Although Project ThunderGo (雷動計劃) has raised concerns in the pro-democracy camp, the results of the Legislative Council election point to its success.
Among the 23 candidates recommended by the project, 19 got elected, a remarkable 82 percent success rate.
In January this year, I put forward the idea of ThunderGo for the first time.
I proposed that pro-democracy candidates should coordinate their campaigns among themselves in order to avoid infighting and maximize their vote share.
Unfortunately, my suggestion only received a lukewarm response from the aspirants even after months of canvassing.
After that I adjusted my proposal and replaced campaign coordination with strategic voting.
I called on pro-democracy voters to cast their votes smartly and strategically on the day of the election in order to maximize their gains.
In less than four month’s time, ThunderGo progressed from just a proposal on a piece of paper to a large-scale strategic voter campaign spearheaded by civil society to make sure the pro-establishment camp would not gain a two-thirds majority in the next Legco.
Rather than a top-down campaign to manipulate the election outcome, ThunderGo was an initiative spearheaded by the pro-democracy voters themselves.
With the help of data and forecasts provided by the project’s researchers, tens of thousands of our fellow citizens organized spontaneously to cast their votes evenly to different candidates.
The objective was to avoid votes being concentrated on one or a few candidates and therefore wasted.
The project aimed to maximize the number of seats won by pro-democracy candidates.
Data provided by Votsonar (雷動聲納）proved instrumental in the success of the project.
It was an online device developed to calculate the odds of pro-democracy candidates based on both our own poll results and those provided by the Public Opinion Programme of the University of Hong Kong.
By accessing Votsonar’s data, voters gained a comprehensive picture of the odds of various pro-democracy candidates, based on which they could exercise their informed judgment and decide who to vote for, in a way that could maximize the number of seats of the pro-democracy camp.
Votsonar would never have become fully operational in just a few months’ time had it not been for the tireless efforts and selfless devotion of a wonderful team of IT experts, statisticians, data analysts and computer program writers, who had volunteered to take part in its development against all odds.
The success of ThunderGo not only marked the triumph of democracy, but also illustrated that our civil society has already come of age.
In hindsight, however, there is still a lot of room for improvement, and we are determined to come up with a better mechanism.
Firstly, we had grossly underestimated public enthusiasm in strategic voting and its implications on the election outcome.
As a result, some long shots received a lot more votes than we had expected, whereas some favorites only won by a narrow margin.
Secondly, we could have done substantially better in updating our data on election day itself, so that voters could get real-time updates on the latest voting developments.
Also, we must improve our system to make it less susceptible to cyber attacks.
Thirdly, due to a lack of resources, most of our analyses and calculations were based on the poll results provided by HKU 10 days before the voting day.
As a result, we were unable to keep up with the latest changes in public opinion leading up to election day, which to a certain extent had undermined the accuracy of our analyses and predictions.
This is something we must work on aggressively to improve the system for future elections.
This article appeared in the Hong Kong Economic Journal on Sept. 12.
Translation by Alan Lee
[Chinese version 中文版]
– Contact us at [email protected]