Why has technology left handicapping in the dust?

[Dave Schwartz]

With all due respect to the software authors who still frequent this board (Dave Schwartz, Jeff P, Ted Craven, etc.), why does it seem that nothing new has come out in YEARS...

It probably isn't worth it any more to create a new software program anyways with the declining market for such things. Probably more of a hobby thing for someone good at programming.

Anything customer facing that has to do with racing data in North America, Equibase feels that is their data you are developing off of. Therefore there is no development community in racing or it's extremely small. Literally a handful of people in the entire country who have intimate knowledge of how racing's data works.

I've worked for many different stakeholders in racing and the data costs are insane for everyone. There is not a lot of room to breath if you have an idea and want to develop something. Years and years of not offering machine readable licenses to developers at reasonable development prices has basically put a big fat sign up saying "Don't bother".

All of these are right on the money.

(note: And if you think data is expensive here, you should see AUS. To get csv data comparable to what most would consider 2nd tier here, would cost you over $2,000 USD per month!)

Despite all of that...
Almost 4 years ago I committed to building something really new and innovative.

In that time I've spent almost 1,200 hours developing new stuff, but always I become convinced that it was just too complicated.

Early last summer I started completely over again - for the 4th time. I committed to simple, simple, simple.

The challenge is that a Real Solution is, by its very nature, COMPLICATED!

Tossed it for the 5th time in October and began again.

We all know that the damn game is hard.
That's why so many people are resigned to enjoying the play rather than the results.


Incarnation #6 is a different beast.
Way different.

First, by default it uses NO DATA.
What? How is that even possible?


As soon as I have a solid beta, I'll put up some screenshots.

The first (very early) beta goes out to one person tomorrow.

The Handicapper's Notebook is coming.
Really.

[cj]

Originally Posted by AndyC

How big is the market for the technology desired? I doubt that it is big enough to entice the investment necessary to produce such a product.

Unless, as I stated, you are the one using it to extract money from the pools at a greatly discounted price. That is the only way to recoup the cost.

[PaceAdvantage]

I'm not even talking about some money making black box system.

I'm talking about giving me the proper tools to discover and create my OWN money making method.

In 2023.

On my hardware that I built in 2021 (so it's already pretty obsolete...LOL)

[ranchwest]

Originally Posted by PaceAdvantage

With all due respect to the software authors who still frequent this board (Dave Schwartz, Jeff P, Ted Craven, etc.), why does it seem that nothing new has come out in YEARS...

Yes, there's been stuff like Stats Lens...but where is the breakthrough ROBUST handicapping package that lets me do essentially ANYTHING I want with the data?

As some might know, I've gotten heavily into trading equity futures...and there are lots of applications out there for this type of thing that let you do basically anything...including creating robust automated trading systems, as well as in-depth research, backtesting and optimization of trading systems...some of them also include AI options that actually will CREATE AUTOMATED TRADING SYSTEMS FOR YOU based on parameters and indicators you deem important.

Why has technology seemingly passed the handicapper by? Where is the package that EASILY allows me to create and test handicapping methods based on PP data I deem important and using betting methods I define? Then test it on years of racing data and also allow optimization of the method when and where possible...and most importantly, in a relatively user-friendly manner?

One would think racing would be the perfect place for this kind of development and innovation. There is obviously TONS of back data available (not cheaply for the most part, but that's another story).

Where endeavors exist that allow people to make money (like pari-mutuel pools...don't laugh...but it IS possible), you usually see no end to the offerings available to those willing to dive into the deep end of the pool. Not so for the handicapper.

Is the potential user base that nonexistent that nobody thinks that it would be worthwhile to pursue the development of these new generation of handicapping tools? I find that kind of hard to believe, actually, even though I know more than a few of you are going to tell me that's precisely why there is an absolute dearth of new and exciting offerings on the application front.

Two reasons:

1) It takes a long time to develop, driving up the selling price.

2) Horses don't understand AI.

[ScottJ]

Originally Posted by PaceAdvantage

Then how do you explain stocks?

What is the technical analysis data for stocks?

A Bloomberg workstation costs on average $3000 ... PER MONTH ... for access to real-time data across equities and bonds. Start there and ask yourself how many horse racing workstations would sell at that price. The reason that analytics innovation exists in trading is that there is the financial backbone to support the investment in IP, research, development, testing, and support.

[ScottJ]

Originally Posted by PaceAdvantage

I'm not even talking about some money making black box system.

I'm talking about giving me the proper tools to discover and create my OWN money making method.

In 2023.

On my hardware that I built in 2021 (so it's already pretty obsolete...LOL)

Bring me back to the point - is that infrastructure worth $3000 per month to you? That's what it costs ... in 2023 ... to gain access to the data science analytics that you are talking about in the trading world. And that does not include proprietary models that each cost millions to produce.

[elhelmete]

I always thought that poor quality and inconsistently-collected primary data played a role.

[classhandicapper]

Originally Posted by o_crunk

Meanwhile, privately there are plenty of home spun tools and others who will pay privately for exactly the tools they want.

Personally, I like it that way.

The more people that have databases and develop their own tools, reports and metrics via research, the more efficient the pools are going to get and the harder it will be to beat the game.

When I started you compiled trainers stats manually, had to keep piles of DRFs if you wanted to look at the PPs of the other horses in a back race up or see back past the last 10 races, you had no idea about many trainer changes unless you kept track, you had to make your own pace and speed figures etc…

People keep asking for more and better data and tools, but that’s what’s making the game tougher and tougher over time. The last thing we need is to make tech even more widespread or to take the technology to an even higher level.

I put a ton of work into creating my database and all reports and metrics I generate from it. Why would I want everyone to have what I have or better?

I’m sure Andy Beyer made a lot more selling books and figures to the DRF than he could from gambling, but giving away the methodology certainly didn’t make it easier for him and others that had figures to win gambling using speed figures.

[Bustin Stones]

Back before Warren Buffett became an investment legend, he sold tip sheets at Ak-Sar-Ben in Omaha as a teen. But, he abandoned that when he saw the limitations. How could he build wealth even if he knew how to take down the entire pool? The pools are not big enough to finance those dreams.
So he turned to a very large pool of wealth in American stocks.
The value of solving the racing question using AI won't financially support spending capital to solve it beyond what has been offered. Whatever programs have been developed are commensurate with the opportunities to profit.
Compare that to what has been spent in programming to profit from the financial markets.
It'd be like using a surgical laser to cut board at Home Depot. The price of the tool has to make sense in relation to the job.
Perhaps someday in the future, Americans will abandon casino games, realizing they cannot beat the house, and turn to racing. Once the well of larger pools can support the investment we could see more sophisticated tools.

[PaceAdvantage]

Originally Posted by ScottJ

A Bloomberg workstation costs on average $3000 ... PER MONTH ... for access to real-time data across equities and bonds. Start there and ask yourself how many horse racing workstations would sell at that price. The reason that analytics innovation exists in trading is that there is the financial backbone to support the investment in IP, research, development, testing, and support.

Who's talking about that?

I use NinjaTrader...you can buy it outright for about $800 or pay about $200 a quarter to use it for as long as you want. And it does everything for me that I wrote about in my original post.

$800 is in the ballpark for the higher end handicapping programs these days (the few that still exist).

And NinjaTrader isn't even one of the best trading platforms out there for the non-pro/retail trader.

I'm talking about tools for a retail player, just like tools for a retail trader.

A Bloomberg terminal? Why would you think that would be an appropriate comparison for this discussion?

[PaceAdvantage]

Originally Posted by ScottJ

Bring me back to the point - is that infrastructure worth $3000 per month to you? That's what it costs ... in 2023 ... to gain access to the data science analytics that you are talking about in the trading world. And that does not include proprietary models that each cost millions to produce.

You have not grasped my point at all.

You're talking about professional end systems.

I'm talking about what is available for the RETAIL TRADER today...which certainly doesn't cost $3000 per MONTH or whatever the heck you're talking about.

And which certainly outstrips what is available for horseplayers by comparison in terms of technology, ease of use and functionality.

[classhandicapper]

It sounds like you are asking for an extension of the trainer pattern tool of Formulator that you could apply to other factors in the PPs. That was a project that was actually outlined at DRF 4-5 years ago. It would have been called “angle finder”. You would have been able to program and test various angles and also program it to give you a list of any horses that fit that angle that were running that day. It never got off the ground, most likely due to lack of resources.

[thaskalos]

Isn't Betmix something like that?

[thaskalos]

Originally Posted by thaskalos

Isn't Betmix something like that?

Just guessing...I never used it.

[steveb]

Originally Posted by cj

The people with the best software aren't selling it, they are the CRWs.

CRWs?