Track Variant

[GameTheory]

Originally Posted by TrifectaMike

Let me add the following to your interesting post.

In a response to a question on machines and human minds, J. von Neumann replied: " You insist that there is something a machine cannot do. If you tell me precisely what it is that a machine cannot do, then I can always make a machine which will do just that!"

J. von Neumann is generally regarded as one of the greatest mathematicians in modern history. He made major contributions in quantum mechanics, hydrodynamics, game theory, set theory, computer science, numerical analysis, statistics, and functional analysis.

That's the problem, we often can't say precisely because that would mean describing a subconscious mental process precisely, and we just don't understand exactly what's going on in our own heads. What's interesting is that early artificial intelligence researchers thought it would be fairly straight-forward to model the brain, and then they realized they really had no idea how the brain actually worked. AI was a big boon to human brain science probably more than the other way around after it put in such stark relief what we didn't know.

[thaskalos]

I keep on reading that the human brain is more powerful than any computer...

Is that not so?

It has long been debated whether handicapping is an art or a science.

In my opinion, it is BOTH...an art AND a science.

The computer can undoubtedly handle the scientific aspect of it...but the artistic side? I have my doubts...

Can a computer be depended upon to create masterpieces in art or literature?

How then can it be expected to artfully put together the great puzzle known as "handicapping a race"?

[CBedo]

My opinion is that a computer can do anything I can do with regards to handicapping...if I can accurately communicate to it what I'm doing. That's the problem. Sometimes much of what we think of as "art," "hunch," or "gut feeling" is really just our brain's way of black box processing all our past experiences. If we're not smart enough to know what we're really doing, how in the heck do you tell a computer to do the same thing?

[TrifectaMike]

Originally Posted by thaskalos

Can a computer be depended upon to create masterpieces in art or literature?

How then can it be expected to artfully put together the great puzzle known as "handicapping a race"?

A thorough knowledge of Bayesian thinking, a dozen or so Bayesian Belief Networks, and a couple of Bayesian learners and I think you might arrive at a different conclusion.

Mike (Dr. Beav)

[thaskalos]

Originally Posted by TrifectaMike

A thorough knowledge of Bayesian thinking, a dozen or so Bayesian Belief Networks, and a couple of Bayesian learners and I think you might arrive at a different conclusion.

Mike (Dr. Beav)

If only I had father time on my side...

[CBedo]

Originally Posted by thaskalos

Can a computer be depended upon to create masterpieces in art or literature?

How then can it be expected to artfully put together the great puzzle known as "handicapping a race"?

Infinite monkey theorem?

[Tom]

Originally Posted by TrifectaMike

A thorough knowledge of Bayesian thinking, a dozen or so Bayesian Belief Networks, and a couple of Bayesian learners and I think you might arrive at a different conclusion.

Mike (Dr. Beav)

The where are the products for us to evaluate? The materpieces of art and music? the records of wagers made before the race?

[TrifectaMike]

Originally Posted by Tom

The where are the products for us to evaluate? The materpieces of art and music? the records of wagers made before the race?

Fortunately, for me, there are no products to evaluate. I know of only three people aside from myself that use a probabilistic framework based entirely on Bayes probability in horse racing. It is not a simple process of putting together a piece of software. It is way of thinking. And the process is continual.

My metrics are constantly evolving.

Mike (Dr Beav)

[Tom]

OK, we can enjoy the art and symphonies while we wait.

[PaceAdvantage]

Originally Posted by thelyingthief

What the hee-haw boys need to ask themselves is: Why, when a player says he makes profit, do you feel it necessary to slam him? Could it be, YOU don't make money? And perhaps YOU lack the faith in yourself to do so, and must deny the ability in others?

As for PaceAdvantage, I'm sorry, I'm not to blame that I have a larger package than you.

tlt-

You sir, are in dire need of a sense of humor. Your name is "thelyingthief" is it not? My reply was a mere little joke on your user name. Lighten up Francis.

And who is slammed when they say they make a profit?

The only people I see slammed around here are those using pace/speed figs and/or computers. They're always being told (complete with a heaping serving of arrogance) how what they're doing can't work and why. And that's when the fights start...

[thelyingthief]

I ask you to remove the very post you quote. I ask you again, please remove that post. It is out of character for me.

tlt-

[PICSIX]

Originally Posted by Cratos

Mike, I understand your question, but plain and simple the track variant under current construction and use is bogus. There is no way that you can measure anything with all variables. At best the so-called track variant is a non-parametric indication of how the track is playing with respect to speed, but in the end you can do just as well without it.

I have worked very hard over the years using the principles of physics to develop the Surface Speed Resistance Value (SSRV) which is measured in one thousandth of seconds because the smallest winning margin which is a nose, is estimated to be .009 seconds.

From Michael Pizzola's Handicapping Magic; "You may notice that in calculating the PBS (Pace Balanced Speed) rating, I have made no attempt to adjust the numbers for daily variants or track to track adjustments.....I've tried various methods of adjustments, and for a period some work better than others. To keep track of it all, you would need a computer with the ability to do simulations and automate those comparisons. Bottom line for me is that they don't make all that much difference in the final betting decision....Save yourself the headache."

[PICSIX]

Originally Posted by GameTheory

Don't get me wrong, I'm telling people all the time that I can program things for them they don't think can be programmed, but some things you just can't, or it would be so complicated and cost-prohibitive to not be practical.

I have also found when making programs for horse players, the "mechanical method" they think they have isn't as mechanical as they realize, and when we sit down to turn into a set of rules the computer can implement, they are using a lot more human judgement that they thought and even though they win with the method in real-life, we can never get the computer to do so well because they are throwing out races and horses for intuitive reasons (often without realizing it). But the computer needs a reason for everything.

Agree 100%! Even the simplest (believed to be simplest) selection process is not so simple when you set down and try to create a set of rules. Even more complicated is the actual betting process....In my opinion this is more of an art than anything else.

[Capper Al]

Originally Posted by PICSIX

From Michael Pizzola's Handicapping Magic; "You may notice that in calculating the PBS (Pace Balanced Speed) rating, I have made no attempt to adjust the numbers for daily variants or track to track adjustments.....I've tried various methods of adjustments, and for a period some work better than others. To keep track of it all, you would need a computer with the ability to do simulations and automate those comparisons. Bottom line for me is that they don't make all that much difference in the final betting decision....Save yourself the headache."

Like I've been jumping in this thread and saying, the feet per second guys seem to do well without a variant. It is better to have one than not. The vendors and suppliers of racing past performances do a good enough job. Focus on other handicapping factors.

[Hoofless_Wonder]

Originally Posted by thaskalos

I keep on reading that the human brain is more powerful than any computer...

Is that not so?

Did you see IBM's Watson computer spank the competition on Jeopardy earlier this year? Not everyone can afford a system with almost 3000 power7 cores and 16 TB of RAM, but the software and AI systems are getting better and better. Even "smart" analytics is really in its infancy, which much of the reason having to do with the limiting cog in the process (humans) that try to apply it can't wrap their brains around it.

Interesting thread on the track variant and how to calculate it. As a recovering meteorologist, it reminds me of attempting to solve some of the problems in fluid mechanics. For example, when dealing with 26 equations and 26 unknowns in a model trying to predict air parcel movement, a lot of those numbers are very small and approach zero, and get thrown out to simplify things. Over time, however, they influence model projections drastically, which is why the weather forecast for the next two hours tends to be more accurate then the one two days or two weeks out. For the ponies, this often translates into the data point of a race four months ago doesn't carry as much weight as the race 14 days ago.

Another concept I don't see being mentioned is the difference between accuracy and precision when calculating variants. Does anyone take that into account? I wouldn't argue with some of the math or physics applied to the calculations, but isn't the data's accuracy somewhat suspect? Even if the track's timing equipment is working properly, I'd think that any calculations other than the leader's splits (i.e., lengths behind, pace adjustments, class and race ratings, etc.) would introduce quite a bit of error....