superrugby.py

Tool for backtesting superbru.com predictor models and using them for making future picks in Super Rugby seasons. Includes a tuned ELO-based prediction model that's currently the best I can come up with.

Pre-COVID the ELO+regression model scored a few top 1% finishes and never finished worse then top 5%. Can confirm that it does not work for beating the bookies though.

Name		Name	Last commit message	Last commit date
Latest commit History 34 Commits
.gitignore		.gitignore
1802.00527.pdf		1802.00527.pdf
LICENSE.TXT		LICENSE.TXT
README.md		README.md
data.txt		data.txt
dataanalysis.xlsx		dataanalysis.xlsx
elo.py		elo.py
generate.py		generate.py
make-plots.py		make-plots.py
melo_sr.py		melo_sr.py
melomodel.py		melomodel.py
ratings.png		ratings.png
spreads.png		spreads.png
sr_hist.txt		sr_hist.txt
srdb.py		srdb.py
super_rugby_historical_data.xlsx		super_rugby_historical_data.xlsx
superelo.py		superelo.py
superrugby2017.py		superrugby2017.py
test.py		test.py
test.xlsx		test.xlsx
totals.png		totals.png

License

alan-mcl/superrugby

Folders and files

Latest commit

History

Repository files navigation

superrugby.py

About

Topics

Resources

License

Stars

Watchers

Forks

Languages