Originally Posted by redietz View Post
Originally Posted by accountinquestion View Post
Originally Posted by Half Smoke View Post
.
I knew about Bill Benter but not about the other guy that mickey posted about
I should have limited my post about how few if any win to the U.S.
I believe U.S. racing is much, tougher to beat than Hong Kong racing
in the U.S. the field size is much smaller, creating fewer opportunities
there are other reasons I don't want to get into right now

in Hong Kong the average field size is 10.63 starters per race and the average win price is $14.26

in the U.S. the average field size is 7.40 - I couldn't find the average win price in the U.S. but I'm sure it's much, much less than that

field size in the U.S. has consistently shrunk over time and it greatly impacts the payouts

and as discussed before re Billy Walters - Benter beat racing in Hong Kong using in depth statistical models

Andrew Beyer and others like him, when they were able to beat racing, were betting based on their personal opinions re the races and the horses entered

Benter's winnings and accomplishments dwarfed Beyer's - but I find Beyer much more interesting - betting by forming an opinion of the race is much more appealing to me, it makes the game fun, but I acknowledge it can't be anywhere near as profitable

to sum it up, right now with the small fields and the small payouts, the gigantic takeout and with the everyday players who know the game well even if they don't win - the best of them probably about break even - I don't think anybody can beat U.S. racing - foreign racing - I don't know

.
Advanced statistical models are far closer to AI than opinions.

Opinions may be more impressive but always subject to change in various ways.

My main point was that advanced statistical models need more data. Actionable data. This will be the same for ai.


LOL. So close, yet so far. You do realize that the entire discussion hinges on the proper definition of "data?" How does one define "data?" More importantly, how SHOULD one define "data?"

I'm giving away half the farm with this but looking where the light shines is a real problem when you deal with definitions.

And finally, you do realize there have been people running programs 24/7 for 20 years to create statistical brushstroke equivalents of real events? Do you really think anything someone can do today wasn't investigated five, or even 10, years ago by people with virtually unlimited resources?

And you're missing the main gist of Hong Kong racing, by the way, that makes it theoretically solvable and that separates it in essence from US racing. But I'm sure you can figure that out for yourself.
This thread is for professional Sports betting not professional clown college. You are in the wrong thread. It was a discussion about AI.

Clown you should read the Wiki page on expected value to learn even how more wrong you are about so much. There are likely other reasons that the Hong Kong guys were able to beat it I really do not remember the details but part of that is to have more data. This is what ai relies on. This is why the style of AI that has taken over relies on bots that are just hammering every website all over the web. Because they want more and more data to train the ai off. So when someone asks a question about AI in relation to sports betting then yes the model matters. AI is just a catch all. All these programs that you speak of having run for 20 years could also just as well be considered AI.