	1702.0	1685.0	1578.0

• To save/see the winners and losers

self.winners_and_losers()