McCabe's Artificially Intelligent Tipper (MAIT) was devised by Dr Alan McCabe, who at the time was a Ph.D. student and lecturer at the School of Information Technology at James Cook University in North Queensland, Australia. MAIT is the "ultimate computerised tipping system, far more powerful than your standard statistical database".
The software itself is a by-product of Alan's existing research into handwritten signature verification. This existing research incorporates a similar artificially intelligent model that learns to recognise handwritten signatures and determine whether they are genuine signatures or attempted forgeries.
MAIT's primary focus was originally the Australian National Rugby League (NRL) competition and it uses concepts from machine learning and artificial intelligence to examine various features of these sporting contests. Since the original system, expansion sports have included Super 12 Rugby Union, the Australian Rules Football League (AFL) and the Rugby World Cup. The model learns which features (such as home ground advantage, team winning percentage, points scored etc.) are most important and applies what it has learned to tipping future matches. Technical details can be found in the research paper on this site.
During 2002 MAIT underwent its first "real-time" testing (and has continued since then) where tips were published in newspapers and placed on the web site several days (typically 5 to 7 days) in advance. The testing went quite well and you can take a look at the performances and some of the media attention that MAIT received in the MAIT's Rates section. Since then, sporting tips have appeared in various regular articles throughout Australia and overseas.
Further expansion sports are being considered and you can check out the details in the future plans section.