README.md
In pbulsink/HockeyModel: NHL Hockey Model for 2018-2019
HockeyModel
This is a model to predict NHL game winners & their likely performance for the rest of the season.
Read more about the package at https://pbulsink.github.io/HockeyModel
Current predictions are below, and are always posted on Bluesky at @bot.bulsink.ca.
Team Ranking
![Team Rankings]("https://github.com/pbulsink/HockeyModel/raw/master/prediction_results/graphics/current_rating.png)
Total Point Predictions
Point Liklihood Ranges
Playoff Odds
President's Trophy Odds
Recent Changes
Installation now doesn't automatically install documenting, testing, tweeting, graphics and parallel processing requirements. You may wish to run install.packages(c('knitr', 'rmarkdown', 'testthat', 'webshot', 'progress', 'markdown', 'covr', 'tictoc', 'parallel', 'ggrepel', 'gt', 'ggalt', 'ggridges', 'scales', 'ggplot2', 'rtweet', 'devtools', 'usethis')).
To Do
- ~~Twitter user @MOCallanain highlighted that the predicted tie rate is < ~0.2, when in actuality it's higher. Likely due to teams playing for loser point, can we parameterize the model to include a tie boost? Diagonal enhanced metric should work - calculate like the DC 0/1 goal enhancement. Model currently re-scales to increase odds to a reasonable amount.~~ DONE
- ~~Currently, only scores are used for model generation. Moneypuck has an expansive expected goals model available for download and updated regularly (see http://moneypuck.com/data.htm). Deriving the team performance by expected goals instead of actual could reduce the impact of luck on expected future performance.~~ Won't Do - Use my own xG model instead (see BulsinkBxG).
- ~~Switch to NHL API for scores and schedule~~ DONE
- ~~Twitter user @joseph__ii picked up on a quirk of the OT/SO odds assignment (see https://twitter.com/joseph__ii/status/1357785234285109248). Try rebalance with league or teams' OT performance measure? ~~ DONE
pbulsink/HockeyModel documentation built on Dec. 16, 2024, 8:03 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
HockeyModel
This is a model to predict NHL game winners & their likely performance for the rest of the season.
Read more about the package at https://pbulsink.github.io/HockeyModel
Current predictions are below, and are always posted on Bluesky at @bot.bulsink.ca.
Team Ranking
Total Point Predictions
Point Liklihood Ranges
Playoff Odds
President's Trophy Odds
Recent Changes
Installation now doesn't automatically install documenting, testing, tweeting, graphics and parallel processing requirements. You may wish to run install.packages(c('knitr', 'rmarkdown', 'testthat', 'webshot', 'progress', 'markdown', 'covr', 'tictoc', 'parallel', 'ggrepel', 'gt', 'ggalt', 'ggridges', 'scales', 'ggplot2', 'rtweet', 'devtools', 'usethis')).
To Do
- ~~Twitter user @MOCallanain highlighted that the predicted tie rate is < ~0.2, when in actuality it's higher. Likely due to teams playing for loser point, can we parameterize the model to include a tie boost? Diagonal enhanced metric should work - calculate like the DC 0/1 goal enhancement. Model currently re-scales to increase odds to a reasonable amount.~~ DONE
- ~~Currently, only scores are used for model generation. Moneypuck has an expansive expected goals model available for download and updated regularly (see http://moneypuck.com/data.htm). Deriving the team performance by expected goals instead of actual could reduce the impact of luck on expected future performance.~~ Won't Do - Use my own xG model instead (see BulsinkBxG).
- ~~Switch to NHL API for scores and schedule~~ DONE
- ~~Twitter user @joseph__ii picked up on a quirk of the OT/SO odds assignment (see https://twitter.com/joseph__ii/status/1357785234285109248). Try rebalance with league or teams' OT performance measure? ~~ DONE
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.