Loading ...
Sorry, an error occurred while loading the content.

MAIT: Artificially Intelligent Tipster

Expand Messages
  • NHNE
    NHNE News List Current Members: 758 Subscribe/unsubscribe/archive info at the bottom of this message. NHNE 2002 Fall/Winter Fundraiser: Money needed = $2090.00
    Message 1 of 1 , Dec 13, 2002
    • 0 Attachment
      NHNE News List
      Current Members: 758
      Subscribe/unsubscribe/archive info at the bottom of this message.

      NHNE 2002 Fall/Winter Fundraiser:
      Money needed = $2090.00
      Donations to date = $1200.00
      Number of people who have helped = 25
      Funds still needed = $890.00
      To make a tax-deductible donation:
      Web: http://www.nhne.com/main/donations.html
      eMail: nhne@...
      Phone: (928) 282-6120
      Fax: (815) 346-1492
      Mail: NHNE, P.O. Box 2242, Sedona, AZ 86339
      Credit Cards: Please include number, expiration date, phone number.
      Thank you!


      By Paul Marks
      New Scientist
      December 11, 2002

      Human gamblers are all too fallible, but some sure-fire tips are now on hand
      from a distinctly non-human source: intelligent software.
      A safer bet

      Alan McCabe, an IT researcher at James Cook University in northern
      Queensland, has developed a software-based results tipster for Australian
      Rugby League -- although it could just as easily be adapted for soccer,
      baseball or cricket. The program outperforms the best human tipsters.

      McCabe unveiled his Artificially Intelligent Tipster -- MAIT for short -- at
      AI 02, an artificial intelligence conference in Canberra last week. The
      project is a spin-off from research into handwriting recognition. But
      instead of being fed data on writing shape and pen stroke velocity to verify
      signatures, MAIT uses a neural network to learn which features of a team's
      performance make them winners.

      A neural network is a computer system that can learn which combinations of
      inputs (representing team performance statistics, say) lead to a particular
      output (such as a team's chance of winning). Crucially, the network's inputs
      can be "weighted", so the influence of each can be increased or decreased
      depending on how significant its contribution is.

      Big-name stars

      MAIT has been trained with match data obtained from a national bookmaker.
      His range of input features includes many you would expect, such as each
      team's current success rate and the points scored for and against them each

      There is one notable facet that MAIT does not consider: the availability for
      a game of big-name sports stars. If they are ruled out by injury, suspension
      or national team duty, their absence might dent team morale and bolster the
      opposition's will to win. So a human tipster might bet against the missing
      star's team.

      But assessing star quality is too human a feature for a system that is
      supposed to be automatic, says McCabe. "A conscious effort was made to
      ensure that no subjective features were included," he says.

      It is a strategy that seems to be paying off. Across the season, MAIT is
      outperforming human tipsters and getting its predictions right more than 66
      per cent of the time.


      NHNE News List:

      To subscribe, send a message to:

      To unsubscribe, send a message to:

      To review current posts:

      Published by NewHeavenNewEarth (NHNE)
      eMail: nhne@...
      NHNE Website: http://www.nhne.com/
      Phone: (928) 282-6120
      Fax: (815) 346-1492

      Appreciate what we are doing?
      You can say so with a tax-deductible donation:

      P.O. Box 2242
      Sedona, AZ 86339
    Your message has been successfully submitted and would be delivered to recipients shortly.