I just finished the most important section in the entire program.

This is where the user gets to enter THEIR handicapping for each horse.

DeTERMINATOR is designed to be an INTERACTIVE ADVISOR in the handicapping process. That means you must be able to enter your own opinions, compare them to the AI, and even COLLABORATE with the AI.

Well, maybe "finished" is the wrong word because I will be writing collaboration functions for several more days.

But getting the data in was really a big deal.

Think of it this way... Calculations depend upon probabilities. But the only handicappers I know of who actually work with probabilities are those using a very rigid tool that says, "Horse #1 is 27%."

Get this...
Imagine you have an 8-horse field.
You have strong opinions about the 1,2 and 3 so you use the ODDS column:
1: 2.0/1
2: 3.0 /1
3: 2.5/1

You think the 4 & 5 are like 4th choices, so you use the CHOICE column.
4: 4
5: 4
(Note: If you entered actual odds for a horse, it will not use the CHOICE column. Instead, it locks your odds in.)

You think everyone else is just a crazy longshot, so you leave the 6,7,8 blank. The engine will take care of those.

That's how you get your picks in.
Then the engine converts the ODDS and CHOICES to percentages.

Then the fun starts.

BTW, CHOICES column goes from 1 to 5, but you can also enter an "8."
if you enter an "8" in the CHOICE it has the AI generate the percentage for that horse based upon what the public opinion SHOULD BE based upon the training process.

Remember that any horse you enter odds for will never change. Only the others will change - based upon the strengths you entered.

There are more tools.

I'm really doing well with it.
Hope to start demos by the end of the month.

The idea is to make it SIMPLER instead of more complex.
(BTW, nobody actually USES 4,000 factors, right? Or even 400.)

You accomplish that by constantly working to slim down your data set.

First, you have to actually HAVE a data set.

No.
While winning margin does make a small difference, it is not what people would expect.

Well, perhaps I should ask what you'd expect it to mean first.
LOL

I don't teach coding.
But I am willing to discuss strategy for accomplishing what you have in mind.

One of my favorite movies.
Lancey Howard!
vs
The Kid

Can't comment more specifically.

1. Did the mechanic after several hours of practice figure out how to arrange the cards into the shoe?
Not knowing exactly how you would play each hand-- hit(s), hold(s), split(s), etc-- seems impossible. — Jim Pommier

It would be impossible - unless it was a completely stacked deck. That's known as a "Cold Deck" in the vernacular, and demands a floor man be in on it. (Somebody has to switch the previously prepared deck in and then fake the shuffle.)

But, no. That's not how it is done.

He arranges sequences of cards that wind up serving his purpose (i.e. win or lose).

BTW, must be a face-up game.

It's all in how you pick the cards up.
Once he has a sequence built into the discards, you top the section off with a "signal sequence." Think of something like Ace of Spades, King of Hearts, Jack of diamonds. (A little obvious - only to illustrate a point.)

BTW, I could never do this myself. When the guy showed me I was shocked.

A common way of arranging the cards would be like putting together 2-card combinations that add up to 10 or 11. Thus, once the sequence is live, when you see an "8" you know that the next card is likely to be a 2 or 3. When you see a "7" it is reasonable to expect a 4 next.

So, if you need a small card to prevent breaking, you hold that next small card on top until it's your turn.

Again, this guy was really masterful. This would be above literally all but a half-dozen or so dealers in Las Vegas in that era. (Of which, I knew 2.)

2. Was he taking 1 card from the shoe each time or was he able to slide out 2 with just the one hand? Again, watching the shoe closely seems it would be easy to see more than 1 card coming out of the shoe. — Jim Pommier

Dealing seconds out of a shoe is not difficult. The difficult part is knowing what the top card is so that you have a reason to store it. See #1.

A function of field size and perhaps the style of track in terms of angle, i.e. like Hastings in B.C. is like a bull ring. — Charles Lucas

Surprisingly, not a function of field size except with the tiny tracks (i.e. TIL, SAL, etc.)

But at those tracks, top trainers dominate so strongly that everything points to them anyway (i.e. speed, top jockeys, pedigree, etc.)

Yes, pedigree.
It's a sleeper factor I found and it matters in cheap claimers for OLDER horses!.

Really great questions.
________________

I did a little reading on "Agents" in AI". "Each of the 8 Agents has their own approach to the game". Is each approach somewhat predetermined by the AI programming? — Jim Pommier

The goal of the AI engine is to develop a system that allows each agent to become their own person, so to speak.

Imagine you're going to teach...
... a brand new horse player how to handicap. What do you teach him?

• You teach what you know and believe.
• This becomes his own personal starting point.
• He will begin with the factors you gave and proceed from there.

While he can - and most certainly will - eventually change how he handicaps, his roots will always be in his beginnings.

Will one Agent look at say Pace Shape while another looks at FTS's? Then each one builds from their experience in their particular approach? One agent may like horse #1, while another may like horses #2 and #3?

This is a very simplistic way to explain exactly how it works.

____________________

For example-- "Is there a vulnerable favorite in this race? Which horses' figure to improve or digress? Will only some Agents review and respond to the question, while other Agents do not based on their AI approach to handicapping? Thanks.

Again, precisely right.

There is one significant difference between HUMAN handicappers and AI AGENTS:
While humans are dedicated to improvement, there is a natural tendency to want to continue doing what they are doing now.

To some degree, we just naturally defend our current approach and resist changing it beyond minor improvements.

Agents are not encumbered by such complications.

HOWEVER, the will always be trapped (to some degree) relatively near their starting points; their POINT OF ORIGIN.

____________________
I've designed the 8 agents with HANDICAPPING PERSONALITIES that will have VERY GOOD HIGH contender hit rates.

There is one such agent whose approach will likely have the very highest contender hit rate. However, he will have a difficult time hitting price horses.

Each agent's tendency will be to stay somewhat close to his roots in contender selection, but his (or her) actual handicapping is completely free form in the sense that they have ALMOST ALL of the 3,000+ factors available to them.

Knowing that there's a vulnerable favorite, I can then pass the race or look at other entries — Jim Pommier

Yes, this is valuable information.

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? — Jim Pommier

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.

That's a big step towards understanding what people want and could use.
Thank you.

I'm trying to design something that allows for almost free-form notes.

I think that the key is to have them readily available - so you don't have to go looking.

Do you keep the notes?

Good fortune to you with this.

Hi Dave, decided to join your Horsestreet blog site and I was just perusing. One of the questions I wanted to ask in terms of the HDW website and it's data is what Past Performance Data do you use. Is it Bris, Equibase, Drf? Also, one of your factors, or can or is this built-in, is horses in cycles tend to show either losing speed on the turn or from 2nd Call to the stretch as negative length loss. I tend to look for horses that have gone 1-2 races whereby this happens particularly if they are chasing and not close to the lead. I have had sucess with these types or will wait to they show improvement in their running. — Charles Lucas

1. Only HDW data.
Cost is $142 per month for all tracks with results.

2. Yes, there are several factors that refer to that.
Things like positions/length lost between calls.

Those are shown in several formats.
Last race
Per race in a graph
Time-decayed
Averages
_______________
Truth: Our suggested approach to speed already recognizes that the 2nd call (i.e. EP) is not as important as it is made out to be.

Of the 52% of all winners that are within 1 length at the 2nd call, 44% of those were already on the lead at the 1st call.

IOW, the 2nd call challengers only account for 8% of all winners and their prices are low.
________
"Do you know what you call a horse who makes a big move on the turn?"
The odds-on favorite.
________

WOW!
Congratulations, @Conley!

Nice article.

&

Tony is correct.
The challenge is that a $16k claiming race may really be anything from a $10 to a $25k race. Usually plus/minus 2 levels.

That's why I recommend using a strength of race rating instead of par.

My new software is really getting closer now.
It will blow your mind with stuff just like what you're talking about.

I guarantee that you've seen nothing even approaching it before.

Stay tuned.

Do you mean "Have ever run to today's par?"

I like this idea.

I am aware of no such tool, but it could be accomplished with some code in my future software - target for release is a few weeks away.

I pretty much pass nothing. LOL

I play races differently, but I believe there is always a way to leverage a bet out of a race.

Must admit, that this position is getting tougher to maintain.

Pete,

They're not coming to HSH. :)