Computer handicapping

[Capper Al]

I've been playing with about 12,000 lines of my own C++ code for handicapping. It takes in a BRIS Multicaps file and outputs a csv file which is upload into one of three Excel spreadsheets. All the spreadsheets have about 20 tabs to them each. The preliminary tests are looking very good. Now when I reflect back at all that is going on with the code and the spreadsheets, I can't see how a paper and pencil capper can compete with a computer. Computers can look at so many aspects of racing all in a second that it boggles the mind.

[Dark Target]

This should generate some debate.

May I ask, if you know how to code in C++ why are you using Excel for anything?

[Turkoman]

I've been playing with about 12,000 lines of my own C++ code for handicapping. It takes in a BRIS Multicaps file and outputs a csv file which is upload into one of three Excel spreadsheets. All the spreadsheets have about 20 tabs to them each. The preliminary tests are looking very good. Now when I reflect back at all that is going on with the code and the spreadsheets, I can't see how a paper and pencil capper can compete with a computer. Computers can look at so many aspects of racing all in a second that it boggles the mind.

Maybe the computer will make the job easier, but you still need to execute, just like in baseball, when a team needs a good bunt from a batter, with men on base in the late innings of a game. I'm a paper and pencil handicapper and doing quite well, and only use the computer to save some stats. After several years, you realize that what you actually need to do is simply follow the same methods that have worked for you in the past, and not change your mind just before the horses leave the starting gate.

[dansan]

as DR.Phil would say Capper how 's that workin for you

[windoor]

As a former pen and paper guy of 40 years I can relate.

Some years ago, I had at least 1000 DRF's with the morning Sun newspaper result charts stabled to them in my living room. A fire hazard to say the least.

Very long hours were needed to find a factor that had some merit, and that was just for a few tracks.

Fast forward to today. I have every race that has been run with results on my PC for the year 2011. Close to 50 thousand races I think. It was a huge undertaking to get my factor set into the database so I could run the queries I wanted, but it was worth it.

What would have taken weeks if not months to investigate, I can now do in minutes. Even complicated formulas for a spot play with a lot of "If" and "Or" statements to rearrange and investigate, only takes of few days. Looking at the big picture over time is a real eye opener.

That great play that has been killin it the last few weeks, looks a lot different when you can see how it does for a full year. Most are just anomalies.

I never thought I would say this, but I have come to trust the selections the program spits out and will soon have my "black box" up an running for all playable tracks.

Though I don't think the research will ever end, finding the gold nuggets has never been easier.

Regards,

Windoor

[Actor]

This should generate some debate.

May I ask, if you know how to code in C++ why are you using Excel for anything?As delivered, the files are not Excel friendly. C++ is not people friendly. Use C++ to crunch the files into an Excel friendly form. Excel is very people friendly (to certain people). :)

[Capper Al]

This should generate some debate.

May I ask, if you know how to code in C++ why are you using Excel for anything?

I use Excel for the presentation layer and final computations. Multicaps does not have post statistics. Post statistic are entered last minute by hand from the BRIS Ultimate program.

[Capper Al]

Maybe the computer will make the job easier, but you still need to execute, just like in baseball, when a team needs a good bunt from a batter, with men on base in the late innings of a game. I'm a paper and pencil handicapper and doing quite well, and only use the computer to save some stats. After several years, you realize that what you actually need to do is simply follow the same methods that have worked for you in the past, and not change your mind just before the horses leave the starting gate.

I like paper and pencil systems. They are fun and can be profitable, but not too dynamic. The program can change with the circumstances of a race in a flash.

[Capper Al]

as DR.Phil would say Capper how 's that workin for you

So far pre-beta testing is going well.

[Capper Al]

As a former pen and paper guy of 40 years I can relate.

Some years ago, I had at least 1000 DRF's with the morning Sun newspaper result charts stabled to them in my living room. A fire hazard to say the least.

Very long hours were needed to find a factor that had some merit, and that was just for a few tracks.

Fast forward to today. I have every race that has been run with results on my PC for the year 2011. Close to 50 thousand races I think. It was a huge undertaking to get my factor set into the database so I could run the queries I wanted, but it was worth it.

What would have taken weeks if not months to investigate, I can now do in minutes. Even complicated formulas for a spot play with a lot of "If" and "Or" statements to rearrange and investigate, only takes of few days. Looking at the big picture over time is a real eye opener.

That great play that has been killin it the last few weeks, looks a lot different when you can see how it does for a full year. Most are just anomalies.

I never thought I would say this, but I have come to trust the selections the program spits out and will soon have my "black box" up an running for all playable tracks.

Though I don't think the research will ever end, finding the gold nuggets has never been easier.

Regards,

Windoor

Agree. And good luck to you. I'm not to the database stage of the game. I'm checking for bugs and double checking results. It is a big commitment to getting a system going. Figuring on the $40.00 an hour for a programmer, posted elsewhere in the forum, I have spent up to $100,000 of my time to write this program.

[Capper Al]

As delivered, the files are not Excel friendly. C++ is not people friendly. Use C++ to crunch the files into an Excel friendly form. Excel is very people friendly (to certain people). :)

You got it. I started off to have C++ write a report and found it not friendly and ugly-- Old school programming. Excel does a nice job with the presentation and allows for last minute tweaking of numbers. Some formulas are best computed in C++ while others are better computed in Excel.

[cj]

You got it. I started off to have C++ write a report and found it not friendly and ugly-- Old school programming. Excel does a nice job with the presentation and allows for last minute tweaking of numbers. Some formulas are best computed in C++ while others are better computed in Excel.

I use C++ and output in HTML. You can't beat web page presentation.

[jerry-g]

I am told that technology will advance so fast in the next 5 years that it will make todays stuff look like the dark ages. I dread it because I don't have any 5th graders around the house that can help me understand it. I think programing will be done in chunks which you download from a web site and put together a program in near warp factor six speed. The billfold companys will go out of business as our money will be in our finger prints. I dread things going so fast we can't keep up, only the younger generation will benefit.
:confused:

[classhandicapper]

Now when I reflect back at all that is going on with the code and the spreadsheets, I can't see how a paper and pencil capper can compete with a computer. Computers can look at so many aspects of racing all in a second that it boggles the mind.

I used to be a computer programmer for a living. Unfortunately, I don't have the C++ skills to develop any kind of handicapping programs. I'm not even sure I want to turn playing horses into a job. However, I was just thinking the other day that my greatest hope for the game getting easier for me is that more and more money is being pushed through the windows by computer handicappers using "similar" methodologies and screens.

I may no longer have any unique information, be able to weigh all the main factors as well as people using staffs, advanced mathematics and screens to compute odds lines and spot plays in many situations. But I think I can understand some of the nuances, exceptions, etc... that fall through the cracks in some races better. If many of the computer guys share similar weaknesses and plays (and I suspect they do), those weaknesses will become the new fertile ground for profits for non computer guys.

[Dave Schwartz]

I may no longer have any unique information or be able to weigh all the main factors as well as people using staffs and advanced mathematics to compute odds lines. But I think I can understand some of the nuances, exceptions, etc... that fall through the cracks better. If many of the computer guys share similar weaknesses (and I suspect they do), those weaknesses will become the new fertile ground for profits.

Exactly!

This is precisely what one must do to be competitive. If you try to compete on the level of the "big boys" you will fall short.

You must do something different!

We must be like a small guerrilla-warfare army against a larger, superior force, we must specialize in what they are likely to miss because their model does not allow for it.

However, one common mistake today are the guys that say, "My brain is simply superior to their computer programs..." That's what Kasparov thought until Big Blue showed him who the sharpest tool in the shed was.

IMHO, anyone who thinks they are going to stare at a Racing Form for a few minutes and out-perform the competition in the long run is kidding themselves.


Actually, I think that is not a good thing, but it is what it is.


Regards,
Dave Schwartz

[classhandicapper]

Exactly!

This is precisely what one must do to be competitive. If you try to compete on the level of the "big boys" you will fall short.

You must do something different!

We must be like a small guerrilla-warfare army against a larger, superior force, we must specialize in what they are likely to miss because their model does not allow for it.

However, one common mistake today are the guys that say, "My brain is simply superior to their computer programs..." That's what Kasparov thought until Big Blue showed him who the sharpest tool in the shed was.

IMHO, anyone who thinks they are going to stare at a Racing Form for a few minutes and out-perform the competition in the long run is kidding themselves.


Actually, I think that is not a good thing, but it is what it is.


Regards,
Dave Schwartz

My knowledge and understanding has exploded since the mid 90s when I first started really breaking through on a more consistent basis. But my ROI has been steadily declining and I've actually adjusted my bet size down. 15-20 years ago, I had information that few other people had. Now I don't.

The only thing that's helping me now is that since I understand more than I used to, I find more of those cracks. But it's tough competing head on with equally bright (or brighter) people armed with computers and advanced math.

[serp]

As a computer/math guy I know no other way than to write my own programs which I feel gives me considerable capabilities that i might not otherwise have. With that said, there are certainly factors to every race that involve data that I do not have access or won't be modeled to use. I can easily see how someone who knows how to properly identify those factors could have an edge in certain situations.

My concern, for myself at least, would be that without using my scripts to pick which horses/races to bet on that I would not be as disciplined in my betting. I know that when I'm actually at a track I want to bet every race and I want to make picks based off emotions/fandom. It makes for a good story when I make a nice hit this way but I certainly don't make money over the longhaul that way.

I imagine there are people who specialize in these factors and are disciplined in their betting that can do well. I am not one of them though.

[jerry-g]

I believe Dave is really telling it like it is. Human thinking power out does super computer every time. We all know this is a game of being able to think for your self...not follow the hoards to the windows to bet false favorites. We are close to getting an edge now than ever before in history. For instance we can soft focus on the race to see who's come to the party and then review each horses resume to see how he fits in today's race. We can start at the beginning of a horses career and work upwards to see how the trainer is performing and whether or not he is placing the horse properly. We can see how the horse responds to that effort and if he is improving and is placed today to win. No computer can do that for us nor can we get it from anyone else other than our own thinking power. "Think and Grow Rich" by Napoleon Hill is a good book to help us with this. ;)

[thaskalos]

...However, one common mistake today are the guys that say, "My brain is simply superior to their computer programs..." That's what Kasparov thought until Big Blue showed him who the sharpest tool in the shed was.

I admit to being one of those guys...

I believed what I had read about the human mind being more powerful than any computer...and I welcomed "computer" opposition...both at the racetrack and at the online poker tables.

I remember reading a magazine article on the Kasparov vs. Big Blue battle...and the article was accompanied by a photograph.

The photo showed two men, sitting at a small table, ready to engage in a game of chess.

One man was clad in a lab coat, and was sitting behind a laptop computer...which was connected to a mega-computer, the size of an entire room.

And across from him sat the Great Kasparov...armed only with his pen and notepad.

The classic battle of man against mega-machine...and the machine was defeated time and again.

Tears rolled down my cheeks as I looked at this photo...and my chest swelled with pride for the mental prowess we possess.

But the machine came back with a vengeance...and the great man was finally beaten, and walked away shaking his head.

And, along with him, HUMANITY was beaten...HUMAN INGENUITY was beaten...and we were ALL left shaking our heads.

Sure...the scientists proclaimed that the creation of Big Blue was ALSO a testament to "human ingenuity" -- but that was small consolation...because it was not the same.

And I remember that, as a horseplayer, I could feel the ground underneath my feet shaking...because I realized that, ultimately, our battle at the betting windows would also come down to "man against machine"...

And if Kasparov couldn't prevail in HIS battle...how could we mere "mortals" be expected to prevail in ours?

[Capper Al]

I used to be a computer programmer for a living. Unfortunately, I don't have the C++ skills to develop any kind of handicapping programs. I'm not even sure I want to turn playing horses into a job. However, I was just thinking the other day that my greatest hope for the game getting easier for me is that more and more money is being pushed through the windows by computer handicappers using "similar" methodologies and screens.

I may no longer have any unique information, be able to weigh all the main factors as well as people using staffs, advanced mathematics and screens to compute odds lines and spot plays in many situations. But I think I can understand some of the nuances, exceptions, etc... that fall through the cracks in some races better. If many of the computer guys share similar weaknesses and plays (and I suspect they do), those weaknesses will become the new fertile ground for profits for non computer guys.

Time will tell. My app wasn't build on stats, although I do use them, it was build on horse racing logic. A similar style to what you say you do.

[Capper Al]

I use C++ and output in HTML. You can't beat web page presentation.

I'll have to get with you on this. Thanks for your help years ago when I first started this project. You helped me with reading in the BRIS files.

[Capper Al]

thaskalos,

I didn't like it either when Big Blue won. It seemed to be jab to human intelligence and creativity. I didn't think it was a fair challenge 256 processors against one man. Of course, at the teller's window my thoughts are different.

[Capper Al]

A computer is only a tool. It can be used in several ways. Although I'm a number guy too, my approach and coding are from a racing fan's prospective first not an academic statistical approach. I haven't gotten to the database part of the game yet. It's coming. But instead of saying this happens 1 out of 4 times so I'll pick this horse, my code says this horse has class, this horse has speed, etc. The 1 out 4 studies will follow. I'm a one man show, and this is a lot of work.

[Dave Schwartz]

Human thinking power out does super computer every time

Jerry, I think you misunderstood. I said that the human has no chance.

[Sinner369]

Jerry, I think you misunderstood. I said that the human has no chance.


I going to use an baseball analogy..........the computer can hit a fastball better than us humans..........but we can hit the curveball better than a computer!

[Dave Schwartz]

Sinner,

For now, you must move out of the realm of the physical. We are still superior in that realm, at least for awhile longer.


Dave

[thaskalos]

If a fifty year old, thirty-year veteran of the handicapping and betting wars were to decide to finally take the leap to computerized handicapping...what should his next step be?

[Robert Fischer]

things to consider

cpu vs man

-------
quality of information(data) for each rival
complexity
robustness
finiteness (and definition of contest)

[Dave Schwartz]

If a fifty year old, thirty-year veteran of the handicapping and betting wars were to decide to finally take the leap to computerized handicapping...what should his next step be?

Tough question. Probably call me, especially if it is you.

Because, with all your contributions to this forum, I would be forced to award you a free copy of my software.

Dave

[thaskalos]

Tough question. Probably call me, especially if it is you.

Because, with all your contributions to this forum, I would be forced to award you a free copy of my software.

Dave
Thanks very much for your kind words and generous offer Dave; expect to hear from me in the next day or two.

But I must warn you; it violates my principles to take advantage of the kindness and generosity of friends...

[windoor]

Thaskalos:

If a fifty year old, thirty-year veteran of the handicapping and betting wars were to decide to finally take the leap to computerized handicapping...what should his next step be?


Other than using Excel since 1996 to keep records, I knew next to nothing about data bases a year ago. I still don't know much, only how to use it.

I use the data only to verify whether or not a angle or spot play can be successful over time and at which tracks, Distance, class, age, sex, surface and time of year. I call it "the seven" and it is the backbone of what I do.

Once you find something that works, it just easier to let the program spit them out on a daily basis. No emotion, No second guessing, No looking at the "D" Horse that almost qualifies and going off at great odds. Oh so tempting.

If you would like to know what I use, PM me and I'll give you a shopping list.

Regards,

Windoor

[dnlgfnk]

The flow of this discussion could leave the non-tech handicapper in a state of depression. One thing seems true however, whether it be the intuitive specialist of the past or the high-tech, comprehensive number cruncher of today. Both were/are generally obsessed with the finish of races.

I run the race in my mind's eye according to the probable public odds, weighing the personal, very detailed and subtle trip factors with the most predictive value according to my experiential database that would allow for the field to run to it's public odds. This usually begins with the early leaders, comparing individual matchups from best speed to stalkers. That hypothetical running is compared to a horse's past races/trips which engineered today's probable odds.

Then I replay the race inversely according to probable public odds, upgrading higher odds horses where relevant factors allow, while downgrading favorites as personal, historically reliable factors allow, again beginning by matching up likely pacesetters and proceeding through the field.

Hopefully holding both runnings simultaneously in mind, I attempt to assess the true odds of the combatants. I do owe the effort to accord the favorites more respect, to the "Contingencies" article where Bill Benter discussed "tote smoothing".

Any predictive value given a factor by a program, has merely an accidental connection to the dynamics within the running of a race, which apparently can produce patterns of similar data, and which it tries to describe with a number. I would rather deal with the reality itself. I do cede the point that the technical approach solves the problem of the limitations of human stamina.

[Dave Schwartz]

http://www.horsestreet.com/BBSImages/PickingWinnerAtDmr.JPG

The checkered pencil pusher in action at Del Mar.

[Capper Al]

The flow of this discussion could leave the non-tech handicapper in a state of depression. One thing seems true however, whether it be the intuitive specialist of the past or the high-tech, comprehensive number cruncher of today. Both were/are generally obsessed with the finish of races.

I run the race in my mind's eye according to the probable public odds, weighing the personal, very detailed and subtle trip factors with the most predictive value according to my experiential database that would allow for the field to run to it's public odds. This usually begins with the early leaders, comparing individual matchups from best speed to stalkers. That hypothetical running is compared to a horse's past races/trips which engineered today's probable odds.

Then I replay the race inversely according to probable public odds, upgrading higher odds horses where relevant factors allow, while downgrading favorites as personal, historically reliable factors allow, again beginning by matching up likely pacesetters and proceeding through the field.

Hopefully holding both runnings simultaneously in mind, I attempt to assess the true odds of the combatants. I do owe the effort to accord the favorites more respect, to the "Contingencies" article where Bill Benter discussed "tote smoothing".

Any predictive value given a factor by a program, has merely an accidental connection to the dynamics within the running of a race, which apparently can produce patterns of similar data, and which it tries to describe with a number. I would rather deal with the reality itself. I do cede the point that the technical approach solves the problem of the limitations of human stamina.

I'd go dizzy doing all this. Seriously, I have to remember to read the past performance. So much of the information that I'll need to handicap has already been digested for me by my app.

[Capper Al]

things to consider

cpu vs man

-------
quality of information(data) for each rival
complexity
robustness
finiteness (and definition of contest)

There is one sobering fact a comprehensive programmer has to realize, and this happens more than we'd like to admit. Let's say we have an eight horse field and the computer rates them 3, 5, 1, 7. And then the class or speed of the field #3 scratches. Us humans would expect the computer would then pick 5, 1, 7 since the 3 was scratched. It ain't necessarily so. It might now come out 1, 5, 4, 7.

[Capper Al]

Tough question. Probably call me, especially if it is you.

Because, with all your contributions to this forum, I would be forced to award you a free copy of my software.

Dave

You celebrity bloggers sometimes go too far with the XOXOXO. :=}

[Robert Fischer]

There is one sobering fact a comprehensive programmer has to realize, and this happens more than we'd like to admit. Let's say we have an eight horse field and the computer rates them 3, 5, 1, 7. And then the class or speed of the field #3 scratches. Us humans would expect the computer would then pick 5, 1, 7 since the 3 was scratched. It ain't necessarily so. It might now come out 1, 5, 4, 7.

great point.

However that could be an argument for either side!
Some of these computers will actually account for that.

A reason why I mention "quality of information(data) for each rival"
is that the data need not be, can't always be, and probably isn't, the same for both man and machine.
Statistically(handicapping w/ stats) a computer can be far superior to me. Visually for the time being I have an edge.

We talk about CHESS = both parties have the Same, Perfect info. Chess is relatively simple. The human is locked into betting or playing. The computer becomes superior (is now superior) to the Pro when the CPU can "think" more moves ahead. I say locked in to bet because a human can't pass. However a master can draw a supercomputer fairly often.

another game = WEIQI(go) = more complex, so much less linear than Chess that even now, the master can toy with a supercomputer. Do both parties have the same perfect data?? I have to admit I'm not sure that the computer has the same perfect data, but I think it is more that the game is just less linear. I do expect the computer to catch up in the next 25 years.

with HORSEPLAYING depending on the technique of the player, the human can actually have a DIFFERENT set of data. This isn't always good for the human. However I have some opinions here from my own experience. I feel that in certain races I can have a complex insight that isn't very linear and a significant part (sometimes the majority) is from visual appraisal to where the data that the computer is using may be "distorted". I could rephrase that the accuracy of data available can become distorted around the different realities of the sport itself. If I am allowed to spot play those distortions I can still eat.

[KingChas]

No dispute that a computer program could pick the best contenders to win the race.

But let's remember that the computer program has no idea what the human aboard,directing the best contenders in the race is going to do that day.

If one were to bet strictly what the program spit out,
I would say an experienced pencil player would have the edge with the recency of watching the humans performance as of late...... (Humans= Jockey's?) :D

[JustRalph]

No dispute that a computer program could pick the best contenders to win the race.

But let's remember that the computer program has no idea what the human aboard,directing the best contenders in the race is going to do that day.

If one were to bet strictly what the program spit out,
I would say an experienced pencil player would have the edge with the recency of watching the humans performance as of late...... (Humans= Jockey's?) :D

In the larger scheme of things I have found that the jocks make little difference in the numbers. So much so that I don't pay attention unless they are under 5% Winners. Just based on my numbers.....it shakes out this way......for the last 3 years of data I have. The good jocks will be on the best horses most of the time and those horses show up as 'good horses' in other areas already.

[Dave Schwartz]

You celebrity bloggers sometimes go too far with the XOXOXO. :=}

I do not understand. Please explain.




But let's remember that the computer program has no idea what the human aboard,directing the best contenders in the race is going to do that day.

But you think a human brain can figure this out any better?

This is what always puzzles me. So many players say, "It doesn't work with a computer, therefore, I guess I'll just have to figure it out." The computer provides good, consistent answers. Maybe the are good enough, maybe they aren't. But whether they are good enough or not, does not determine whether or not the human brain would be better.

To assume it would is folly.


Dave

[thaskalos]

You celebrity bloggers sometimes go too far with the XOXOXO. :=}

Like Dave said...how about some clarification here Al?

I have seen you use the term "celebrity bloggers" here before...but I don't know what you mean by it.

Explain it to us...and who knows; you too might qualify to be included in our club...

[pondman]

I can't see how a paper and pencil capper can compete with a computer. Computers can look at so many aspects of racing all in a second that it boggles the mind.

The computer aids in the scienctific process:

1. Recording observations
2. Forming conjectures
3. Deducing a prediction
4. Testing and affirming consequences

We all emphasize different parts of the process. I prefer to spend most of my time on the observations, and trying to find those over looked angles, by avoiding the crowd and most contemporary speed handicapping methods. I also think it's best not to over analyze the game with too much math, because the unknowns will always be greater than the know. Therefore, my model will be different than someone who does exhaustive statistical testing.

My point:


To have a successful software product, all of these steps need to be done before any of the code can be written, or the software won't be worth anything. In other words the programmer must be an expert at the game prior to writing any code. I'd argue it doesn't happen the other way around. Software will not make you successful unless you understand the thought process behind the design. And therefore need to be careful with an over the counter product.

[Dave Schwartz]

To have a successful software product, all of these steps need to be done before any of the code can be written, or the software won't be worth anything. In other words the programmer must be an expert at the game prior to writing any code. I'd argue it doesn't happen the other way around. Software will not make you successful unless you understand the thought process behind the design. And therefore need to be careful with an over the counter product.

Mr. Pondman,

The purpose of today's highly flexible software is not to lead you down the path that the programmer has chosen for you. Rather, it is (or should be) to help you discover your own path.


1. Recording observations
2. Forming conjectures
3. Deducing a prediction
4. Testing and affirming consequences

All of the above is/should be true, but not before hand by the programmer. What you have done is created a wonderful definition of what the user needs to do to develop his own strategy.


Regards,
Dave Schwartz

[pondman]

This is precisely what one must do to be competitive. If you try to compete on the level of the "big boys" you will fall short.


Is there really such a thing as a "big boy"?

I assume you mean the resources to manipulate and move a market, therefore whipping the guppies out of their money.

IMO...it's the contrarian that will take the rings. They have an understanding of volume and more efficient measurements than a whale with an ego and the majority (the herd or crowd.) They won't risk their money until a majority has made a mistake; And then the contrarian walks away with a wheelbarrow. It's true of most mutal games-- poker, racing, and the equity markets. The money isn't earned in little baby steps. It's raked in with a few well placed home run bets, where the margins are large, and the other players make mistakes.

Does it take a super computer to do it? No. It takes organized data (and an iron stomach), but most people on the site with a 6th grade education could do it.

[thaskalos]

Is there really such a thing as a "big boy"?

I assume you mean the resources to manipulate and move a market, therefore whipping the guppies out of their money.

IMO...it's the contrarian that will take the rings. They have an understanding of volume and more efficient measurements than a whale with an ego and the majority (the herd or crowd.) They won't risk their money until a majority has made a mistake; And then the contrarian walks away with a wheelbarrow. It's true of most mutal games-- poker, racing, and the equity markets. The money isn't earned in little baby steps. It's raked in with a few well placed home run bets, where the margins are large, and the other players make mistakes.

Does it take a super computer to do it? No. It takes organized data (and an iron stomach), but most people on the site with a 6th grade education could do it.

I disagree!

If we have an "edge"...then the money should be earned in "little baby steps".

When bettors attempt "killings"...they often only kill THEMSELVES in the process.

Take a look at the casino...

It only holds a tiny edge over the player...and yet is able to turn that edge into MONSTEROUS profits over time.

We players often speak of "patience"...and yet we endeavor to carry our profits away in a "wheelbarrow"?

[Dave Schwartz]

Is there really such a thing as a "big boy"?

Well, I don't know. I would think a guy wagering $100m-$300m a year is a "big boy."

When you consider that $1 out of every $7 wagered in the U.S. last year was wagered by one of six people, yes, I would say there are big boys.


IMO...it's the contrarian that will take the rings. They have an understanding of volume and more efficient measurements than a whale with an ego and the majority (the herd or crowd.)

I think you have to be contrarian just to survive. The problem is that with all those "big boys" getting theirs first, there just aren't very many rings left to be grabbed.


They won't risk their money until a majority has made a mistake; And then the contrarian walks away with a wheelbarrow.

That is a great theory. The problem is that there are fewer mistakes being made than ever before. As for the "wheelbarrow," I'd say that you have two choices:

* find a longshot that makes you a big profit (but you can't wager any big money). These come up fairly often.

* find a low-priced horse that was overlooked. These come up more often, produce far less ROI, but you can bet a ton on them.

Of course, I guess that depends upon your definition of what "wheelbarrow-size" really is. If you mean, make a few hundred bucks, fine. But if you mean make steady money, banging out (say) $80k per year or more, well, I think that is going to be very difficult.

One of my users has said for years that he bets $200 per race about 30 times per day. He handicaps and watches about 60 races to get the 30 plays. That is $6,000 per day, $30,000 per week. For him to make a "working wage" of (say) $1,500 per week, he needs a 5% advantage. That is very do-able, especially with today's rebates.

However, if you use the "patient wheelbarrow" approach (as I imagine you mean), you are not going to find 30 bets per day. You might very well look at 60 races and find 5 or 6 bets. In that case, even if you can wager $200 per horse, you are now betting just $6k per week.

In this scenario, you've got to show a 25% advantage to make a good "living wage." IMHO, that math simply doesn't work today, if it ever did.

Sorry to paint such an ugly picture, but this is the reality that most people face: get seriously hi-tech or become a fun player (and expect to lose).


This is, of course, just my opinion.


Regards,
Dave Schwartz

[Capper Al]

Like Dave said...how about some clarification here Al?

I have seen you use the term "celebrity bloggers" here before...but I don't know what you mean by it.

Explain it to us...and who knows; you too might qualify to be included in our club...

Dave and Thaskalos,

Anyone that likes to post a lot is a celebrity blogger. I guess I'm one too.

[Capper Al]

In the larger scheme of things I have found that the jocks make little difference in the numbers. So much so that I don't pay attention unless they are under 5% Winners. Just based on my numbers.....it shakes out this way......for the last 3 years of data I have. The good jocks will be on the best horses most of the time and those horses show up as 'good horses' in other areas already.

My findings are the same.

[Capper Al]

great point.

However that could be an argument for either side!
Some of these computers will actually account for that.

A reason why I mention "quality of information(data) for each rival"
is that the data need not be, can't always be, and probably isn't, the same for both man and machine.
Statistically(handicapping w/ stats) a computer can be far superior to me. Visually for the time being I have an edge.

We talk about CHESS = both parties have the Same, Perfect info. Chess is relatively simple. The human is locked into betting or playing. The computer becomes superior (is now superior) to the Pro when the CPU can "think" more moves ahead. I say locked in to bet because a human can't pass. However a master can draw a supercomputer fairly often.

another game = WEIQI(go) = more complex, so much less linear than Chess that even now, the master can toy with a supercomputer. Do both parties have the same perfect data?? I have to admit I'm not sure that the computer has the same perfect data, but I think it is more that the game is just less linear. I do expect the computer to catch up in the next 25 years.

with HORSEPLAYING depending on the technique of the player, the human can actually have a DIFFERENT set of data. This isn't always good for the human. However I have some opinions here from my own experience. I feel that in certain races I can have a complex insight that isn't very linear and a significant part (sometimes the majority) is from visual appraisal to where the data that the computer is using may be "distorted". I could rephrase that the accuracy of data available can become distorted around the different realities of the sport itself. If I am allowed to spot play those distortions I can still eat.

I'm not saying that the paper and pencil cappers will be eliminated. It's just so much easier using a computer. It frees the mind to make better bets.

[Capper Al]

The computer aids in the scienctific process:

1. Recording observations
2. Forming conjectures
3. Deducing a prediction
4. Testing and affirming consequences

We all emphasize different parts of the process. I prefer to spend most of my time on the observations, and trying to find those over looked angles, by avoiding the crowd and most contemporary speed handicapping methods. I also think it's best not to over analyze the game with too much math, because the unknowns will always be greater than the know. Therefore, my model will be different than someone who does exhaustive statistical testing.

My point:


To have a successful software product, all of these steps need to be done before any of the code can be written, or the software won't be worth anything. In other words the programmer must be an expert at the game prior to writing any code. I'd argue it doesn't happen the other way around. Software will not make you successful unless you understand the thought process behind the design. And therefore need to be careful with an over the counter product.

Absolutely or, at least, given the specification by a handicapping expert to code. Too many programs are written by computer/math types. These guys look at a databases and work the odds from there. No great insights here. My program comes with years of handicapping experience and reading and interviewing a lot of handicapping authors. I'm still struggling with lightly raced horses. I thought that was going to be easy to figure. At the moment, I find mysel regrouping.

[Capper Al]

Dave,

How do you know the following to be true?

"When you consider that $1 out of every $7 wagered in the U.S. last year was wagered by one of six people, yes, I would say there are big boys."

[Capper Al]

Off to testing my app for the PA contest.

[dnlgfnk]

I'd go dizzy doing all this. Seriously, I have to remember to read the past performance. So much of the information that I'll need to handicap has already been digested for me by my app.

One could, Al. The consolation is that you're dealing with reality.

It's a bit like projecting par figures, though. Once I see the interpretation of a race verified by the outcome, much of the capabilities of the horses in that race are noted, and much of the effort has already been done for their next race--a microcosm of the computer constantly updating. After a few races one can be very familiar with the entire colony, save for firsters.

For instance, in Santa Anita's opener today Sir Allison was a legitimate even money risk. Some fairly obvious evidence from his only possible early speed rivals (except for the 12-1 firster) suggested that they did not possess contentious speed, and he would sail.

I've already gauged him as a very confident underlay if he draws inside another comparable speed rival well outside, or draws outside a few speed types with the turn in mind, goes off a tepid favorite or in the money choice, plus some other factors. The crowd, or perhaps a computer, might consider his winning margin, etc. as having significant predictive value. At least I'm hoping.

[dnlgfnk]

http://www.horsestreet.com/BBSImages/PickingWinnerAtDmr.JPG

The checkered pencil pusher in action at Del Mar.

Dave,

I laughed, but some other individual may think that rather than tolerate non-tech handicappers, or say "whatever works for you", you might actually condescend towards them and reject the idea that they have anything to offer.

[Dave Schwartz]

Well, you are right. Perhaps I should lighten up a bit.

Besides, the guy in the picture is not a paper-and-pencil handicapper anyway. Sure, he's got his paper, and he's got his pencil, and he might be an absolute genius with a slide rule, but it appears to me that he is working without the benefit of modern information.

That is really my point, you know. I am not saying that to win you must use hi-tech software. I am saying that you must use the output of hi-tech software.

In my own software there are over 4,000 factors for every horse, but I don't use anywhere near that many. Where the hi-tech part comes in is deciding which dozen or less of those 4,000 factors work well enough together to allow me profit.

In fact, what I use is my own version of NewPace (http://store.thehorsehandicappingauthority.com/products/-NewPace%3A-Early-vs.-Late-Video-Seminar-%28Download%29.html), which is comprised of 5 "handicapping objects," each made up of from 2 to 5 factors.

I could, in fact, do precisely what I am doing with paper-and-pencil and calculator. It would be fatiguing to handicap 8 races per hour like I do now, but I could probably do 4-5 per hour (once I got practiced at it).

I do love that picture, though. The guy is really sad looking. As the saying goes, "but for the grace of God..."


Regards,
Dave Schwartz

[bob60566]

From a handicapping book I read and now use is in my handicapping . If you are using more than four factors to arrive at any angle it is to many.
Intresting??

Mac:ThmbUp: :ThmbUp:

[Dave Schwartz]

Mac,

The question is "which 4?"



Dave

[bob60566]

Mac,

The question is "which 4?"



Dave

Dave

That is where the the individual handicapper has to get to and that does not come overnight.
Mac:)

[benzer]

Mac,

The question is "which 4?"



Dave

Many things comprise every horse race, which 4?.
Factors alone mean very little.
Is it just the nose pick? Doubtful
Is it just ego? Doubtful
Is it just speed? Doubtful
Is it just Pace? Doubtful
Is it just Trainer? Doubtful
Is it just Jockey? Doubtful
I could go on and on and on and on and (you get the idea).

[windoor]

From a handicapping book I read and now use is in my handicapping . If you are using more than four factors to arrive at any angle it is to many.
Intresting??

Mac:ThmbUp: :ThmbUp:

I guess it depends on what you call a factor. Would any filter be considered a factor? My way of thinking says it is.

I get very excited when I see just two or three filters (when combined) return a win percent in the high teens across all races, even though it shows to be a big loser if played in that state.

After I find a such a thing, it always gets my standard eight to twelve filters added in. Some are just negative expectations filters. These are nothing more than filters that eliminate horses that have shown certain, less than desirable qualities.

The hit rate is so low on these type of horses that it behooves you to throw them out, regardless of other factors that may be pointing to the horse as a contender.

Other positive expectation horses usually show up as a group. I have one that collectively wins about 40% of the time when one or more is in a race.

Picking the right one, is of course, the dilemma.

Regards,

Windoor

[raybo]

Expert handicapper before a program is not necessary, as Dave stated, but if you're going to create a new program, it certainly doesn't hurt.

Personally I started with 20+ years of studying the game, before writing my first "program", (what a joke that was!). Once I had the experience and quite a bit of gained knowledge, it still remained, after creating my first spreadsheet, to research, use trial and error, later using a crude database, etc., etc., etc., before what I do today ever came to light.

But, without the "knowledge and experience", I doubt I would have ever gotten this far.

[Dave Schwartz]

Many things comprise every horse race, which 4?.
Factors alone mean very little.
Is it just the nose pick? Doubtful
Is it just ego? Doubtful
Is it just speed? Doubtful
Is it just Pace? Doubtful
Is it just Trainer? Doubtful
Is it just Jockey? Doubtful
I could go on and on and on and on and (you get the idea).

I think I do get the idea. What you are saying is that different things win different races.

However, even if you consider the 4,000 factors which make up my program as The Short List of factors, it is safe to assume that looking at every race as a completely unique event will cause one to re-invent the wheel every time. That will not work.

We must have some path or thread to follow that takes away the perceived uniqueness. In other words, we need a finite number of strategies to use in our process.


Regards,
Dave Schwartz

[bob60566]

I think I do get the idea. What you are saying is that different things win different races.

However, even if you consider the 4,000 factors which make up my program as The Short List of factors, it is safe to assume that looking at every race as a completely unique event will cause one to re-invent the wheel every time. That will not work.

We must have some path or thread to follow that takes away the perceived uniqueness. In other words, we need a finite number of strategies to use in our process.

Regards,
Dave Schwartz
Dave I agree but each to our own in that process.

Mac:)

[Dave Schwartz]

Dave I agree but each to our own in that process.

Absolutely!

[benzer]

That was my point. Single factors are not profitable long term. 3 or 4 factors applied to every race at every track will also come up short.

Dave just so you know I have no experience with your program and I was not implying anything good or bad about it. 4,000 factors in your program would seem like you are not keeping any stone unturned. :)

[Overlay]

That was my point. Single factors are not profitable long term. 3 or 4 factors applied to every race at every track will also come up short.
I agree about the eventual unprofitability of individual factors, and the same would apply to any multi-factor system designed to arrive at a single selection to be played regardless of odds. But I believe that a manageable number of factors can be used to develop a full-field betting line that can be used on a long-term, generic basis with respect to tracks and race types to accurately find wagering value.

[Capper Al]

Let's get real. Take a look at Saturday (1/7/12)GP race 11. It is event #8 in the PA contest. After scratches, and I post for the contest, let's discuss this race?

Here's what I do to myself. The program makes selections, but it also flags horses with some potential to watch out for. In this race, the selections are different from the horses with the hidden potential. What does a human do with conflicting data? The human has to decide, not the computer. I'm going with the selections, but the trackman and/or BRIS go with the others. We'll see who's right. More details later.

[bob60566]

That was my point.Single factors are not profitable long term. 3 or 4 factors applied to every race at every track will also come up short.

Dave just so you know I have no experience with your program and I was not implying anything good or bad about it. 4,000 factors in your program would seem like you are not keeping any stone unturned. :)

That is 100% correct.
So how many factors should i use every race at given track every day.

Mac:confused:

[raybo]

That is 100% correct.
So how many factors should i use every race at given track every day.

Mac:confused:

The ones that work at that track, at that time of year, at that distance, at that race shape, on that surface, at that class, in those weather conditions, etc., etc., etc..

Nothing works, at a particular track or at any track, for all races. Each race is unique, so your method must be dynamic and current.

[benzer]

That is 100% correct.
So how many factors should i use every race at given track every day.

Mac:confused:
That my friend is the million dollar question. A question that when answered I will certainly tell you about it.

A few factors is not enough to have any real profit potential long term.

Too many factors can be confusing and will many times overlook the bombs.

[raybo]

One must have the ability to access ALL factors, but there are "core" factors that apply to individual race types. Discerning which handful of factors make up the "core", for individual races, is the key to finding the true contenders, then it's all about knowing how to "gamble/invest".

Most don't have the ability, or the patience and work ethic, to find the "core". And, of the ones that do have those things, very few of them know how to "gamble/invest".

It's a dual problem set, 1 portion is dependent on the other (handicapping is dependent on wagering skills), but the other portion is not dependent on the other one, if you know how to gamble/invest your handicapping doesn't have to be quite so involved.

One thing is for sure, you must be very good at wagering.

[windoor]

One must have the ability to access ALL factors, but there are "core" factors that apply to individual race types. Discerning which handful of factors make up the "core", for individual races, is the key to finding the true contenders, then it's all about knowing how to "gamble/invest".

Most don't have the ability, or the patience and work ethic, to find the "core". And, of the ones that do have those things, very few of them know how to "gamble/invest".

It's a dual problem set, 1 portion is dependent on the other (handicapping is dependent on wagering skills), but the other portion is not dependent on the other one, if you know how to gamble/invest your handicapping doesn't have to be quite so involved.

One thing is for sure, you must be very good at wagering.


Bingo, on the first part. Certain "types" of races are more predictable than others. Finding the right factor set for each type is the key.

And we do not have to be right all the time. The average odd will dictate how often we do, and whether we make a profit or not.


The second part is spoken like a true exotics player. While I am sure ticket structure is of paramount importance for Super players, I assure you it is not needed for "win only" players. A simple flat wager works well enough here :)

At least my own records says this is so.

Regards,

Windoor

[Capper Al]

One must have the ability to access ALL factors, but there are "core" factors that apply to individual race types. Discerning which handful of factors make up the "core", for individual races, is the key to finding the true contenders, then it's all about knowing how to "gamble/invest".

Most don't have the ability, or the patience and work ethic, to find the "core". And, of the ones that do have those things, very few of them know how to "gamble/invest".

It's a dual problem set, 1 portion is dependent on the other (handicapping is dependent on wagering skills), but the other portion is not dependent on the other one, if you know how to gamble/invest your handicapping doesn't have to be quite so involved.

One thing is for sure, you must be very good at wagering.

You're right, but is it a blessing to be driven so much or a sickness? I'm just glad to see others like me, blessed or ill whatever we are.

[Capper Al]

Picks: 12-2-3 (9-11)
Flags: 3-8-11 not necessarily in order

We'll talk about these after the race.

[bob60566]

I believe Dave is really telling it like it is. Human thinking power out does super computer every time. We all know this is a game of being able to think for your self...not follow the hoards to the windows to bet false favorites. We are close to getting an edge now than ever before in history. For instance we can soft focus on the race to see who's come to the party and then review each horses resume to see how he fits in today's race. We can start at the beginning of a horses career and work upwards to see how the trainer is performing and whether or not he is placing the horse properly. We can see how the horse responds to that effort and if he is improving and is placed today to win. No computer can do that for us nor can we get it from anyone else other than our own thinking power. "Think and Grow Rich" by Napoleon Hill is a good book to help us with this. ;)
I must have missed this post earlier
Mac:ThmbUp: :ThmbUp: :ThmbUp:

[riskman]

I must have missed this post earlier
Mac:ThmbUp: :ThmbUp: :ThmbUp:

I think you missed post#24 by DS.

[bob60566]

Jerry, I think you misunderstood. I said that the human has no chance.#24

That to me seems out of context in handicapping with computers,

Mac:)

[Capper Al]

Picks: 12-2-3 (9-11)
Flags: 3-8-11 not necessarily in order

We'll talk about these after the race.

This time I've got egg all over my face. My guys lost big time. They finished 10-4-5-6. None of my picks in sight. I'd say looking back I should have dropped any horse in the 12 post but the 10 horse won. My computer app just blew it, so it's back to the code.

[jerry-g]

If we are to learn anything from these types of races we need to look at how they did it. Here is my thinking of why the 10 won.

Only two horses had made any moves on the turn with any recency in its last race. They were the 1 horse and the 10. The 1 had 5 races where it raced w/i 5L of the lead from 1st call to Str. The 10 had 7. The one had never raced w/i 1L of the lead in any of its last 10 races, the 10 had done 8. The 10 was w/i 1L in its last race at the 2C and Str at a mile race. Todays race was 7F. These are just things worth noting on horses that win and why.

[TexasDolly]

Jerry what race are you referring to. Thank you,
TD

[raybo]

You're right, but is it a blessing to be driven so much or a sickness? I'm just glad to see others like me, blessed or ill whatever we are.

I guess that depends on what your ultimate goal, in horse racing, is. If you would "like" to be profitable, while enjoying this great game, then I suppose being so "driven" is not necessary.

However, if your ultimate goal is to "be" profitable", then it is a necessity. Would you expect to be profitable, in almost any other endeavor, without displaying "patience and work ethic"? Especially in an endeavor where 98% are "unprofitable".

[thaskalos]

If we are to learn anything from these types of races we need to look at how they did it. Here is my thinking of why the 10 won.

Only two horses had made any moves on the turn with any recency in its last race. They were the 1 horse and the 10. The 1 had 5 races where it raced w/i 5L of the lead from 1st call to Str. The 10 had 7. The one had never raced w/i 1L of the lead in any of its last 10 races, the 10 had done 8. The 10 was w/i 1L in its last race at the 2C and Str at a mile race. Todays race was 7F. These are just things worth noting on horses that win and why.

Here is my question concerning this race:

If we insist on playing maiden claiming races, for three year-olds, ON THE TURF...then what races are we supposed to pass on?

We all talk about the virtue of "patience"...and then, not only do we wager on one the most chaotic and confusing races of all...we endeavor to rehandicap the race after it's run -- in order to make sense out of its outcome.

We marvel at the computer's ability to process mountains of data in mere seconds...and then we use that computer to handicap a race in which there are barely any relavent data at all to process.

I know, I know...it's "different strokes for different folks"...and some handicappers consider these types of races to be the most profitable ones out there...

IMO...races like these reduce handicapping to just an uneducated guessing game. And we cannot afford to just be guessing out there; not when we are facing takeouts of 17% to 30%.

Having said that...the 10 horse had more than a few things going for it.

Along with an impressive-looking, fast-closing, first start on the turf...the horse was also coming out of a "key race".

Both horses who finished ahead of him in that race WON their next start...also on the turf.

Pass races gentlemen...computer or no computer.

[ranchwest]

Here is my question concerning this race:

If we insist on playing maiden claiming races, for three year-olds, ON THE TURF...then what races are we supposed to pass on?

We all talk about the virtue of "patience"...and then, not only do we wager on one the most chaotic and confusing races of all...we endeavor to rehandicap the race after it's run -- in order to make sense out of its outcome.

We marvel at the computer's ability to process mountains of data in mere seconds...and then we use that computer to handicap a race in which there are barely any relavent data at all to process.

I know, I know...it's "different strokes for different folks"...and some handicappers consider these types of races to be the most profitable ones out there...

IMO...races like these reduce handicapping to just an uneducated guessing game. And we cannot afford to just be guessing out there; not when we are facing takeouts of 17% to 30%.

Having said that...the 10 horse had more than a few things going for it.

Along with an impressive-looking, fast-closing, first start on the turf...the horse was also coming out of a "key race".

Both horses who finished ahead of him in that race WON their next start...also on the turf.

Pass races gentlemen...computer or no computer.

I suppose everyone is going to have a different notion of which races to play, but I agree that selecting specific types of races can be highly beneficial. Right now, I am only betting races of a single distance and surface, no non-winners in a time frame, and only certain race types. I find this a lot easier than betting 4 1/2 f on the dirt at LA on the same day as betting a 1 1/4 turf race at GP and a 6 f race on AW at AP. Keeping it simple.

[Capper Al]

Here is my question concerning this race:

If we insist on playing maiden claiming races, for three year-olds, ON THE TURF...then what races are we supposed to pass on?

We all talk about the virtue of "patience"...and then, not only do we wager on one the most chaotic and confusing races of all...we endeavor to rehandicap the race after it's run -- in order to make sense out of its outcome.

We marvel at the computer's ability to process mountains of data in mere seconds...and then we use that computer to handicap a race in which there are barely any relavent data at all to process.

I know, I know...it's "different strokes for different folks"...and some handicappers consider these types of races to be the most profitable ones out there...

IMO...races like these reduce handicapping to just an uneducated guessing game. And we cannot afford to just be guessing out there; not when we are facing takeouts of 17% to 30%.

Having said that...the 10 horse had more than a few things going for it.

Along with an impressive-looking, fast-closing, first start on the turf...the horse was also coming out of a "key race".

Both horses who finished ahead of him in that race WON their next start...also on the turf.

Pass races gentlemen...computer or no computer.

Good point about being a race to pass. On our topic about computer handicapping, I just read my applicaton's output and never checked the past performances about this race. An easy bad habit for those of us who use computers to handicap to fall into. The computer output is organized better and easier to read, but one can't take the race conditions and the horse's past story as displayed in their PPs out of the equation.

[Dave Schwartz]

IMHO, deciding which races to pass are a function of good handicapping output. That is, you must have a mechanism which tells you, "This is a playable race."

Every race that I handicap ultimately delivers me one of 3 messages:

1. PLAY because there is at least one low-priced horse that I can take a stance against.

2. PLAY because there the low-priced horses together will not likely be beaten and profit is a reasonable long-term expectation.

3. PASS because neither of the above scenarios exist.


If the race is a "play," then I need to know how much to wager. This is a function of advantage and hit rate. I am still amazed that I now see the true beauty of Kelly. It has completely changed my play. I was playing live last week with a small audience and had a race which called for an 80-unit wager spread across 3 horses, followed a couple of races later by a 1-unit bet on a single horse. Obviously, the message was clear: The 1-unit race should not be played at all. (My average race-play is around 35 units.)

My stats tell me that the more units I wager in a race the better I do, both in hit rate and $ return.

IMHO, this is how it is supposed to work.


Regards,
Dave Schwartz

[thaskalos]

IMHO, deciding which races to pass are a function of good handicapping output. That is, you must have a mechanism which tells you, "This is a playable race."

Every race that I handicap ultimately delivers me one of 3 messages:

1. PLAY because there is at least one low-priced horse that I can take a stance against.

2. PLAY because there the low-priced horses together will not likely be beaten and profit is a reasonable long-term expectation.

3. PASS because neither of the above scenarios exist.


If the race is a "play," then I need to know how much to wager. This is a function of advantage and hit rate. I am still amazed that I now see the true beauty of Kelly. It has completely changed my play. I was playing live last week with a small audience and had a race which called for an 80-unit wager spread across 3 horses, followed a couple of races later by a 1-unit bet on a single horse. Obviously, the message was clear: The 1-unit race should not be played at all. (My average race-play is around 35 units.)

My stats tell me that the more units I wager in a race the better I do, both in hit rate and $ return.

IMHO, this is how it is supposed to work.


Regards,
Dave Schwartz
What about races such as this one Dave...which offers very little pertinent information to assist us in making a decision.

A maiden claiming race, for three year-olds, on the turf...featuring a mix of first-time starters, horses with no established turf form...and others who only show a lone turf start, or two...

Is there really a need to play races like these, in today's full-card simulcasting era?

Why not stick to races where our information is a little more dependable?

[Dave Schwartz]

LOL - That race falls into the category that I call "obscure." I would never handicap such a race.

Why bother when there are so many good ones around?

[thaskalos]

LOL - That race falls into the category that I call "obscure." I would never handicap such a race.

Why bother when there are so many good ones around?
I knew it!

Great minds think alike...:)

[dnlgfnk]

One of the stated advantages of computer handicapping, cited when one questions the "mere" 24% ROI of Benter, for instance, is the sheer volume of plays which overcomes the more selective, higher ROI player. But I'm seeing many passed races being advocated here. I believe Benter rarely passed races save for those containing firsters.

Secondly, I was under the impression that a substantial number of factors for each horse, multiplied by their respective coefficients according to strength of predictability, then measured against the rest of the field's "points", produced a percentage for each horse, i.e., true odds. But it seems the large number of factors are actually corollaries of a few major factors, differing little within their grouping (best overall fig, 2nd best, best last fig, 2nd best last fig., etc.)

I think I actually apply more factors, albeit roughly the same cornerstone factors( non-numerical judgements of pace, position on track, running styles, odds, etc.) significantly dealing with the dynamics of the race in my approach, which I wouldn't have believed after reading of Benter's exploits. Or are the number crunchers here aware of whether they differ greatly from his methodology?

[Speed Figure]

What are units??

[RXB]

What about races such as this one Dave...which offers very little pertinent information to assist us in making a decision.

A maiden claiming race, for three year-olds, on the turf...featuring a mix of first-time starters, horses with no established turf form...and others who only show a lone turf start, or two...

Is there really a need to play races like these, in today's full-card simulcasting era?

Why not stick to races where our information is a little more dependable?

I do well in those races.

If I were to project my own experience onto everyone else, then no one should ever bet any race for winners at 6.5f or shorter. Just because I have a difficult time with those conditions doesn't mean that others can't excel in those races.

[thaskalos]

I do well in those races.

If I were to project my own experience onto everyone else, then no one should ever bet any race for winners at 6.5f or shorter. Just because I have a difficult time with those conditions doesn't mean that others can't excel in those races.
I believe you, without a doubt...and I hope your success continues.

There are no "absolutes" in this game...and to every rule, there is an exception.

I am not imparting universal truths here; I am only relating my own experiences...gained from many years of enthusiastic participation in this game.

And I happen to think that the vast majority of players would benefit greatly by ignoring obscure races such as these.

I intended no offense...and I apologize if any was implied.

[RXB]

There was no offence taken. I just think people should remember that the takeout rates are the same for each race at any track and thus the opportunities are equal. No higher percentage of money is lost by the public as a whole in MCL grass races than in open older claiming sprints.

[thaskalos]

There was no offence taken. I just think people should remember that the takeout rates are the same for each race at any track and thus the opportunities are equal. No higher percentage of money is lost by the public as a whole in MCL grass races than in open older claiming sprints.
Good point.

I'll try to remember that...

[Capper Al]

IMHO, deciding which races to pass are a function of good handicapping output. That is, you must have a mechanism which tells you, "This is a playable race."

Every race that I handicap ultimately delivers me one of 3 messages:

1. PLAY because there is at least one low-priced horse that I can take a stance against.

2. PLAY because there the low-priced horses together will not likely be beaten and profit is a reasonable long-term expectation.

3. PASS because neither of the above scenarios exist.


If the race is a "play," then I need to know how much to wager. This is a function of advantage and hit rate. I am still amazed that I now see the true beauty of Kelly. It has completely changed my play. I was playing live last week with a small audience and had a race which called for an 80-unit wager spread across 3 horses, followed a couple of races later by a 1-unit bet on a single horse. Obviously, the message was clear: The 1-unit race should not be played at all. (My average race-play is around 35 units.)

My stats tell me that the more units I wager in a race the better I do, both in hit rate and $ return.

IMHO, this is how it is supposed to work.


Regards,
Dave Schwartz

Dave,

Are you playing Kelly by the book, or have you come up with a way that works for you?

[jasperson]

No dispute that a computer program could pick the best contenders to win the race.

But let's remember that the computer program has no idea what the human aboard,directing the best contenders in the race is going to do that day.


Who can do this?

[Dave Schwartz]

I use Kelly "by the book" but I know my probabilities are not right. I have some problems. For example, favorites are predicted a little low - the ones at $1.91 to $2.00 are slightly profitable but I cannot bet them.

If I dial up the favs just for being favs, then I lose the $net on some of my longshots.


What I am actually doing is something I call "the Betfair approach." If you look at my previous post with the 3 scenarios, you'll get an inkling of what I am doing.

Perhaps someone who has played live with me can give you a handle on how it all comes out. Pretty amazing, actually. I mean, I certainly lose races - doesn't everyone? But most of the time, when they head for home, I'll have 2, 3 or even 4 horses going to the wire together.

It all begins in the contender process.

Regards,
Dave Schwartz

[raybo]

It all begins in the contender process.

Regards,
Dave Schwartz

There it is!! :ThmbUp: :ThmbUp:

[gm10]

LOL - That race falls into the category that I call "obscure." I would never handicap such a race.

Why bother when there are so many good ones around?

I love those races. It gives you an opportunity to employ handicapping factors (pedigree, late moves by horses that go unnoticed, trainer patterns) that are less understood, and which will be soon be replaced by form lines as the best predictors - and most people know how to read form lines!!

Late speed that goes unnoticed, I think that I am pretty good at finding that, but I don't claim to be an expert in pedigree/trainer patterns. However, they do help me from time to time, and that certainly encourages me to keep learning about this game.

[Dave Schwartz]

Just goes to show that one can choose to specialize in the obscure and gain an edge.

[Robert Goren]

100% Kelly will break you about as fast as anything if you over estimate your edge by very much. But if you have good numbers it is the way to go.

[gm10]

Just goes to show that one can choose to specialize in the obscure and gain an edge.
When you're living on the edge of obscurity, you are equally close to the regular and the obscure . I'm not sure whether that is good or bad overall :confused:

[jerry-g]

It sounds like a good job for the record keeper to me. I have my records kept for free by my online betting account. I can get an R.O.I. on just about any comb of plays I have made in the past for any given time frame. Free's me from keeping records.

[raybo]

It sounds like a good job for the record keeper to me. I have my records kept for free by my online betting account. I can get an R.O.I. on just about any comb of plays I have made in the past for any given time frame. Free's me from keeping records.

Better than no records at all, however, the ADWs I've used only give you data on your bets; race date and race #, type of bet, horses bet, amount of bet, and profit/loss. They don't give you race distance, race class including restrictions and sex/horse ages, surface or surface conditions, or final odds, etc., etc..

A good record keeper will have all this data, and maybe more.

[Dave Schwartz]

It sounds like a good job for the record keeper to me. I have my records kept for free by my online betting account. I can get an R.O.I. on just about any comb of plays I have made in the past for any given time frame. Free's me from keeping records.

Really? Does it allow you to do things like search by track, distance or type of race?

[jerry-g]

Dave

Right now for me I have to keep it simple. I am only interested in what my win bets are doing over my exotic ones. I do keep things I jot down in my Think and Do book I notice unusual about the races to help me know what is winning. Today, I picked 2/3 races at Parx but didn't bet because of value. I can get winners at all distances, tracks, classes etc. MY base right now is not large enough for me to get anything predictive out of it. One day, however, I may get interested in more complication compilations. And I'm gonna write that down in my book.

[sjk]

Secondly, I was under the impression that a substantial number of factors for each horse, multiplied by their respective coefficients according to strength of predictability, then measured against the rest of the field's "points", produced a percentage for each horse, i.e., true odds. But it seems the large number of factors are actually corollaries of a few major factors, differing little within their grouping (best overall fig, 2nd best, best last fig, 2nd best last fig., etc.)

I think I actually apply more factors, albeit roughly the same cornerstone factors( non-numerical judgements of pace, position on track, running styles, odds, etc.) significantly dealing with the dynamics of the race in my approach, which I wouldn't have believed after reading of Benter's exploits. Or are the number crunchers here aware of whether they differ greatly from his methodology?


My program does not operate by multiplying factors by coefficients or by assigning a number of "points" to each horse. For better or worse it is a great deal more complicated than anything like that.

I would think there are plenty of ways of getting to an odds line.