There was a time in a prior life when I was enthusiastic about the Melbourne Cup - in fact any form of *beast* racing.
That was because I had a system. It was a complicated system that used huge amounts of data to build *predictive models* of a race event - and then used those models to find a selection of *beasts* that would likely be at the front of the field - toward the end of the next race.
And then I used that selection in multiple exotic bets - mainly trifecta's and first four's.
This was my "pseudo scientific" approach to gambling. And it usually worked - wins were much more numerous than losses. For a couple of years it became my primary source of income - but that is another story.
But what quickly became apparent is that the data collection was onerous and time consuming. To be a consistent winner - huge amounts of data are required - some of which is not obvious to the casual punter. Weather and track condition and *beast* temperament are quite significant factors - which the form guides never seem to reliably quantify.
But the most significant factor - has to do with "the insider problem". The industry folks who work and live with *beasts* have insight that is difficult to reliably capture in *predictive models*.
I stopped mucking about with it all when I realised that I couldn't reliably capture that insider knowledge.
But that hasn't stopped others from trying.
The Macquarie Equities Research folks have a good track record - and today they recommend - 1 Dunaden, 24 Kelinni, 15 Maluckyday, 20 Lights of Heaven, 2 Americain and 12 Galileo's Choice.
If you are feeling adventurous then a few trifecta's built with these numbers might be fun.
PS: You know what is hilarious - I won $150 on the sweep at my local pub - with #14 the winner.
That was because I had a system. It was a complicated system that used huge amounts of data to build *predictive models* of a race event - and then used those models to find a selection of *beasts* that would likely be at the front of the field - toward the end of the next race.
And then I used that selection in multiple exotic bets - mainly trifecta's and first four's.
This was my "pseudo scientific" approach to gambling. And it usually worked - wins were much more numerous than losses. For a couple of years it became my primary source of income - but that is another story.
But what quickly became apparent is that the data collection was onerous and time consuming. To be a consistent winner - huge amounts of data are required - some of which is not obvious to the casual punter. Weather and track condition and *beast* temperament are quite significant factors - which the form guides never seem to reliably quantify.
But the most significant factor - has to do with "the insider problem". The industry folks who work and live with *beasts* have insight that is difficult to reliably capture in *predictive models*.
I stopped mucking about with it all when I realised that I couldn't reliably capture that insider knowledge.
But that hasn't stopped others from trying.
The Macquarie Equities Research folks have a good track record - and today they recommend - 1 Dunaden, 24 Kelinni, 15 Maluckyday, 20 Lights of Heaven, 2 Americain and 12 Galileo's Choice.
If you are feeling adventurous then a few trifecta's built with these numbers might be fun.
PS: You know what is hilarious - I won $150 on the sweep at my local pub - with #14 the winner.
No comments:
Post a Comment