Dr Alistair Legione

Dr Alistair Legione

Lecturer in Veterinary Microbiology and Public Health at the University of Melbourne

NBA Game Of The Day

2 minute read

Published:

A shiny web application that ranks daily NBA games for your viewing pleasure.

This application is primarily designed for those of us outside the USA, who have to watch NBA games on delay due to time zone differences. This app will show you, based on a formula, which game is going to me most worth your time investment. Particularly useful if you only have time to watch one game of the night, and you don’t want to waste it on the 40 point blowout

Currently in alpha form, please feel free to submit comments, suggestions, or pull requests

All data comes from the NBA Stats API, and all data and any team logo images belong to the NBA.

Hopefully one day it will be available online, but currently only local versions work.

Run your own local version

You can run ‘NBA Game of the Day’ on your own computer.

To run, start by installing the programming language R and the GUI interface RStudio

Then either clone this repo and run the RunShiny.R script

git clone https://github.com/alegione/NBA-Game-Of-The-Day.git
cd NBA-Game-Of-The-Day
Rscript RunShiny.R

OR run the below commands in the Rstudio console

packages = c("shiny", "jsonlite", "dplyr", "DT", "curl", "httr")
install.packages(packages, repos = "https://cran.rstudio.com/")
library(shiny)
runGitHub("NBA-Game-Of-The-Day", "alegione")

This will download and run the Shiny application locally

Game Ranking Algorithm

The aim of the ranking system is to assign value to a game based on ‘watchability’ or ‘excitement’. This is hard for numbers alone to achieve but we’ve given it a go (and open it ideas). Current methodology follows the below basic formula:

Score = 'Number of overtimes' x 'Overtime Value' (default: 3) +
        'Final Margin Value' -
        'Final game margin' +
        Individual player highest points scored x 'Top Player Points Value' (default: 0.1) +
        (Team highest points scored - 100) x 'Top Team Points Value' (default: 0.1)

There are two final score adjustments possible, which can both be utilised at the same time.

1) Win Percentages: Multiply final ‘Game Rank’ score by the sum of the win percentages of the competing teams. For example, two teams above .500 will increase the total ‘Game Rank’ score 2) Nightly Margin: Rather than using an arbitrary value for the ‘Final Margin Value’, replace it with the maximum margin for the night.

The Interface

The basic features of the interface, and how to run the program are pretty simple. Select the date of the games you want to rank, adjust any parameters, and click on the button. A table will generate ranking all the games of the night. If you adjust any of the parameters from this point, a new table will be generated.

"Example of program interface"

Comments

Quick tips for using ffmpeg for video editing

4 minute read

Published:

In this brave new world of 2020, many of us are finding we have to use Zoom or other similar video conferencing software tools to teach classes. Then we wish to take those recordings and quickly edit them to trim, compress, and upload them to an online server. I’ve found the command line tool ffmpeg to be extremely helpful in this endeavour. Below are a few quick commands you can use to get your recording from its raw form in to a final version in just a few commands.

Building your Online Research Profile - Twitter

14 minute read

Published:

Twitter is the social media hub of choice for a vast array of academics. It’s a great place to share your latest work, ask questions of experts, communicate your research field to the general public and engage in the general discourse of your community. It can potentially be a hive of politically charged vitriol too, but we can hopefully avoid swimming in those circles more often than not.

This guide will take you through the basics of setting up a profile, and working through the basics of a ‘tweet’, your ‘timeline’, your ‘followers’, ‘likes’ and ‘retweets’. There is a decent amount of platform specific terminology (much of it general parlance for social media these days), but we’ll get through it together.

Building your Online Research Profile - Google Scholar

4 minute read

Published:

The goal of improving your online profile is that when people search your name, you’ll appear near the top (or ideally at the top) of the search results. I’d say globally the majority of people at time of writing will be searching using Google, so it makes sense to set up your Google Scholar profile too. Since lots of people search for papers through this service nowadays too, having a scholar page has the added benefit of adding a ‘clickable’ link to your name when your papers are searched. This link takes them through to your profile page where they can find even more of your papers! How convenient.

Building your Online Research Profile - Find an Expert

5 minute read

Published:

Most researchers in a University system will have a staff profile page. Here, your contact information, a photo, and your research outputs are going to be listed. Also, for most institutions this will have a brief biography on who you are as a researcher. This can be a valuable first contact for people searching you out on the internet as a University profile is likely to be considered one of the more trustworthy sites when people are trying to find you.