Machine Posted May 28, 2011 Author Share Posted May 28, 2011 Re: A.I. horses 2 2nds :sad both lost bank now 243.40 Quote Link to comment Share on other sites More sharing options...
Machine Posted May 30, 2011 Author Share Posted May 30, 2011 Re: A.I. horses 1 for today BOG @ B365 :hope Redcar 2.25 Kohala :ok Quote Link to comment Share on other sites More sharing options...
Machine Posted May 30, 2011 Author Share Posted May 30, 2011 Re: A.I. horses winner today Bank now 269.40 :ok Quote Link to comment Share on other sites More sharing options...
Machine Posted May 31, 2011 Author Share Posted May 31, 2011 Re: A.I. horses Last day today and 2 bets to go at . :hope The least the bank can be is 243.40 which would be a ROI of 219.97% and a 15.94% yield since reverting back to Ver 1. Redcar 4.30 Lady Chloe Yarmouth 5.50 Wise venture :ok Quote Link to comment Share on other sites More sharing options...
Machine Posted May 31, 2011 Author Share Posted May 31, 2011 Re: A.I. horses 2 winners Bank now 314.74 Since changing back to progamme ver 1 in post 92 (explained in post #100,new version 1.1 was crap) Starting with the 76.04 I had left the stats are: Starting bank 76.04 Bets 84 NR 3 S/R 35.71% Staked 1018.45 Profit 238.70 ROI 313.91% Yield 23.43% Ave Win Odds 3.74 expected SR Odds 2.8 Edge 33.51% Not bad for a rag arse computer program. Will the last one to leave please turn the light off :ok Quote Link to comment Share on other sites More sharing options...
jimbo1234 Posted June 1, 2011 Share Posted June 1, 2011 Re: A.I. horses fantastic please do it again for june i need to win some money cheers jimbo :clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap:clap Quote Link to comment Share on other sites More sharing options...
Machine Posted June 1, 2011 Author Share Posted June 1, 2011 Re: A.I. horses Due to requests on a biblical scale (1) :eyes and the prospect of jimbo1234 threatening to chain himself to my recently installed cathedral style wrought iron gates (with circles,baskets and rosette panels) and go on hunger strike if I dont, I have bowed to peer pressure and will post up Junes selections. As this thread is not a glory hunt I'll start a new one upstairs in the racing systems forum. :unsure So polish up your brogues don your best bib and tucker, no trainers or baseball caps allowed in there, you wont be mixing with the hoi poloi any more so watch your p's and q's ;) new thread HERE :ok Quote Link to comment Share on other sites More sharing options...
Cowan Posted June 4, 2011 Share Posted June 4, 2011 Re: A.I. horses Just to say I've been involved with systems for a long time and I believe the most you can hope for in terms of strike rate is about 30% with any (profitable) system. I do have one system which is at 42% but this produces few selections and a small profit. Consider that 30% of favourites win year after year. That is a great strike rate but you will make a heavy loss. The most profit is made at around the 20% strike rate but of course that means long losing runs and many people can't tolerate it. I believe that 50% SR is both unachievable and unprofitable. But prove me wrong. :notworthy I see the SR is now at 35% which is less than you expected, but hey, its doing very well, making a great profit! When you created the system was it based on SP? You'll be doing better than that now because you are getting BOG prices. Quote Link to comment Share on other sites More sharing options...
Machine Posted June 4, 2011 Author Share Posted June 4, 2011 Re: A.I. horses Dont want to go into to deep a discusion. The system was derived using neural networks (Artificial inteligence its called now) A database of horses was fed in and the percentage that each variable affects the outcome of the combined races was given. eg.(not the actual prog %ages) H/Cap 8% N/hcap 15% non d or cd winner 8% D/winner 11% cd/winner 12% going 15% SP 11% etc etc up to 100% Then a rating system was devised by using NN for each variable So say the fav had a rating of 100 for race type it would score 15(100*15%) for a n/hc race or 8 for a hc race If the same horses rating for distance was 150 it would get 16.5 for d/winner or 18 for cd/winner or only 12 for neither All the adjusted ratings per race are added together and run through an Analytical Hierarchy Program to sort the potential winners. You should get the idea how it works now The final piece of the jigsaw is the price if I set the price @ 1/2 all gg's with that or above will be fully rated . At the moment Ive set the price @ 2+ so only the ones over even money will be rated fully and then if the price is acceptable to the criteria level from the original NN database tests they will be a bet . Also yes the BOG has helped in improving profit :dude :ok Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.