Alumni day showcases Statistics research

26 September 2017

Associate Professor Rachel Fewster
Associate Professor Rachel Fewster

How can we use rats’ genes against them? How long’s the human genome? And what’s the link between astrostatistics and cricket?

All three questions were answered at the Alumni and Postgraduate Showcase 2017, held on Saturday, September 23 at the Science Centre. The departments of Statistics, Computer Science and Mathematics came together to present short and lively talks on current research, display hands-on technology and postgraduate students’ work and provide course information.

Starting off the day, Associate Professor Rachel Fewster explained that analyses of rat genes gave us a very good tool for tracing the origin of rat invasions or reinvasions. Isolated populations had different genetic makeups, she said, and by doing genetic assignment analyses, it was possible to identify whether a rat was local or from elsewhere, providing routes to stem an influx.

Genetic analysis established that a post-eradication re-infestation on the Broken Islands, just off Aotea (Great Barrier Island) had most likely been started by a rat that had swum 300m from the mainland population. On a small island in the Bay of Islands, genetic analysis on a rat found under a family’s holiday tent established that it had probably come by boat with them, as its genetic makeup was unlike that of local rats.

Dr Brendon Brewer
Dr Brendon Brewer

Professor Thomas Lumley shared the startling fact that the human genome, were it a zipper, would stretch 7000km from New Zealand to Hawai’i. Each grandparent gives you about 60km; a gene would be a few metres long. In discussing the accumulating knowledge about genes and their links to disease, he raised a thorny question: “When should someone’s genes actually affect what health advice they get whether it’s diet or medication choices or whatever? That turns out to be a much harder question than it sounds it should be.”

The link between astrostatistics and cricket was Dr Brendon Brewer. His work involves applying statistics to astronomical problems – but he’s a keen cricket fan and applies statistical methods to cricket, too. He used Bayesian survival analysis to find out how long it took batsmen coming to the crease to adjust to the conditions and start batting consistently – it’s called “getting your eye in”.

Brendon compared Don Bradman, the Australian cricketer who died in 2001, and New Zealand player Ross Taylor, and found there was only a 52% probability that Bradman, a national hero in Australia, was better than Taylor when they were both on a score of zero.

Brendon, a transplanted Australian, has teamed up with New Zealand Cricket and doctoral student Oliver Stevenson to explore issues such as players’ typical career trajectory and how much a player can be expected to contribute to a team’s possibility of winning.

 

See photos from the day on Facebook.