Another excellent question.

Obviously, it has to learn from somewhere, so, yes, there IS a database. It's just not part of the distributed system.

The training module -which is not part of the distributed system either - DOES contain a database.

I actually got the idea from the LLMs - Large Language models.
When I saw they were all pre-trained, I thought of how one of the big handicaps with HSH was that it just didn't come with enough data on day one to study or build much of anything.

Hello Dave
1. Does a program like this help a tournament player?
2. I saw a recent video about reports and it reminded me of hsh program. I didn’t fair well with hsh due to its complexity (basic computer experience here) is the determinator more straightforward or does the user need to generate their own factors and weights, etc. ? Is it out of the box ready? Thxs!

Hello Dave
1. Does a program like this help a tournament player?
2. I saw a recent video about reports and it reminded me of hsh program. I didn’t fair well with hsh due to its complexity (basic computer experience here) is the determinator more straightforward or does the user need to generate their own factors and weights, etc. ? Is it out of the box ready? Thxs! — john

__________________
Yes, in fact that is covered a bit in a couple of the videos.
Three things of special interest
1. CHAOS RACE LEVEL (0-5)
2. BALOs (Bet Against Low-Odds horses)
3. AI BEST HORSE
_______________
1. CHAOS RACE LEVEL 4s & 5s indicate that one should look for big payoffs.
_______________
2. BALOs indicate that there is a strong favorite who should under-perform.
_______________
3. AI Best Horse (Odds Ranges)
This tells which horses fit the race by odds ranges.
_______________
The most powerful race for *tournament longshots would be a CHAOS RACE with a BALO.
What's a tournament Longshot? It's a longshot that you can actually handicap to the top.

_______________
In the graphic below, I've taken the projected longshots in the race and indicated how they rank in their appropriate odds categories. Typically you are looking for horses with 1s and 2s.



What the AI trained on could best be described as:
"If a horse in the 17/1 up category won the race, what factors would likely point to the winner?"

Obviously, we are not getting factors like "best speed rating ever" because such horses don't typically pay $36 and up.

Instead, it uses some pretty esoteric stuff. Things like:

• Time-Decayed Stretch Gain.
• Average Positions Gained from 2nd call to the wire.
• Regression Speed Ratings - Low (The worst speed rating the horse could be expected to run.

Also interesting to note that those 1s on the screen indicated that they were ranked #1 or #2 in the field, not just among the longshots. IOW, they were better than the favorites.