A pinboard by
George Ng

Ph.D in Biotechnology who have joined a eCommerce startup in Hong Kong


When you track the NFL using chips in footballs and shoulder pads, analyse player

When Scalable Machine Learning Meets The NFL The NFL vernacular entered my world in June 1984 when Machintosh was featured in the Super bowl half-time ad. Tiltilating video of busty tank-top woman hammer thrower and a new computer didn't make sense, neither did the NFL game. Coming from Asia, footballs were round and didn't have pointy ends. The NFL didn't show up on my radar again until Jerry Maguire and again when I wept over Forrest Gump. But every other American male I met gushes over the NFL, convincing me that I've missed out on a beautiful game. A former boss berrated that the alltime NFL best picks were Tom Brady and Ray Lewis in 2000 and 1996 respectively (read more).

Too Much Of A Good Thing? The NFL have been late to using statistical analysis to improve game performance, playing catch-up with Major League Baseball and NBA. But when is a good thing too much? Besides live-streaming the entire game on Twitter, The NFL football and shoulder pads have embedded chips to digitally track a player's on-court coordinates in relation to how the ball moves on the field. Motion-tracking cameras in every arena were installed to further track player and ball positions 25 times a second(read more). Post-game real-time telemetry event data points allow both fans and the media to pour over their favourite sport team. Once worshipped and awe inspiring sporting heroes can now be statistically examined for every shots created, defensive play and ball touches.