Since there is no database there cannot be any testing unless it is done manually.

The challenge with the approach you have outlined has many challenges.
_____________________
1. Since all the handicapping is dynamic, there is no mechanism that would allow automated testing.
Thus, to test against (say) 5,000 races, SOMEONE would have to literally bang those 5,000 races in manually.

In addition, if you DID pound in (say) 5,000 races, you'd logically want to do it live. Those 55 races took me about 6 hours with the tracking. That would put a 5k test at around 550 hours.
_____________________
2. Because I cannot have the data being exported and, therefore, easily available to the whale world, there can never be an automated testing process.
_____________________
3. Because the whales are the great influencer of our time, and they retrain their models annually, this new testing will need to be done once a year.
_____________________



FWIW...
In my 3+ decades in the horse betting business, I have found that the typical handicapper has no problem using the concepts and factors that he THINKS are important, and will use those without hesitation, despite the fact that he has done no real vetting.

He will consider a single pretty good day as enough proof to start using his new idea.



I suggest that the only way forward with The deTERMINATOR is to adopt a similar belief system.

Remember, I never said that AI BEST HORSE is the answer that will make you a professional. You will probably need to add your own secret sauce to make it work the way you want it to. (Truth be told: I already have created my own, derived BEFORE I did the 55 races.)

I really don't want 200 guys doing the same thing.
In fact, I'd make that a requirement if I could.

(On that particular topic, I have a practical answer.)

Hey Dave, will the video be available to view on this site? — Mark

Soon.
Check your messages.

what if I don't want to Dutch the horses, play only 3 even if 4 are possibilities, etc. — Andrew Kramer

Perhaps share your questions here?

And, did you see this thread?
About That 55-Consecutive-Race Test

As a follow up to a question I asked Dave during last week's meeting, can we get a general sense about what data the AI models are trained on? Is it a fair assumption that they were fed all of the 3000+ factors from HSH during training? Similarly, were they then evaluated against out-of-sample test data? — Atakante

A GENERAL answer would be: It started with one year of everything, plus multiple years for some of the small sample types.

As for what factors were permitted, the answer is all but a handful that had some known issues.

I do not understand what "evaluated against out-of-sample test data" means. If you mean was the AI TESTED, the answer is no.

I ask b/c it feels right to look deeper into how AI predictions fare in real life with an eye towards their performance based on factors/variables the models were NOT trained on. Given tens of thousands of pace lines the models are using to make inference, it seems to me an uphill battle trying to prove them right/wrong with even couple hundred manually tracked races for those variables they already considered during training. — Atakante

You are correct.
There is zero proof.

If memory serves, the chaos race variable/factor was NOT an input to the models so that's one candidate factor worthwhile manually diving into. Is it a safe assumption that race-level factors were excluded from training data? If not, is there a short list of other candidates or a shorthand logic to find others to manually "study"? — Atakante

Chaos = correct, but there was a fitness function involved.

Race Level: Nothing was left out.
___________________
The questions you have asked are actually pushing the envelope of need-to-know.

The AI is a very strong engine.
After my recent small sample test that was a day chosen at random, I am actually surprised that it performed so well.

I simply threw together a logical "system" based upon what has been discussed and it crushed the game. After going back and correcting a handful of races I had tagged, and correcting a few mistakes, manually assembled these statistics.



This is pretty amazing output.
DUTCHING: TOTE ODDS represents the actual dutched betting for the 55 races.

FLAT BETTING (A, A+) represents betting $2 on each A or A+ horse. IOW, betting only the horses that computed as having a $2.20 or higher $Net.

FLAT BETTING (BAD BETS) represents all the horses bet that were not A or A+ HORSES (according to the rules provided).

Very deep questions.

Not simple enough to answer in writing.

We'll have to discuss this in a meeting.
(Also, some will likely be proprietary.)

I believe the Grades (A+, D, etc.) were initially portrayed as ways to find who the whales are likely to bet on (mainly A+ and As) but I'm curious to find if anyone here really rely on those for their handicapping. It's still not clear to me what/how they contribute as opposed to say Magnitude and Rank. Any responses appreciated... — Atakante

Short version from my POV:
First, Grades are derived from pure math.

As Tom said, it is always the payoff from the "selected odds" times the hit rate.

The correct formula is (Odds x 2 +2) x Pct = $net.
"Grade" is also as Tom said; a function of the $net.
____________________
In a perfect world, All we'd need is to bet all the A+ horses and retire to wherever we want that has an internet connection.

Here's the problem:
$Net is a function of odds x hit rate and NEITHER OF THOSE IS ACCURATE!

Still, it is a good estimate of both probability and profit, but... honestly, it takes more.

BTW, Optimum% is probably a better measuring stick in the sense that you'd be better off betting more on a low odds A+ than a high odds A+.

While there is a time and a place for just using Grade, $Net, or Opt%, the game isn't that simple.

There are complications to the entire picture. At this morning's live play session we discussed a process that I call "Using Boxes" to build systems. In the coming days I will try to find time to expand on it.

Although this is not a great form - I'm sure many of us could do better...

Here it is if someone wants it to print out.

There's a potentially excellent thread going on "under the hood" that I decided should "have the hood open."

To that end, I've moved it where it can be seen.

Direct Link to that thread

↪Neal Freedman Hello Neal. I've been working on a strategy similar to Ram. First I have one (1) report that I use to pick contenders. Then in the Final Step I meld the two (2) Green "3" and "4" AI pacelines. For Turf and Tapeta, I meld a little different. Under GENRES I've been using ODS 63 with an Obj Wgt of "46". 46 will elevate a horse in this object versus using 50 or higher.
I think the key is being consistent, especially with the initial Prime AI melding. Dirt, sprints, routes, turf, track specific-- be consistent with what you use. I've found using the two (2) Green "3" and "4" AI pacelines works pretty well for dirt races. — Jim Pommier

This is the stuff that makes Super Stars!
Thank you!
______________
Just a heads up... make sure that you track the Genres by Object Name instead of numbers because the numbers are regenerated alphabetically as new genres are added.
______________
I'd like to describe what you are doing with BOXES.

You'll want to make your own, but here's a starting point.


This is meant to be a simple worksheet with 4 races per page.

• The top set is the explanation.
• There are 4 COMPONENTS to your approach, numbered 1 thru 4.
• "OOP" stands for "Order of Preference."

• The 2nd set is a race that you won.
• After the race is over, you use a highlighter to mark the winning horse.
• The Fail Box contains a GREEN 0 to indicate that none of the COMPONENTS failed.

• The 3rd set is a race that you lost.
• After the race is over, you use a highlighter to mark the winning horse - GREEN to indicate the COMPONENT that had the winner, and RED to indicate that the COMPONENT FAILED.
• The Fail Box contains a RED 3 to indicate the precise COMPONENT THAT FAILED.

Surprisingly, it was only the ones that actually had big payoffs, as I recall. LOL

Did you check out the Chaos5 races with A+ horses on Friday?

(I only looked at Friday.)

Just to be clear... Last November's races were not included in the training set. The AI was trained during the summer and for Graded races I went all the way back to 2013.

Must be some good things! — Mark

Truly.
They're working on how to win and I'm working on documentation.

Where's everyone at? — Mark

Lots going on under the hood.

Dave Schwartz, some questions for you:
What do you consider deTerminator's best feature (AI Line, AI Best Horse Pct, $Net, ...) to be, and why?
What is the easiest way to apply it to generate profits? — Tom Atwater

Considering what I've been working on for the last week, I'd have to say:

"THE BEST FEATURE" is NONE OF THE ABOVE!

Brian and Tony have the prerelease. Maybe they can give you a first impression.
Ping @ponyplayer & @Tony Kofalt

I hope to have it finished by tomorrow night.

35 fixes it and is in your DropBox.

It's working but I left in a stop.
I'll explain all of this tonight.

But Upgrade coming to get rid of the UNNECCESSARY stop.

Dang.
Upgrade in 15 minutes

Georgeontherun-- good name for a race horse. — Jim Pommier

LOL
I like it!

As I always say when new people come over...
"$500 for the bird and the cage."
They say that they're not looking for a bird.
Then I say, "How about I give you $800?"

It is not.