Artificial Intelligence: Questions they should answer

[Dave Schwartz]

Imagine...

• ... that someone (who shall remain nameless) is getting ready to (finally) release (in beta) his new Artificial Intelligence software.
• ... there are AGENTS who know quite a lot.
• ... that you can ask them questions.

Examples of potential questions...

• What is the PACE SHAPE?
• Who's likely to go to the front?
• Which paceline(s) should be selected?
• Could this be a CHAOS RACE?
• Is there a vulnerable favorite?
• What factors did YOU find important?
• How should FTS be handled?

WHAT QUESTIONS WOULD YOU WANT TO BE ABLE TO ASK?

[Bustin Stones]

Tell me if this makes sense to you:
Most people can figure out which horses have the pace advantage in a given race based on the strengths of each entry.
The connections draw very similar conclusions.
the connections of the disadvantaged horses realize they can change the dynamics of the race by disrupting the predictable pace.
The attempt to change dynamics either succeeds, yielding a result not predicted or the attempt fails for various reasons and it may or may not change the results.
Or the advantaged horses fail to duplicate their advantage sufficiently and, in doing so, change the result.
So, you could theoretically handicap a card to perfection and never bet a winner.
I don't see AI changing this because it cannot know which connection(s) decide to depart from what they are expected to do.

[goforgin]

Is there a vulnerable favorite? Also, does the AI take into consideration the trainer/jockey win and ITM percentage, class rating, earnings per start and the actual horse's win and ITM percentage?

[Dave Schwartz]

Originally Posted by goforgin

Is there a vulnerable favorite? Also, does the AI take into consideration the trainer/jockey win and ITM percentage, class rating, earnings per start and the actual horse's win and ITM percentage?

It considers just about EVERYTHING.

Of course, it doesn't actually USE everything in every race.

BTW, there isn't actually ONE AI.

They are called AGENTS and there are 8 of them.

Each has their own approach to the game. You can select as many as 5 to be active in any given race.

In fact, you can create your own TEAMS.

For example, imagine you could have different teams for:

• Claiming Races
• Graded Stakes
• Turf Races
• 2-year olds

[traveler]

Originally Posted by Dave Schwartz

It considers just about EVERYTHING.

Of course, it doesn't actually USE everything in every race.

BTW, there isn't actually ONE AI.

They are called AGENTS and there are 8 of them.

Each has their own approach to the game. You can select as many as 5 to be active in any given race.

In fact, you can create your own TEAMS.

For example, imagine you could have different teams for:

• Claiming Races
• Graded Stakes
• Turf Races
• 2-year olds

Is this similar to the Agents in HSH?

[Dave Schwartz]

Not even remotely close.

[paulbenny]

The data converted to information such as the jockeys and trainers know who believe there is lone speed and the before data shows that in the pace analysis and there is history with the jockey or trainer moving to the front forcing the pace could be available if someone loaded this kind of information into data that lead to information and the AI included it. The point is that AI is going to still be limited unless there was information out there that deals with scenarios that are documented. Not one case although one case could be written up and it was swept in.

If someone did this analysis before and after the race and it was substantive it would be swept into the equation.

[mountainman]

Originally Posted by Bustin Stones

Tell me if this makes sense to you:
Most people can figure out which horses have the pace advantage in a given race based on the strengths of each entry.
The connections draw very similar conclusions.
the connections of the disadvantaged horses realize they can change the dynamics of the race by disrupting the predictable pace.
The attempt to change dynamics either succeeds, yielding a result not predicted or the attempt fails for various reasons and it may or may not change the results.
Or the advantaged horses fail to duplicate their advantage sufficiently and, in doing so, change the result.
So, you could theoretically handicap a card to perfection and never bet a winner.
I don't see AI changing this because it cannot know which connection(s) decide to depart from what they are expected to do.

Great post. And the instances in which (very) human intervention willfully alters a race-flow can often be predicted.