Nice one,
giuseppe_esq. Good luck.
Hi
Steeplechasing and
simon boardman
Many thanks.
I have seen both of your posts and you have provided some amazing write-ups about the reasoning behind your selections! Well done. I am sure you will be very popular on this forum and will help many others learn from you.
I fall into the category of "don't always provide the reasoning behind my selections" which can be frowned upon. I have got to the stage where I think that this forum may not be suitable for how i make my selections, and feel that it may be more appropriate to post on a more technology based forum.
This is why I put the message
'I try to predict WHAT horse will win, not WHY it will win'. in my message footer. This is not meant to offend, it just says what I do on the tin.
Now my background, I am not a horse race bettor, I only watch horse racing on weekends and I have a very limited knowledge.
My knowledge has improved by learning from some of the amazing people on here, whom i cannot thank enough, nor will I ever forget.
I am self-employed, working as a Business Intelligence Consultant, which can be very demanding at times.
So why am I here? To be honest, I have spent the past 2 years or so trying to develop a fully automated system to create my own horse racing rating using historical data using a variety of data processing techniques, ranging from data mining, creating speed ratings, my personal official ratings etc. I love working with data to try and predict future outcomes (probabilities).
The difficulty isn't in creating the automated systems, it is in picking enough winners to be profitable.
One thing I have learnt, is that, despite the vast amount of data on horse races that is available, this task is very difficult and complex. I often feel it is borderline impossible. I am not a mathematician, so have tried to learn a variety of mathematical techniques in my efforts to create some effective ratings.
I run several automated scripts to generate the ratings for every daily racecard in the UK and Ireland, which now takes me approximately one hour per day. I could probably refine this to get this down to less than an hour.
I have got to the stage where I now need to tweak some variable values to try and improve the selection accuracy and returns.
I may never succeed, but I won't ever stop trying.
Goodluck in all your future bets.
Regards
Giuseppe