It’s been an awesome adventure with the P104 team; from writing policy proposals, to organising local community events, to drafting the alternative budget, attending to residents’ concerns, and the weekly badminton sessions (my favourite!).
One of my inspirations do study Mathematics during my university days was from the TV series Numb3rs, where two brothers (one’s a Mathematics professor at a US college, while the other’s an agent with the FBI) solved crime using methods like k-means clustering and geographic profiling. When I discovered the theories that were used on the show were real, I thought to myself, “Now that’s who I want to be when I grow up.”
My skills came in useful in 2014 when there was a need for someone to create a fiscal model for the alternative budget: to find a way to estimate future revenue and expenditure based on a number of inequalities and assumptions. This eventually evolved into electoral modelling for the first by-election after GE2013 at Teluk Intan. Using information like ethnic proportions in each voting stream and polling station, I was able to come up with a range of expected outcomes (i.e. specifically which party would get how many votes etc.). It turned out my model was off by over 6,000 votes, as one of the factors I heavily underestimated was voter turnout. My models for future by-elections were improved once I factored in proportions of voters who worked out of state.
Fast forward 3 years later, I’ve been recruited into INVOKE, where a team of data analysts are using highly advanced technologies like cloud computing and machine learning to model electoral outcomes for the upcoming general election. We’ve developed our own predictive models to determine, at the individual level, how fencesitters can be expected to vote, how huge numbers of newly registered voters will affect electoral outcomes at the state and parliamentary seat level, and even which party should stand at which seat to maximize its chances of victory. Before INVOKE started, I never would have thought of creating mathematical models for political campaigns. It just feels surreal to be in the shoes of the Mathematics professor from Numb3rs, explaining the use and results of statistical techniques like regression analysis, analysis of variance (ANOVA), and Markov chains for forecasting purposes.
I’ve been lucky to be able to work with colleagues who taught me the values of humility, humour, and empathy. I’ve worked with highly intelligent, down to earth people I deeply respect: Wong Chen, Rafizi Ramli, Saifuddin Nasution Ismail to name a few. If you’re looking for a place to intern, to learn more about what happens behind the scenes in an MP’s office (or to put your mathematics/computing skills to really awesome use), do consider working here at P104:)