Software: what's missing, what's next?

[RonTiller]

A recent thread, Software Development Experience, concerning a programmer writing a handicapping program for a friend, provoked some interesting responses. Just a sampling of some comments:

GaryOz: "As someone who has been a software junkie--and have great things to say about many of the software producers--it is a pretty mature market--unless you are doing something novel."

Red Knave: "If there is any innovation to be had it's probably going to be more here [refining betting algorithms] than with the handicapping side."

JustRalph: "somebody else mentioned it is a mature area. That's an understatement."

Jeff P: "Handicapping software is a niche market - one that's pretty well saturated...it helps to invent and offer something that is both brand new and useful to the player...it had better be something that truly helps the player's win rate and roi...It also helps if the software interfaces with a database."

D Schwartz: "The ironic part is that very few software programs are even designed to win without some user input. In other words, it is safe to say that in most approaches, the user determines who wins more than the software does...Now, this does not mean that a competent handicapper will do equally well with "good or bad" software. Rather, it means that there are good tools and bad tools. Winners may find a way to win with bad software but will do better with good software."

All this got me wondering, where is handicapping software in its development cycle? What hasn't been done that should be done? Can be done?

Does the future portend pretty much the same kind of stuff, maybe just done better or prettier: better numbers (better pace, speed, power, workout ratings, trainer ratings, etc), more sophisticated algorithms for developing odds lines, prettier screens, a new razzle-dazzle set of AI paradigms under the hood (ants, termites, etc.), easier querying of databases many programs create, better integration of various components (data, handicapping, betting and record keeping)? Is everybody catching up to everybody else, converging on the same inevitable set of tools, numbers and screens? Will there be more "home grown" software in the future?

Are there conceptual leaps waiting to happen, or are novel innovations just at an end? Are we at the same place physics was at the end of the 19th century, when Lord Kelvin proclaimed all the questions in physics had been answered (but maybe we can go out a few more decimal places in our answers), and then BOOM, relativity and quantum physics bust out from nowhere and changed everything? Is this where we are with handicapping programs, doomed to ever more precise numbers, odds lines and clever approximations for questions already asked and answered? Are there questions waiting to be asked? Are there questions nobody has answered because nobody could, till now (or 10 years from now)? Will the power of computers change things?

And what about the platform? We have handicapping programs installed on individual computers, linking to data on the internet (sometimes doing live data lookups, sometimes downloading data files); in their infancy, we have various web applications, either run entirely on web servers or run in a hybrid mode, all run without installing a program on a customer's computer; we have cloud computing, cloud databases hosted by Amazon, IPhone apps, 3G this, 3G that. Will that replace existing apps, supplement them, or just allow more options?

I have no hidden agenda here and I don't mean this to promote yabadabado over scoobydoobydo. I am interested in what people think the future holds, what it should hold, what they would like it to hold, regarding handicapping software. That is, other than the obvious: give me the winner in every race, tell me what to bet so I can make a million dollars after a few weeks of fiddling around after coming home from a bar fight, etc.

Ron Tiller
HDW

[BillW]

I can see detailed trip analysis come about when raw datasets are available from Trakus. Just think what you could do with a 1/30th of a second snapshot of each horse throughout the whole race.

[facorsig]

I see strong parallels between thoroughbred handicapping software and technical software I use in my profession as a corrosion engineer in oil/gas production. Both software families originated at similar times based on a few factors and a limited number of users who actually had computers, knew how to use them and saw an advantage to computerizing the large number of calculations (these arguments could probably be made for all technical software because of the relatively recent advent of computers).

Over time additional factors have been discovered to refine the calculations. Additionally, a new generation of users have come on board. Some of the users don't know how the problem was solved before computers. Many of the users don't have a strong technical background to know when to question the software output based on their own experience. The software has grown more powerful and enabled more people to get correct answers, but also a large number of incorrect users getting incorrect answers.

Until now, all of the programs have been empirical, taking observations from experiments under controlled conditions and trying to extrapolate those to today's condition. In actual fact, there are about 10 - 12 software programs used in my profession which all go back to a single set of experiments by Shell. Possibly, a case could be made that many of our handicapping software programs today go back to work by Sartin. Sartin will have made some observations about horses and developed some algorhythms to quantify his observations. In my profession, I am just now seeing the first couple of new software programs which are principles based. I have been indirectly associated with one of the program's fundamental research for about 12 years. The quantity of work that has gone into a principles based software is staggering and involves specialized (read expensive) people.

A principles based thoroughbred handicapping software might have genetic, biological and physical factors combining to predict today's race. This is deeper than pedigree and would almost be based on some sort of genetic coding, like DNA. Biological factors could be heart size, blood pressure, leg angles/lengths, muscular stiffness, and medications. Physical factors could be predicting a bump at the starting gate (dog racing software uses this), the compaction of the soil in the most likely running path, wind, or other factors.

This is all strictly speculation about principles based approaches because most users are now using the empirical models and the empirical models have made so many improvements that the investment for a principle based model can only be afforded by a very select few.

Hope this offers some insight.

Fred

[Dave Schwartz]

Bill,

I can see detailed trip analysis come about when raw datasets are available from Trakus. Just think what you could do with a 1/30th of a second snapshot of each horse throughout the whole race.

WHat could you do with that?


Dave

[BillW]

Bill,



WHat could you do with that?


Dave

It could replace viewing replays for trip anaylsis. With that dataset you (or your computer) could see each check-up, each change in momentum/direction etc. that occurred during a race. Picture computer generated chart comments that were complete and accurate. Nothing revolutionary, but more readily available to the casual user.

For an example, look at the Trakus simulated replays that are shown at tracks now (Keeneland for instance), a handicapper could have that capability on his/her own laptop.

Of course, like the DRF publishing the Beyer Speed figs. this would cause great pain to those willing to put in, and benefit from the hours of work viewing video to extract this information now.

[chickenhead]

one thing that became very apparent to me when I began using a database, and still holds true, is that for what we will call an average user (let's just use me as a stand-in)...a database without real built in statistical analysis that can test the potential validity of whatever is uncovered possibly causes more harm than good.

Rather than more whiz bang factors, the ability to do testing of what already exists seems to be the biggest thing missing. So far as I can tell, the way most people do this sort of thing right now is just a very basic in-sample, out-of-sample kind of thing.

You develop over here against this stuff, and you run it against that stuff over there. I don't know jack about statistics, but it seems you can do that often enough, with enough stuff, and you'll find all sorts of things that appear valid but are really garbage. Powerful database querying tools require powerful analysis tools to determine how meaningful the results of those queries are.

[Tom]

Chick.....the new HTR "Robot is already doing that......truly and amazing tool.
One neat thing is you can just let the robot tell you what is winning, what is losing, what is profitable.

I agree with Bill...entering the Trakus data and getting trip and very good pace info would be great. Anyone who has used Sartin's KGen will see the great use of having a graphical pace line to judge how typical a race was. The only problem with KGen is it is manual input. But suppose you were looking at the "red" line..you might see it is not how the horse generally runs his races.

[Grifter]

Great topic, Ron....
Does the future portend pretty much the same kind of stuff, maybe just done better or prettier: better numbers (better pace, speed, power, workout ratings, trainer ratings, etc)... Is everybody catching up to everybody else, converging on the same inevitable set of tools, numbers and screens?
I think "yes", on all counts. As I understand it, everybody pretty much works with the same raw data... whatever Equibase provides. All the data calculations and data manipulation provided by firms such as HTR, HSH, etc., and sold through HDW, basically slice, dice and manipulate that same data. How much more significance can we torture out of that raw data?

Bill W. said....

I can see detailed trip analysis come about when raw datasets are available from Trakus. Just think what you could do with a 1/30th of a second snapshot of each horse throughout the whole race....It could replace viewing replays for trip anaylsis. With that dataset you (or your computer) could see each check-up, each change in momentum/direction etc. that occurred during a race. Picture computer generated chart comments that were complete and accurate. Nothing revolutionary, but more readily available to the casual user.

For an example, look at the Trakus simulated replays that are shown at tracks now (Keeneland for instance), a handicapper could have that capability on his/her own laptop.
Have not seen the Trakus simulated replays, but take it a step further. Imagine a program that takes video or Trakus data and analyzes it in seconds (milliseconds?) and generates a report that could be yet another tool. A very cool tool, indeed, removing the tedium of watching endless video replays. Wasn't trip handicapping part of what Bentner did? But, of course, if that technology becomes generally available, the edge is gone.

-- Grifter

[BillW]

Have not seen the Trakus simulated replays, but take it a step further. Imagine a program that takes video or Trakus data and analyzes it in seconds (milliseconds?) and generates a report that could be yet another tool. A very cool tool, indeed, removing the tedium of watching endless video replays. Wasn't trip handicapping part of what Bentner did? But, of course, if that technology becomes generally available, the edge is gone.

-- Grifter

They just look like a cartoon of the real race replay. Very similar in looks to the simulation available in Equisim. The only difference is that the "cartoon" is driven by the positional data (at 1/30th second increments) collected during the running of the race by the Trakus system.

[InControlX]

While we're dreaming... Is the Trackus data availabale anywhere for any track? I haven't been able to find a source, and would appreciate a link.

ICX

[CBedo]

www.trakus.com

??

[formula_2002]

http://www.paceadvantage.com/forum/showthread.php?t=55240.

HOPEFULLY, THE ABOVE THREAD, AND THOSE THAT FOLLOW, WILL DEMONSTRATE THAT GIVEN GOOD DATA, REASONALLY GOOD DATA ANALYSIS,AND A QUICK LOOK AT THE TOTE BOARD, ONE CAN COMPUTER GENERATE SELECTIONS WITH NO SUBJECTIVE INPUT AND ATLEAST BREAK EVEN.
A 5% OR SO REBATE COULD BE THE TIPPING POINT.

IN SOME CASES, THE ONLY SOFT POINT MAY BE THE READING OF THE TOTE BOARD.

[InControlX]

CBedo,

Thanks. Oops! I hadn't checked in awhile. Too bad so few tracks are on the system.

ICX

[garyoz]

Just caught up with this post Ron. As ever, a great post. From the responses, nothing new except better data collection through Trakus (yawn)-I think there are more elegant & dynamic solutions available which can come from quantitative finance theory and practical application. When you get down to it, quasi-Sartin solutions (past performance static comparables) are still the norm in software methodologies. They are so 1990, and its almost 2010.

[DJofSD]

What would be a great adjunct is the ability to capture my thought processes as I handicap a race. As a selection oriented player I would like to be able to "replay" how I got to the winner or the loser.

The value would be in the post race review. What did I do right? What did I do wrong? Did I overlook something that in a post race review now looms off the page as a critical factor? Did I correctly dismiss a false favorite? Capture these decission points into a data base for analysis.

[garyoz]

The Problem is the operationalization of that decision process. "naming" what you are doing and having construct validity across that name. Except for value--which is a very high level concept (e.g., caused by the factors that you want to name) favorites get tossed for many reasons. Is it trainer, is pace, is it bouncing, etc.--and worse, an interaction or combination between these factors. Also it is not necessarily binary.

[DJofSD]

Yes, I understand the decission point is very often not a binary one.

And, yes, there are a lot of factors that have lots of interaction with other factors. I've often thought the reason why NN have little value is b/c there is not any way to understand how a decission was arrived at -- it truely is a black box.

Part of the idea behind my initial posting was having the ability to use some kind of bio-feedback monitoring. One idea is use a tracking device to monitor what you are looking at while looking at the computer screen. In essance, being able to capture on a screen by screen basis what you were looking at on each screen.

[RonTiller]

Rather than biofeedback, how about handicapping with fMRI - maybe different areas of the brain "light up" during different decision events.

How we actually make handicapping decisions is a very interesting topic. When I first met Jim Cramer, we thought that we should be able to capture or embody the decision making process he goes through in deciding what to bet; yet it always eluded us. Is it like embodying medical diagnosis in a program? There are many computerized medical diagnostic tools, aren't there? Is it like computerizing the decision making process a beauty contest judge goes through (on that, biofeedback might very well be enough!). Is it like computerizing which job applicant to pick from a pool of applicants in Human Resources?

Jim Cramer kind of went about it obliquely - he sat down and mad a list of 100 questions he wanted answered in deciding what bet(s) to make in a race and which horses to bet. These were not necessarily 100 questions he consciously asked himself - after all, he tended and tends to handicap very quickly. But these questions were never far from the surface and although he may not have explicitly asked himself question #78, if asked, he could probably give a literate and informed answer.

He arrived at these 100 questions by doing something he rarely did - he sat down and spent several hours on a few races, trying to capture his thought processes. Tom Hambleton took these 100 questions and made them sing like Pavarotti - after doing several races 1 question at a time, he found he didn't really need to go through each question individually. I guess he internalized them somehow.

At any rate, we don't have the questions anymore, nor do we have the typed up analyses from Tom, where he went through example races one question at a time. That kind of thing would indeed make an interesting approach to a handicapping program.

Gary - can you expand on what you mean by "quantitative finance theory and practical application."

I guess in the end, maybe handicapping programs are a lot like automobiles. Their function is to get us from point A to point B, some more comfortably, flashy, quickly or economically than others. Over the years, the engines change, stereos get better, transmissions get smoother, etc, but ultimately, there's a passenger compartment, wheels, something under the hood and the dashboard controls.

Handicapping programs will continue to have better engines under the hood (better numbers, better probabilities, better spot plays, better algorithms, etc.), better dashboards (user interfaces), more stuff to make life easier, but in the end, their function is to...??? Well, how about that, I'm not even sure what the function of a handicapping program is. To assist handicappers in handicapping? To tell them what to bet? To help them get from point A (race 1 is coming up) to point B ("Pass" or "$12 trifecta box with 5, 7 and 8)? To suck money out of hapless victims in a game that is impossible to beat? Sigh...

Ron Tiller
HDW

[ddog]

a very simple question,

has anyone done or seen any db testing of changes in post position of horses from race to race?

I think I saw some service that had post position stats based on ODDS , but that was all?


What about based on a set of "expected" pre-race data points covering all the horses past races or a single set of back to back races?


There was a track , which must go nameless, that I played and the horses a couple of weeks ago from the inside couple of posts had NO shot at all.

When those horses ran back this week(end) you could include them at great odds given the "failures" and really make some bucks.

It was not isolated to a horse or two.

I wonder about testing over a broad range of racing dates, even back to a year, to see if something emerges without needing to see or read all the charts.

Too much noise??????

sorry, that's about all the lighting I can stand there!
:)

[Nitro]

I just happened to catch the “Post of the Day” and thought I might add my 2 sense. Although I doubt that my comments will change the minds of those who are deeply involved with any type of softwhere. It still might help others contemplating the real value of its use.

I have no hidden agenda here and I don't mean this to promote yabadabado over scoobydoobydo. I am interested in what people think the future holds, what it should hold, what they would like it to hold, regarding handicapping software. That is, other than the obvious: give me the winner in every race, tell me what to bet so I can make a million dollars after a few weeks of fiddling around after coming home from a bar fight, etc.

Ron Tiller
HDWI’m not here to burst anyone’s bubble and I also have no hidden agenda either, but I personally feel that the best words to describe all this "softwhere" mumbo jumbo IS “yabadabado” or “scoobydoobydo”. Why? Well very simply put, none of all this great “softwhere” has significantly increased the ROI of anyone using it long term. (Unless of course the ROI is based on softwhere sales!) If it did they wouldn’t be selling it to others for use in a para-mutual type game. And a game it is! The scientific approach hasn’t yielded any better results then what the serious players were accomplishing before the age of computers! Why? Because it’s just that: A game and not a science. Math is an exact science and is being applied to game where there are too many variables and unknowns for even the most sophisticated “softwhere” to handle. The core problem is that the majority of “softwhere” is analyzing the wrong element: the animal. This game is completely controlled by humans who own, train, ride and most importantly bet on this game. So as they say in the softwhere trade, “Garbage in = Garbage out”.


By Formula 2002
HOPEFULLY, THE ABOVE THREAD, AND THOSE THAT FOLLOW, WILL DEMONSTRATE THAT GIVEN GOOD DATA, REASONALLY GOOD DATA ANALYSIS,AND A QUICK LOOK AT THE TOTE BOARD, ONE CAN COMPUTER GENERATE SELECTIONS WITH NO SUBJECTIVE INPUT AND ATLEAST BREAK EVEN.
A 5% OR SO REBATE COULD BE THE TIPPING POINT.

IN SOME CASES, THE ONLY SOFT POINT MAY BE THE READING OF THE TOTE BOARD.I think this is the most sobering response to this entire thread. Why? Well if you agree that the money being wagered might be of some value to those who are placing wagers, then you might also agree that some members of the betting population have more insight into the actual condition of these horses. Something that “softwhere” can never incorporate because it’s a complete intangible to those outside of the game looking in.

Someday the everyday player (serious or not) will understand that the betting population is comprised of both insiders and outsiders. The tote-board very often can provide a crystal clear picture of what one of these groups is doing. The question is whose money provides the most significant changes on the board and why. Is it the “soft” money or the “hard” money? I would have to say the latter. Does it mean anything? Well if we agree that it might be of some real value to someone, then maybe they know something that the “softwhere” doesn’t. They might even know what the “softwhere” is telling the outsiders using it. What an advantage!

[Dave Schwartz]

Why? Well very simply put, none of all this great “softwhere” has significantly increased the ROI of anyone using it long term.

And you know this how?

If it did they wouldn’t be selling it to others for use in a para-mutual type game.

And you know this how?


The scientific approach hasn’t yielded any better results then what the serious players were accomplishing before the age of computers!

And you know this how?



What you have done is stated your opinions as facts.

You know none of these things.


Regards,
Dave Schwartz

[Tom]

Why? Well very simply put, none of all this great “softwhere” has significantly increased the ROI of anyone using it long term.

Let say that this is absolutely incorrect.

[ryesteve]

I also have no hidden agenda either
That's a relief... here I was about to spend valuable time wondering exactly what your hidden agenda is...

[Nitro]

Why? Well very simply put, none of all this great “softwhere” has significantly increased the ROI of anyone using it long termAnd you know this how?
If it did they wouldn’t be selling it to others for use in a para-mutual type game. And you know this how?
The scientific approach hasn’t yielded any better results then what the serious players were accomplishing before the age of computers! And you know this how?

What you have done is stated your opinions as facts.

You know none of these things.
Regards,
Dave SchwartzNo actually Dave I’ve stated the “facts” all right as hard as they might be to swallow. But for someone to say that another “knows none of these things” is purely a subjective judgment call that’s based on who knows what? You don't know me from Adam! I'll just say that I'm not one for BS-ing anyone especially those who are interested in getting a handle on this game. I have nothing to gain by dispelling notions of grandeur that softwhere is the answer to making long-term profits.

As far as increasing the long term ROI goes, it’s a FACT that the percentages of the 1st and 2nd betting choices producing Winners hasn’t changed one iota since the handicapping softwhere was introduced back in the 80’s. In fact, all of the calculated speed, pace and trip numbers (the scientific approach) haven’t had a positive impact either. Of course there are exceptions to the rule for some using figures, computer generated or not. But these players are certainly in the minority. If the softwhere had any type of real long-term impact the result of the betting pools should likewise reflect a change. That is if those using it had any real faith in what it produced. I might be wrong, but wouldn’t they naturally be betting more and of course bringing down the odds of the winning entries. That just isn’t happening! If it is, then the ACTUAL percentages once again prove the point I’m trying to make. So maybe there are only a limited number of players using the softwhere, and those who are don’t have enough confidence to be betting enough to really impact the betting pools. Who knows?

In a para-mutual game where everyone is betting against each other (which is another FACT), what possible advantage could someone gain by selling a product that would permit another competing player to arrive at the same conclusions? Wouldn’t they naturally be betting the same entries and driving down the odds value of the winning entries? (This is if they were in fact winning) So here again we can conclude any number of things: Either the softwhere is in very limited use, or the thousands of players using it aren’t betting very much, or if is being used by a multitude of players that their results are not very good since the percentages mentioned haven’t changed at all.

I will say this in defense of any proven winning system that might be publicly available or not: That its my humble opinion that many people tend to sway very easily, lack self-discipline to follow through, and even find it difficult to arrive at similar conclusions when using identical methods. Why is that? I guess it’s the human experience and the desire to want to accomplish something like cashing a winning bet on their own. How many times have you tried to explain why this entry or that will do well and produce a winning play? Only to be ignored because they either had a better idea or lacked your confidence.

But not to digress any more I think the author of this thread had something more in mind by titling it: "Software: what's missing, what's next?" Maybe its not what I’ve mentioned, but its certainly not an argument for demonstrating a solid approach to this game.

[RichieP]

it’s a FACT that the percentages of the 1st and 2nd betting choices producing Winners hasn’t changed one iota since the handicapping software was introduced back in the 80’s.

All due respect the % has changed on the favorite winning races from 33% to currently just a shade below 36%. That is a significant change.

To say "the fav wins a third of the races and has for decades" is no longer valid.

BTW I learned this from probably the most complete player I have seen on boards Dan G. He gets the stats from his software:)

I am out of the racing game(betting,watching,supporting) for all intent and purpose so I truly have no dog in the fight.

Best of luck and next time you have a cup of coffee remember that:
"Chock Full of Nuts is the heavenly coffee"
:ThmbUp:

[DJofSD]

No actually Dave I’ve stated the “facts” all right as hard as they might be to swallow. But for someone to say that another “knows none of these things” is purely a subjective judgment call that’s based on who knows what? You don't know me from Adam! I'll just say that I'm not one for BS-ing anyone especially those who are interested in getting a handle on this game. I have nothing to gain by dispelling notions of grandeur that softwhere is the answer to making long-term profits.

As far as increasing the long term ROI goes, it’s a FACT that the percentages of the 1st and 2nd betting choices producing Winners hasn’t changed one iota since the handicapping softwhere was introduced back in the 80’s. In fact, all of the calculated speed, pace and trip numbers (the scientific approach) haven’t had a positive impact either. Of course there are exceptions to the rule for some using figures, computer generated or not. But these players are certainly in the minority. If the softwhere had any type of real long-term impact the result of the betting pools should likewise reflect a change. That is if those using it had any real faith in what it produced. I might be wrong, but wouldn’t they naturally be betting more and of course bringing down the odds of the winning entries. That just isn’t happening! If it is, then the ACTUAL percentages once again prove the point I’m trying to make. So maybe there are only a limited number of players using the softwhere, and those who are don’t have enough confidence to be betting enough to really impact the betting pools. Who knows?

In a para-mutual game where everyone is betting against each other (which is another FACT), what possible advantage could someone gain by selling a product that would permit another competing player to arrive at the same conclusions? Wouldn’t they naturally be betting the same entries and driving down the odds value of the winning entries? (This is if they were in fact winning) So here again we can conclude any number of things: Either the softwhere is in very limited use, or the thousands of players using it aren’t betting very much, or if is being used by a multitude of players that their results are not very good since the percentages mentioned haven’t changed at all.

I will say this in defense of any proven winning system that might be publicly available or not: That its my humble opinion that many people tend to sway very easily, lack self-discipline to follow through, and even find it difficult to arrive at similar conclusions when using identical methods. Why is that? I guess it’s the human experience and the desire to want to accomplish something like cashing a winning bet on their own. How many times have you tried to explain why this entry or that will do well and produce a winning play? Only to be ignored because they either had a better idea or lacked your confidence.

But not to digress any more I think the author of this thread had something more in mind by titling it: "Software: what's missing, what's next?" Maybe its not what I’ve mentioned, but its certainly not an argument for demonstrating a solid approach to this game.
If you are going to cite long term statistical trends then you have to assume all other variables are fixed and known. I don't think you can say that when it comes to handicapping and gambling.

First and foremost, you are using broad statistics that apply to a general population and inferring the subpopulation using computers can't be doing very well since the stats haven't changed over time. Maybe the group not using computers has gotten worse while those that use computers have gotten better. Taking a broad average over a long period of time is going to mask the improvements in the computer group.

Next, you assume that any and all software used in the handicapping process is picking the winners. While I'm sure there are programs that do that however you can not lump them in with other programs that do not pick the winner. Just because I am using a computer does not mean the computer is making all of the decissions.

Yes, it is a paramutual game but to say that if software was contributing to success of those playing the game both the average winning percentage would be going up and the average payout would be going down assumes too many things. Not every one betting into the pools is a whale. There could be those using software to make a profit but are happy with a level of income that does not impact the pools in a negative fashion.

Finally, you like to assert statistics as facts. You need to tie the statistics to a well reasoned theory that can not be refuted. Then, maybe, you can say with certainity that computer aided handicapping does not improve the profits of those that use computers.

Good luck, I'll continue using my computer software.

[DanG]

In the interest of bandwidth I won’t quote Nitro directly as several good posts before this have done so already. Real quick and hopefully someone will come up with some great suggestions for Tiller’s thread title.

I. As RichieP stated; this is simply not the case concerning favorites…
rAODD YEAR FLD Plays Wins Win WROI $AVGM TREND
1 2005 8.2 53210 18267 34.3% $0.82 $4.78 XX
1 2006 8.1 57221 19725 34.5% $0.82 $4.75 0.1%
1 2007 8.2 58283 20223 34.7% $0.82 $4.72 0.4%
1 2008 8.2 57205 19975 34.9% $0.82 $4.71 0.6%
1 2009 8.6 7474 2637 35.3% $0.84 $4.76 1.0%

II. One of the most miss-understood aspects of “software”; My favorite example comes from a pub in England after explaining HTR to a man he stated; “I could never let a machine make my selection”. This is not the role of any good software I’m aware of. Check the advertisers on this board and / or HTR and you will see races bet in many different ways depending on user interpretation.

The same is true of the grandmother using her hat pin / the sheets / trip players or DRF. That leads to the next misconception; If you use DRF print version for example…you are using…”software”. Software is used in data collection from Equibase, speed ratings, par times, beaten lengths, formatting print outs and ultimately the formulator that just allows a more customized approach… The word “software” is just entirely misunderstood in terms of application.

III. With all due respect your confusing “system” with “organized information and proprietary ratings.” How and why would anyone market a “system” that only allowed one interpretation? As you said; this is not possible in a pari-mutuel environment.

Back to Ron Tillers original thread;

The future of software is in very good hands. The people who dominate this niche market are very smart / dedicated people who take their work seriously. The data collection will be the next major breakthrough imo. I realize the logistical and contractual issues with “live” data; but real time information via download will take whoever is able to provide it to the next level.

• Be it same day shoe / medication / bandage data.
• Real time result information that continually updates a key race file.
• Live propagating of speed figures.
• Real time / nation wide bias alerts.
• Somehow sitting down all the chart callers in one room and establishing a standardized notation system.
• However; the real breakthrough will come from something that never even occurred to us and the beautiful part is…it will probably be incredibly simple. I sincerely hope ‘Nitro makes the discovery and this thread will have a fitting / yet ironic last chapter. :ThmbUp:

[InFront]

I started handicapping the horses in the early 80's mainly paper/pencil capping and by the early 90's began using all sorts of new computer software that hit the market. From the early non-database programs where you had to manually enter data to database programs that were able to download tons of data automatically once the Internet started to take off. During these years I have tried many of these database software programs even some of the most expensive ones that are still marketed today. I now currently use a customized database software which was programmed by a fellow computer programmer. It has been created over years time to my exact criteria and believe it to be the most advanced software I have ever seen. It does things and has features that I have never seen in any other marketed software. It is basically a creation of all the powerful features that other software I have tested over the years has or at least features that I found somewhat useful with countless other features I have thought of myself. So the software I use is most of all of the good stuff rolled into one huge package.

With that being said the one weakness if you want to call it that with all marketed software I thoroughly tested including my very own customized program is that one can easily take huge databases and find things that showed "past profits" even through large samples. But at the same time that 99.9% ( don't want to say 100%) of those same things fail "forwardly" through new data. It's almost to the point that you start believing that past performances and data is useless to profit from this game. And can drive you nuts at times.

Now when I say that what I mean yes you can take any simple factor such as play the favorite, horse ranked 4th on last race speed, ranked 1st in class rating, won last race, mlo at 5/1, ranked 1st in some proprierity formula, etc. and query such items through a large database. And when forward testing such things they will hold almost exactly as far as win% and roi% goes even when tested forwardly through brand new "un-tested" data. It is almost uncanny. But since such factors show consistent losses doesn't help us much from a betting standpoint.

But it's when you create a algorithm combining all kinds of factors, rankings, proprierity formulas, etc. in creating such a nice profitable winning formula you can easily find tons of such angles that will profit greatly through "past data", that is the data you used to create such an angle. It's those profitable algorithms that almost never hold, not even close, when tested forwardly through new data. And since I believe this is the ONLY way to be sure that you have a "true winning formula or angle" so you can wager on it into future races it is a very important to overcome this problem beforing doing do. I have seen this same occurence happen thousands of times for the past two decades regardless what software I was using at the time marketed or private.

So the question is why is this? If a stand alone losing factor or angle holds constant through any database forwardly why would a winning profitable angle not do the same or at least some of them? So basically what I like to see is software that can overcome and battle this backward/forward querying effect but don't think there is any logical way or equipped software program that can ever overcoming this major flaw and problem.

[Overlay]

I haven't read through the whole thread line-by-line, so pardon me if this has already been brought up (or if it hasn't been brought up, but is unworkable on its face).

What about software that could compare the past performances of the horses in a race against each other based on the results of a large number of past races, and arrive at an accurate projection, not (or not only) of what each horse's chance of winning is, but of the final odds that the public will in fact send each horse off at (including the effects of last-minute or after-the-bell additions to the pool), based on past pari-mutuel behavior? Is such a concept even possible? Would the same backfitting concerns arise as with most systems based on past data? (I'm trying to think of a way that a wager could be made on a horse at any point in the betting cycle, with a high degree of confidence as to what degree of wagering value the horse's final odds would represent.)

[GATEDANCER2]

So maybe there are only a limited number of players using the softwhere, and those who are don’t have enough confidence to be betting enough to really impact the betting pools. Who knows?

I think that sums it up. I've experimented with it and I've tried to reach others that have used computer programs and I come up empty. Maybe I'm out of the loop, but I'm on one other message board and NOBODY there uses software.

I believe the slight increase in winning favorites is because the game itself is more on the up and up. Increased testing, improved photographic technology, etc.

[InFront]

I also will agree that I think even with todays powerful software programs that it hasn't changed the game this much. Yes post favorites have risen about 2% but 2nd choices are similar, average payoffs of winners is very close, horses with best last race speed have similar win%s and roi%s as the old days, etc. So you would think with the wide use of so many software programs out there and these computer cappers pouring money into these pools that such kinds of stats would vary greatly say from 1985 compared to 2005? But as one can see most are very close if not exact.

[Nitro]

If you are going to cite long term statistical trends then you have to assume all other variables are fixed and known. I don't think you can say that when it comes to handicapping and gambling.

First and foremost, you are using broad statistics that apply to a general population and inferring the subpopulation using computers can't be doing very well since the stats haven't changed over time. Maybe the group not using computers has gotten worse while those that use computers have gotten better. Taking a broad average over a long period of time is going to mask the improvements in the computer group.

Next, you assume that any and all software used in the handicapping process is picking the winners. While I'm sure there are programs that do that however you can not lump them in with other programs that do not pick the winner. Just because I am using a computer does not mean the computer is making all of the decissions.

Yes, it is a paramutual game but to say that if software was contributing to success of those playing the game both the average winning percentage would be going up and the average payout would be going down assumes too many things. Not every one betting into the pools is a whale. There could be those using software to make a profit but are happy with a level of income that does not impact the pools in a negative fashion.

Finally, you like to assert statistics as facts. You need to tie the statistics to a well reasoned theory that can not be refuted. Then, maybe, you can say with certainity that computer aided handicapping does not improve the profits of those that use computers.

Good luck, I'll continue using my computer software.Hi DJofSD
Referring to Factual statistics do not involve any variables at all when reviewing the big picture. I will admit that when you view the winning percentages of the 1st and 2nd betting choices by breaking it down on a track-by-track basis it will certainly vary (above and below the average). You can also break it down on a race-by-race basis and find that these percentages rise significantly with 7 or fewer entries. However, I mentioned increasing the ROI. There might be a few heavy hitters that prefer short odds, but for the seasoned player using whatever is at their disposal, I would think that they might prefer making plays that consistently return a decent profit at a lower cost.

I think most of us here are using computers in one way or another. That doesn’t mean we’re using handicapping softwhere. I would think that those not using it have found alternative methods for skinning a cat. Depending on their skill levels, they might also be doing something right to maintain an interest in this game. Sure there very well might be some softwhere advocates that can produce a positive ROI, but just like EVERY other “Outside” group in the betting population they’re certainly in a small minority. And that’s another Fact simply because as we all know only a small portion of this group is in the black.

I’m not sure where I implied that softwhere is doing the decision making, because I think most would agree that no matter what method you’re using, the person using it makes the final call on what to bet and how much to bet. I have no doubt that people using identical methods (no matter what they are) make entirely different selections for play. That’s because even the most sophisticated methodology still leaves room for interpretation.

My purpose for mentioning the type of game we’re playing (para-mutual) was related to implying simple evidence of a conflict of interest. If I’ve got a proven softwhere package that’s providing a solid ROI, why would I want to sell it to you: my competitor? Am I that egotistical and stupid enough to drive my potential ROI down by letting hundreds maybe thousands of other players in on it? Or perhaps I will make enough on these sales to compensate for the loss in ROI. Or maybe I’m just a good-hearted philanthropist who has no worries about maintaining a positive ROI? But then why did I develop the softwhere in the first place? I think it might be a big mistake to assume that the so-called “whale” is relying on softwhere. If I could venture a guess I would also say that they’re not part of the “outside” betting population.

The statistical items I’ve mentioned are indeed well known facts. There are no theories implicated because they don’t involve what bettors might do. Its based solely on what they’ve actually done! And that’s a fact. How many of these bettors are using softwhere or how much they’re pushing through the windows is anyone’s guess. The point is that you have a finite end result to any race. There are only so many possible outcomes. So if one method points you in one direction versus another, the idea is to use something that’s as versatile and adaptable as the conditions of the races are variable. In my opinion I just don’t believe that any current softwhere package operating on a single platform of information has what it takes to deliver the goods over the long haul.

By the way, I really like your quote from Albert Einstein. It’s something that softwhere users might even consider when dealing with the realities of the racing game. It’s not all about numbers!

Good luck to you!

[Tom]

What do long term trends about the crowd as a whole have to do with anyone
using a program?

I don't care what percentage of favorites are wining now and then, the only results that mater are my own, and using software has greatly improved no only my W% and ROI but the number of races I can play as well.

[DJofSD]

I wanted to make an attempt to draw a parallel between the state of handicapping software and the history of physics.

When Issac Newton came up with F=ma, it became the basis for orbital mechanics (Newtonian mechanics) used to predict the orbits of the planets. But after improvements to observations and enough data had been gathered, it was discovered the planet Mercury was showing differences from the predicted orbit.

It was not until Albert Einstein's theory of general relativity was used to explain the difference did we become aware that F=ma can not be used to solve all problems. F=ma is good enough for every day experiences but when you are exploring the fringes, it falls short.

I believe that most handicapping software is of the F=ma variety. It is good enough for most but not all situations.

In this discussion, the state of handicapping software and what is missing, are we looking for a new theory of handicapping or are we just looking at the current state of affairs to point out the obvious: it does not work in all cases?

[garyoz]

Handicapping is not physics. It is a statistical array of probabilistic outcomes, not laws of nature.

[DJofSD]

Handicapping is not physics. It is a statistical array of probabilistic outcomes, not laws of nature.
Handicapping is not physics. OK.

What about the horses when they're running? Does physics apply then?

[GATEDANCER2]

Tom, what software do you use? I'm experimenting, haven't committed to one yet.

[Tom]

HTR, CJ, and Val4.

[DJofSD]

Tom, and here I was thinking you were still using KGEN. Silly me! :)

[Tom]

I would, but it is manual entry.
I do use it for the triple crown and BC races.
Nothing like it!

[GATEDANCER2]

I'm not familiar with these programs. The obvious question is why do you use 3 different programs? do you use data from all 3 to handicap 1 race? seems very time consuming. i've experimented with some of the commercial ones like all-ways and nuerax but haven't felt comfortable with them. what advice would you give to a software rookie that's been handicapping for 40 years the old fashioned way.

[traynor]

<snip>Jim Cramer kind of went about it obliquely - he sat down and mad a list of 100 questions he wanted answered in deciding what bet(s) to make in a race and which horses to bet. These were not necessarily 100 questions he consciously asked himself - after all, he tended and tends to handicap very quickly. But these questions were never far from the surface and although he may not have explicitly asked himself question #78, if asked, he could probably give a literate and informed answer. </snip>



Experts are rarely able to explicitly verbalize the processes they use to reach decisions. A very interesting process called "bootstrapping" is used to reverse engineer the decision processes of experts. Rather than relying on the expert verbalizing his or her decision process, the bootstrap algorithms analyze the relative importance each component of the decision matrix that the expert must have used to reach his or her correct decisions.

The advantage is that after a number of decisions, the process used by the expert can be reduced to a series of algorithms that in nearly all cases produce results (more correct decisions) that are better than the decisions of the expert modeled.
Thanks

[Nitro]

Handicapping is not physics. It is a statistical array of probabilistic outcomes, not laws of nature.You’re right “handicapping is not physics”, but the actual entries in every race experience the basic laws of physics in every event (not of the quantum type at all). Handicapping simply attempts to measure everything that takes place. The “statistical array of probabilistic outcomes” is something that a softwhere program might try to solve, but is more often then not defeated because it can’t incorporate those very basic laws of nature at the time that they’re actually put into motion. (Weather forecasting weather might be good analogy.)

[Nitro]

I'm not familiar with these programs. The obvious question is why do you use 3 different programs? do you use data from all 3 to handicap 1 race? seems very time consuming. i've experimented with some of the commercial ones like all-ways and nuerax but haven't felt comfortable with them. what advice would you give to a software rookie that's been handicapping for 40 years the old fashioned way.GateDancer2 I would think with 40 years under your belt, you already have the foundation for a better understanding of this game then many a softwhere user. (and maybe even developer). Like most of these users you’re probably looking for short-cuts to profitability. I would suspect that your present discomfort with using it might stem from the realization that you feel it’s too mechanical for a game that’s operating on many levels of reality. Perhaps you feel it lacks the intuition or common sense of a human mind. Maybe you feel it doesn’t adequately address some of the many things that you’ve learned about the game after all these years. You’ve probably learned them through first hand experiences and can react to any immediate changes. The softwhere might learn from past statistics gathered and from a comparison of what it produces versus what the outcome actually is. Maybe at some point it will make adjustments to weighing the critical factors it uses to make predictions in order to achieve better results. Unfortunately many things that take place in a horse race can’t be measured before or after the race, particularly a horse’s energy level.

A simple analogy to consider would be to think about all of those great programs written for playing the game of chess. Why is it that a human mind can still beat even the best of these programs? It’s because in many cases these programs can’t anticipate what the player will do next for a future move. It will only react to what the player has done with a past move. The sofwhere will try to develop a statistical advantage based on all of the finite moves available at any point in the game and then use its internal logic to make its next move. On the other hand a chess master will succeed by anticipating the computers next move because he knows how the softwhere will react to every move he makes. You can appreciate this better when you use one chess program to play another. The outcome is about the same as using more then one softwhere package to play the horses.

My feeling is that using softwhere does give the end user a plausible excuse for not blaming themselves for any shortcomings found in their personal results. They simply blame (or cheer) the one or more of the softwhere packages they might be using. Much the same way that an angle player will rationalize the shortcomings of his end result by noting that the winning entry was covered by some other angle. While this is good for the ego, it might not be so good for the bankroll. After all most players can only hedge their bets so far in order to achieve a profitable result.

[ryesteve]

A simple analogy to consider would be to think about all of those great programs written for playing the game of chess. Why is it that a human mind can still beat even the best of these programs?You're living in 1996. The current generation of chess programs are so strong, in matches against grandmasters the software needs to be handicapped by either starting a pawn down, playing black in every match, having draws count as wins for the human, or a combination of all three.

But hey, you haven't let facts get in the way of your anti-software diatribes so far, so why start now?

[JustRalph]

My feeling is that using softwhere does give the end user a plausible excuse for not blaming themselves for any shortcomings found in their personal results. They simply blame (or cheer) the one or more of the softwhere packages they might be using. Much the same way that an angle player will rationalize the shortcomings of his end result by noting that the winning entry was covered by some other angle. While this is good for the ego, it might not be so good for the bankroll. After all most players can only hedge their bets so far in order to achieve a profitable result.


you couldn't be any more wrong. How about ill informed?

you obviously know nothing about the very good software that is out there now. And you make assumptions about how players use it...........

[Nitro]

You're living in 1996. The current generation of chess programs are so strong, in matches against grandmasters the software needs to be handicapped by either starting a pawn down, playing black in every match, having draws count as wins for the human, or a combination of all three.

But hey, you haven't let facts get in the way of your anti-software diatribes so far, so why start now?It seems to bother you that humans can still compete at all!? The facts I referred to might indeed be dated, but the analogy still has some validity since both games involve a decision process. Other then that, the intricacies of both games are worlds apart.

If I’m not mistaken, the title of this thread is “Software: What’s Missing What’s Next”. Apparently some seem to think that there’s nothing missing. Sorry if my discourse seems offensive, but I was simply attempting to point out What’s Missing and What may never come next. But who knows?! Someone will come up with an even better explanation as to why this did or didn’t happen - after the race is over.

[Nitro]

you couldn't be any more wrong. How about ill informed?

you obviously know nothing about the very good software that is out there now. And you make assumptions about how players use it...........
I’m not so sure about that.
Nearly 27 years ago I trained the best OCR program available to actually read the past performances and race results right off the Form. This was accomplished using an $1800 Microtek scanner. (quite a cost difference from today!). All of this data was further automatically screened, error proofed, delimited and entered into a rather large database for use with a few home-grown programs (my personal softwhere package). All of this information permitted a nicely automated source for developing track variants, trainer/jockey stats, horse speed and running profiles and the like. From what I’ve seen that’s available today the newest softwhere simply operates on larger, faster machines with all of the bells and whistles and sophisticated graphics, but the bottom line really hasn’t changed. Why? Maybe its because the basic game hasn’t changed all that much.

I also know a little something about how people use new toys. I learned first hand! A long time ago I would have probably defended my softwhere as well, by noting that I was profitable. I can only say from experience that there are superior methods to attack a game that is far removed from the stats built on bits, bytes and logic. Whether anyone wants to believe that is entirely up to them, but I guarantee that there will always be unknowns that will haunt and frustrate the most avid softwhere user. The only way around this is in my opinion is only play on the strengths of what you’re using and accept the lower value plays that everyone else seems to find.

[DJofSD]

Nitro, in all seriousness, you are probably the best person with good, perhaps very good, credentials to evaluate handicapping software programs. You obvious have no "favorites" and no axes to grind. And having tackled the then arduous task of using a scanner and OCR software to create digitized PP's, you know what the computer is capable of doing.

[fmolf]

i believe whatever method a person uses that is successful for him is all well and good.....people who put their faith in computer programs to handicap is good for them......i only use the computer to download the pp's....i have tried formulator an do not like it......too much information ........it is overwhelming ...and you still have to pick out which information is pertinent and usable and this is the essence of handicapping.....i ask all of you recreational/serious handicappers out there...what is the fun of inputting #'s and taking a computers word.......is the program speed orientated ...class orientated ....pace orientated...one still has to choose.......i would be willing to try to use a program to give me overlayed exacta combos...but in the minutes leading to post the amounts of money bet oneach combo or the odds on each horse would need to be inputted...not practical.....nothing beats a day at belmont ...arrive when it opens with a thermos of coffeee sit on a picnic table and dissect the races.....i guess i'm a bit old fashioned.....

[traynor]

<snip>I can only say from experience that there are superior methods to attack a game that is far removed from the stats built on bits, bytes and logic. Whether anyone wants to believe that is entirely up to them, but I guarantee that there will always be unknowns that will haunt and frustrate the most avid softwhere user. The only way around this is in my opinion is only play on the strengths of what you’re using and accept the lower value plays that everyone else seems to find.</snip>

The underlying fallacy in the logic of handicapping software is that meaningful comparisons can be made using the available data. While that belief is convenient for software developers (myself included) it is conceptually impoverished. That is, most comparisons are based on the false assumptions that each entry is performing to its maximum potential in every race, and that meaningful (a euphemism for "profitable") decisions can be made on the basis of those comparisons.

Rather than simply crunching Equibase numbers in YASR fashion (yet another silly rating), the future of computer race analysis is likely to be a synthesis of available and proprietary information. That is, a blend of the Equibase numbers modifed by "something else" (for example, trip handicapping, onsite physical inspection of "body language" and appearance, or something else entirely).

In a closed system in which everyone is massaging the same basic set of numbers, it is not especially difficult to understand why reliance exclusively on computer readouts as decision tools is not especially profitable.
Thanks

[Buckeye]

It really comes down to "throw it out" or not-- (the program's opinion) about a given horse and sometimes that horse period.

modify it, and it still comes down to me you and maybe eight other people standing there with a Racing Form with ten minutes to post. Who do you like?

Money gets exchanged in real time. Races happen in real time.

Software would need to replicate this kind of analysis to move forward.

Good luck.

[Warren Henry]

A long time ago, I was one of the first to actually make a living selling handicapping software. I also played the horses pretty seriously.

I have been lurking for a while and finally decided to join so that I could offer answers/opinions from what may be a different perspective.

I realize that since this is my first post, I will lack general credibility. However, I am not particularly bashful, so I will barge right in.

Many skeptics ask "Why would you sell a winning method?"

There are a variety of legitimate answers for that question. For me, it was an easier way to make money selling software than to handicap every day, drive 40 miles to the track, etc. I could stay in my jammies most of the time and only leave the house a couple of times a week to take the orders to the post office. Handicapping is hard work even with good tools. I discovered early on that I didn't have the energy to run the busines properly and handicap profitably at the same time. Running the business was easier.

Another answer is that the software I sold could be considered as tools. I was pretty good at producing the tools, but not as good at using them. I was often congratulated at the track by a customer who had just cashed a large bet using my tools when I didn't have a dime on the horse in question. They had different ways of using the software than I did. Two examples -- One guy cashed a nice bet on a horse using a pace line from a year earlier because he knew that the trainer in question always had his horses fit and ready for the beginning of this particular meet. I didn't know about the trainer and judging from the price he got, neither did most of the other bettors. Second example -- a mentor of mine used the same program that I did to produce speed ratings. However, he did his own adjustments to the track variants. At the end of the year, we cashed the same percentage of bets, but his averaged a dollar better than mine per two dollar wager. Why, because he used data that was unknown to me - thus had more value.

Some feel that since we are all looking at the same basic data, we must all arrive at similar results. I disagree. Our results will differ because we look at the data in different ways.

Some feel that because the percentage of winning favorites remains relatively constant over time, it means that use of software hasn't changed the game. I both agree and disagree with this view. The basic nature of the game has not changed. However, the player who changes the way he operates, may significantly alter his/her hit rate (or more importantly ROI).

My tip for the day -- Do what the other guy isn't doing. Since the rage today seems to be speed/pace, why not bet based on CLASS. I believe (but haven't proved) that doing a rating based purely on purse size class analysis. will produce a fair number of winners and their price will be way better than you will get trying to produce a better pace/speed number.

What is next Probably more of the same. Fancier software with more bells and whistles doing the same old job of trying to improve existing factors. What I call evaluation of WHAT HAS HAPPENED.

What is missing? The ability to evaluate WHY IT HAPPENED. My future efforts (for my own use) will be in the area of pattern analysis. Since I want good prices, I will probably develop tools to see if I can find patterns that will tell me WHY a certain horse wins at a good price, or why a certain time of the year at a certain track always seems to produce a higher percentage of longer priced horses. Can I find a pattern that has produced winners at a price and profit from that pattern going forward?

Enough for now.

[Dave Schwartz]

Since Mr. Henry is so now and, (to use his phrase) lacks credibility, I'll stick my two cents worth in here and admit <G> to knowing Warren.

A finer man I have never met. Of course, I never actually met him either but we had many telephone conversations years ago.

The information he offered was highly valued by me.

Warren - call me when you egt a chance. I'd love to catch up. (But not until after UConn beats Missouri today.)

Regards,
Dave Schwartz

[DJofSD]

Finding patterns has always be the key. Not becoming slaves to them is the trick.

[ryesteve]

It seems to bother you that humans can still compete at all!?Not at all... what bothers me is when people continually misstate "facts" as a means of supporting provocative arguments.

the analogy still has some validity since both games involve a decision process.I'd be happy to concede it's a valid analogy, since the current state of the art of the chess software contradicts your underlying argument.

Anyway, it's not as if you're incorrect in some of the points you're making... the problem is that the conclusion at which you're arriving is too extreme. And honestly, I'm not convinced you understand as well as you think you do, the capabilities and approach of software today, compared to what was being done 27 years ago.

[DanG]

Post #53: :ThmbUp:

I hope you write more often Warren; interesting post.

[Nitro]

It really comes down to "throw it out" or not-- (the program's opinion) about a given horse and sometimes that horse period.

modify it, and it still comes down to me you and maybe eight other people standing there with a Racing Form with ten minutes to post. Who do you like?

Money gets exchanged in real time. Races happen in real time.

Software would need to replicate this kind of analysis to move forward.

Good luck.Some great insight there!
But unfortunately there are more involved then you, me and eight other people with Racing Forms in the betting population.

[Nitro]

A long time ago, I was one of the first to actually make a living selling handicapping software. I also played the horses pretty seriously.

I have been lurking for a while and finally decided to join so that I could offer answers/opinions from what may be a different perspective.

I realize that since this is my first post, I will lack general credibility. However, I am not particularly bashful, so I will barge right in.As far as I’m concerned your credibility is self-evident based on your forthright responses to many of the items that I’ve previously mentioned. Your perspective only differs because you sold your software and I would never sell mine.

Many skeptics ask "Why would you sell a winning method?"

There are a variety of legitimate answers for that question. For me, it was an easier way to make money selling software than to handicap every day, drive 40 miles to the track, etc. I could stay in my jammies most of the time and only leave the house a couple of times a week to take the orders to the post office. Handicapping is hard work even with good tools. I discovered early on that I didn't have the energy to run the busines properly and handicap profitably at the same time. Running the business was easier. Your response here corresponds to what I mentioned previously, but my comment is not based on skepticisim. It’s based on common sense.

My purpose for mentioning the type of game we’re playing (para-mutual) was related to implying simple evidence of a conflict of interest. If I’ve got a proven softwhere package that’s providing a solid ROI, why would I want to sell it to you: my competitor? Am I that egotistical and stupid enough to drive my potential ROI down by letting hundreds maybe thousands of other players in on it? Or perhaps I will make enough on these sales to compensate for the loss in ROI.

Another answer is that the software I sold could be considered as tools. I was pretty good at producing the tools, but not as good at using them. I was often congratulated at the track by a customer who had just cashed a large bet using my tools when I didn't have a dime on the horse in question. They had different ways of using the software than I did. Two examples -- One guy cashed a nice bet on a horse using a pace line from a year earlier because he knew that the trainer in question always had his horses fit and ready for the beginning of this particular meet. I didn't know about the trainer and judging from the price he got, neither did most of the other bettors. Second example -- a mentor of mine used the same program that I did to produce speed ratings. However, he did his own adjustments to the track variants. At the end of the year, we cashed the same percentage of bets, but his averaged a dollar better than mine per two dollar wager. Why, because he used data that was unknown to me - thus had more value.
Once again you’re offering additional confirmation of my previous comment.

I’m not sure where I implied that softwhere is doing the decision making, because I think most would agree that no matter what method you’re using, the person using it makes the final call on what to bet and how much to bet. I have no doubt that people using identical methods (no matter what they are) make entirely different selections for play. That’s because even the most sophisticated methodology still leaves room for interpretation.

[Nitro]

Some feel that since we are all looking at the same basic data, we must all arrive at similar results. I disagree. Our results will differ because we look at the data in different ways.There is only so much worthwhile data that’s available to begin with. Most that have attempted to draw any substantive conclusions from the manipulation of this data usually have the same goals in mind. The only difference I see is how certain information is “weighed” in terms of importance to reaching that goal. The problem is that most programs are too rigid to accommodate the dynamics of the ever-changing conditions of each race as well as it participants. I believe Mr. Traynor’s comments are very applicable here: In a closed system in which everyone is massaging the same basic set of numbers, it is not especially difficult to understand why reliance exclusively on computer readouts as decision tools is not especially profitable.



Some feel that because the percentage of winning favorites remains relatively constant over time, it means that use of software hasn't changed the game. I both agree and disagree with this view. The basic nature of the game has not changed. However, the player who changes the way he operates, may significantly alter his/her hit rate (or more importantly ROI). A player may improve the way he operates without the use of any softwhere at all. Hit rate is only one side of the coin. Without an adequate return on those hits the ROI suffers because they won’t compensate for the losses. I would think that basic money management would take a priority here.


What is next Probably more of the same. Fancier software with more bells and whistles doing the same old job of trying to improve existing factors. What I call evaluation of WHAT HAS HAPPENED.I agree completely!

[Nitro]

What is missing? The ability to evaluate WHY IT HAPPENED. My future efforts (for my own use) will be in the area of pattern analysis. Since I want good prices, I will probably develop tools to see if I can find patterns that will tell me WHY a certain horse wins at a good price, or why a certain time of the year at a certain track always seems to produce a higher percentage of longer priced horses. Can I find a pattern that has produced winners at a price and profit from that pattern going forward? It happened for no other reason then the outsiders inability to secure the same information that the insiders already have. They also seek good prices. The horse does well because the insiders place their properly conditioned animal into a race where the potential conditions are ripe for a score. There are developed patterns that I currently use on a daily basis, but they’re no longer related to the things that have happened to race contestants in the past. This historical patterned information only provides a baseline for looking at the potential of future events.

Since Mr. Henry is so now and, (to use his phrase) lacks credibility, I'll stick my two cents worth in here and admit <G> to knowing Warren.
A finer man I have never met. Of course, I never actually met him either but we had many telephone conversations years ago. I found this comment rather amusing. Warren and I are both saying many of the same things concerning softwhere, yet his comments get a pass because the advertiser is an acquaintance.
Now there’s a double standard pattern I’ve seen before!

[Nitro]

Anyway, it's not as if you're incorrect in some of the points you're making... the problem is that the conclusion at which you're arriving is too extreme. And honestly, I'm not convinced you understand as well as you think you do, the capabilities and approach of software today, compared to what was being done 27 years ago.It might seem extreme, but my explanation for what I did many years ago was simply a way to convey the experiences I had trying simplify the game for myself. Many thought what I was doing at the time was also bit extreme. Some even wanted to purchase what I had then, but I wouldn’t sell it. After all I was on my way to self-made riches. I found those thoughts were also a bit extreme then as I do now when thinking about softwhere in general.

Please don’t get me wrong. I’m not trying to convince anyone of anything! I’m simply trying to express my feelings about a topic that I understand very well. Whether you appreciate where it’s coming from is another story completely. A similar analogy can be drawn from what I mentioned before:

I guess it’s the human experience and the desire to want to accomplish something like cashing a winning bet on their own. How many times have you tried to explain why this entry or that will do well and produce a winning play? Only to be ignored because they either had a better idea or lacked your confidence.

Good luck!

[Dave Schwartz]

I found this comment rather amusing. Warren and I are both saying many of the same things concerning softwhere, yet his comments get a pass because the advertiser is an acquaintance.
Now there’s a double standard pattern I’ve seen before!

Whether Mr. Henry's opinions agree or disagree with mine, he is a man that I respect. Since he chose to bring up the issue of his own credibility, I addressed it.

Respect and agreement are not the same thing, although, to be honest, I haven't actually disected his post - I am not really sure what he said on this topic.


Regarding the drum you are beating so loudly, my opinion differs from yours significantly. If Warren agrees with you then I disagree with him as well.

But for the record, I merely said that you stated your opinions as if they were facts. They are not.

The truth is, your issue this is a non-issue for me.

I have been in this business a long time. Some people win and others lose. The winners will believe that their ways are best and, likely, will support those efforts. The non-winners have nothing to support, so often have nothing positive to say.


Every once in awhile there is a supposedly succesful player that likes to beat his chest and say, "MY way is the only way." When they do, they show their ignorance and lose credibility. Generally, I just chalk it up to a "need-to-lead" and forget it. (Which is what I did with you.)


Regards,
Dave Schwartz

[njcurveball]

There are developed patterns that I currently use on a daily basis, but they’re no longer related to the things that have happened to race contestants in the past. !

I feel like I am reading a page from one of those overstuffed envelopes full of offers for riches I get in the mail.

But just for curiousity sake, HOW can you have DEVELOPED patterns that are NOT related to things that have happened in the past? :confused:

As I sat in front of my bulky laptop, circa 1990, many people would come over and engage our group and explain how "computers cannot pick winners". As they did this holding their program (a computer generated report) and Racing Form (a computer generated report), our group would explain the figures on the screen and how INTERPRETING them is the key.

The ANSWER was in using the human brain. When we would give a longshot winner, they would ignore our advice. "I AINT BETTING WHAT NO COMPUTER SAYS!". They would come back right away and ask "Do you have another?". When the horse would lose they would return even quicker and say "I had the winner (usually the favorite), your computer DONT KNOW NOTHING!"

I enjoy reading threads like this, they show how many people have moved forward and how many people have not. I just ask this simple question, are there any handicapping contests where I can be the ONLY ONE using a computer? If so, please point me to them. :ThmbUp:

[Buckeye]

I feel like I am reading a page from one of those overstuffed envelopes full of offers for riches I get in the mail.

But just for curiousity sake, HOW can you have DEVELOPED patterns that are NOT related to things that have happened in the past? :confused:

As I sat in front of my bulky laptop, circa 1990, many people would come over and engage our group and explain how "computers cannot pick winners". As they did this holding their program (a computer generated report) and Racing Form (a computer generated report), our group would explain the figures on the screen and how INTERPRETING them is the key.

The ANSWER was in using the human brain. When we would give a longshot winner, they would ignore our advice. "I AINT BETTING WHAT NO COMPUTER SAYS!". They would come back right away and ask "Do you have another?". When the horse would lose they would return even quicker and say "I had the winner (usually the favorite), your computer DONT KNOW NOTHING!"

I enjoy reading threads like this, they show how many people have moved forward and how many people have not. I just ask this simple question, are there any handicapping contests where I can be the ONLY ONE using a computer? If so, please point me to them. :ThmbUp:

Gee whiz, it comes down to what?

It comes down to Not computer programs that's for sure. Buy one you like and then ignore what it says most (or at least MOST) of the time. If any of these programs had a clue it would be all over. $2.20 to win. They simply can't do it.

[Warren Henry]

My apologies to the board. My real reason for coming here was to look for some tools to help me automate my own data acquisition and bet placement. I did not mean to start any arguments.

BUT.....

My following comments will be more general than software oriented. I promise that my next post will concern software and what I am seeking in that area.

I don't see the analysis of the horse racing data as a closed end proposition. While the past performance data may well be the same for all of us, there are so many different ways to look at this data that we can stay busy for a long time looking for new ways to be profitable. Being profitable is the goal. Picking winners is secondary at best.

What I believe to be true is that doing what the other guy is doing will not lead to profitability. If everyone in the universe is trying to tweak speed numbers, speed by itself will not produce profitability. As I said in my first post, the key to profitability is to bring something to your play that the others are not doing. Whether this is new thinking or whether it is recycling out of favor methods from the past doesn't matter.

I collect racing books. I have most of the current works, but what I find most interesting are the old books. Many of the techniques used "back in the day" worked and still work. I believe (and will attempt to prove for my own profitability) that one can use tried and true handicapping methods from the past and profit from this game.

An analysis of the systems of the past shows that the handicapping game cycles. These cycles are probably based on profitability - both to the bettors and to the system sellers. To restate the point that I was trying to make earlier. If most of the bettors today are playing speed, play class for more profits. When the bettors switch to class or trainer patterns, go back to betting speed or pace.

We have not begun to exhaust the different ways of looking at the data. Most players concentrate on attempting to isolate the eventual winner. What would happen if you concentrated on losers? You could concentrate on finding reasons why a favorite might be a false favorite. Or just find a set of horses not likely to win a given race and dutch the contenders if one or more of those horses was getting serious consideration on the board.

In my opinion, computer software is just one of the many tools that a player may use to assist in his work. I do not believe that there is a universal RIGHT WAY to bet horses. The only thing that truly matters is whether or not the method you use is profitable for YOU.

[Buckeye]

well yes profitabilty is the key. In a pari-mutual system it would seem unlikely that anything you could "purchase" would do much that for you. Why they're selling it would be because most likely it's not worth much on it's own.

[Dave Schwartz]

well yes profitabilty is the key. In a pari-mutual system it would seem unlikely that anything you could "purchase" would do much that for you. Why they're selling it would be because most likely it's not worth much on it's own.

If this were true, then one should not waste their money purchasing books either.

I would contend that books affect mutuels far more than software does. When you consider that the best-selling software packages may wind up in the hands of (perhaps) 1,000 people over its entire lifetime, a single book (such as Winning at the Races or Modern Pace Handicapping) will sell fifty times that many. Furthermore, a book will continue to influence play for a decade or more, while software only lasts through a few incarnations of the computer's OS (i.e. Windows).

Ironically, it is the idea that impacts mutuels more than a single application of the idea.

To illustrate the point, consider Qurin Early Speed Points. Do they still "work?" Yes, but at some tracks 8-pt horses are so over-bet that they actually become a worse bet than a 0-point horse.

I really do not have a dog in this fight. I am not going to lose half of my customer base because someone pointed out that "software doesn't work." Most of us know better. Some of you don't. That doesn't make you wrong if you disagree with me. It means that you perceive it differently.

Difference in viewpoint is what makes the world such an interesting place.


Regards,
Dave Schwartz

[Warren Henry]

Years ago (1979-2000), I was in the horse software business. For various reasons, I burned out and left the game. I am retired now and would like to pursue some theories that were marginally profitable in 2000. With a
little improvement and the current rebate availability, I believe that I can develop a series of plays that I can play pretty much automatically and make decent money. I want to do this for my own personal amusement and
profit. I have no desire to get back into the commercial software (or tout) game.

When I left, I quit following the game completely. My friends from the racing days tried to keep me abreast of some of the changes, but ended up giving up on me as "I didn't even read The Form."

After an absence of nearly ten years, I had expected to see that the software sellers would be selling most of the tools that I wanted back then. I am shocked at what I am NOT finding.

I am seeing massive "do all, end all" types of systems, both black box and tweakable. but I don't see the small tools that would allow a player easily automate some of his handicapping tasks.

Where is the tool that will automatically acquire multiple data files for a racing day?

Where is the tool that will parse/extract the data from these files and load a database?

Where is the template tool that will allow a novice to build his own database using the above tools?

Where is the conditional betting engine that will acquire streaming tote data and automatically post bets with an ADW site according to the users specifications? (The conditional betting capability at the ADW sites is not complex enough for the types of bets that I want to make.)

I am really surprised that these capabilities are only available as part of larger suites of software or to those who develop them themselves . Am I the only guy who wants to do my own analysis and my own selections, but wants to automate the process?


While I WAS a programmer, my skill set is not currently what is needed to build the tools that I seek.

In an ideal world, here is what I want to end up with ---

A) Software that will automatically acquire past performance and results data from any/all tracks running at the time. Also software that will push this data into a database.

B) A complete past performance database including lifetime PPs for all horses currently running in North America.

C) Software to analyze the database looking for profitable patterns

D) Software that will analyze the current past performances and automatically select plays based on parameters identified in prior analysis.

E) Software that will monitor the tote and structure bets close to post time and automatically submit those bets to PTC or other ADW sites.

F) Software that will monitor the results of the wagers and report profit and loss statements.

I plan to do the code for the analysis of the database, the analysis of the current days data and structuring of the plays for the day. I can also do the wager tracking and reports.

Where I need the help is in automatically acquiring the PP and results files and preparing the data for load into the database.

I also need help in producing streaming tote data that I can use for betting decisions.

Last, I would like the capability to push those bets
automatically into one or more online betting sites at x minutes to post.

If necessary, I could probably cobble all of these tools together myself, but I would prefer to concentrate my efforts on the research, and daily wagering decision software.

If these tools are readily available, would someone please point them out to me. If they are not commercially available, would any of you be willing to share concepts/techniques (not necessarily code) with me?

[InFront]

In my opinion based on past research is that the info found in the following things DON'T work: books, paper systems/methods, public newspaper choices, tipsheets, speed sheets or any other similar type of products, tout selections and yes "canned" software programs.

If any such things ever had a chance to work and show profits once they become or became public knowledge you can be rest assured most are now useless. But also believe that most don't work in the first place. As the ole' saying goes "If it doesn't work and you are very sure it doesn't work, then sell it".

Now what may work or should I say at least gives a handicapper a chance is software that is very customizable, can access any of the 1,500+ data fields available, has some decent proprierity formulas built in it and can query large size databases. While these are also products that may be available to the general public cause of their "wide difference in use" and having querying capabilities I believe they the ONLY PRODUCTS any new or old handicapper should invest in. And of course also investing in the data files this software accepts. Now I'm not saying you can even win with any of these software programs. What I'm saying is I think it gives you the best "slight chance" over all of these other sold racing products I mentioned.

One more thing it is OK to buy a "real good book" once in awhile. But this is only for the purpose that it may trigger some different or new ideas that you can use and test in your own database software.

[njcurveball]

Gee whiz, it comes down to what?

It comes down to Not computer programs that's for sure. Buy one you like and then ignore what it says most (or at least MOST) of the time. If any of these programs had a clue it would be all over. $2.20 to win. They simply can't do it.

I am always confused by responses like this. "Buy one and ignore it"? Ignore what? $2.20 to win? How is that going to happen? Racing is a great game, but betting does not have some unification purpose. Look at Desert Party in Dubai yesterday. He had beaten Regal Ransom two races in a row on the same track.

Is Regal Ransom a "pick" to win? Of course not. Is he a play at 2-1? Of course not. But if Desert Party is 3-5 and Regal Ransom is 10-1, WHO do you think the smart money players will be betting. Sure the "non techy" friends will win A LOT of races, but the game is about profit and that profit comes from being smarter than the crowd.

Your post shows you have not got past the "there is only one right answer" stage. Many never do. For those a computer program will not help unless it prints out the exact order of finish AND the exact way to bet it.

Those people are in the majority and would never "let a computer pick their horse!" Which is why I love this game! :ThmbUp:

[Dave Schwartz]

InFront,

That is the most intelligent post I have seen in this thread.

Thank you.


Regards,
Dave Schwartz

[RobinFromIreland]

...I am seeing massive "do all, end all" types of systems, both black box and tweakable. but I don't see the small tools that would allow a player easily automate some of his handicapping tasks...I wholeheartedly agree; the lack of specialization tools is surprising.

...Where is the tool that will automatically acquire multiple data files for a racing day?...I'm working on it now - in fact it's not so far away.

...Where is the tool that will parse/extract the data from these files and load a database?...As above. As I'm based in Ireland I'm designing it to that the parser is modular, meaning that you just plug in implementations that suit your needs. Right now, I'm writing one that will mine the Racing Post website. But an implementation that works off data files on your hard drive or perhaps with an API would also possible, and the rest of the application would work the same.

Thus, the database is a standardized format. Therefore queries written for European racing/wagering will be applicable for American racing et al. You could have parsers that just focus on certain tracks, or some that mine every piece of data available. Whatever one you want, you just plug it in.

...Where is the template tool that will allow a novice to build his own database using the above tools?...Once the database schema is decided, it is not difficult to automate its construction.

...Where is the conditional betting engine that will acquire streaming tote data and automatically post bets with an ADW site according to the users specifications? (The conditional betting capability at the ADW sites is not complex enough for the types of bets that I want to make.)...That's obviously somewhat more bespoke but entirely possible.

...A) Software that will automatically acquire past performance and results data from any/all tracks running at the time. Also software that will push this data into a database...No problem.

...B) A complete past performance database including lifetime PPs for all horses currently running in North America...I don't know much about getting historical PP data for the US - is there a website(s) that hosts all historical results? Or do you have to buy and download the data? Expensive? Can the data be privately shared?

... C) Software to analyze the database looking for profitable patterns...This is something rather more difficult and time consuming to design something that would be abstract enough to be flexible while at the same time encourage experimentation. Again, totally possible though.

...D) Software that will analyze the current past performances and automatically select plays based on parameters identified in prior analysis...As above.

...E) Software that will monitor the tote and structure bets close to post time and automatically submit those bets to PTC or other ADW sites...No problem. Just depends on how to tell it when to bet.

...F) Software that will monitor the results of the wagers and report profit and loss statements...A lot of these questions that come up miss the fact that many other industries have encountered similar problems and have built solutions already. You just have to use your imagination about how they can be utilized for horseracing. For instance, for question F), use online banking services like www.buxfer.com (http://www.buxfer.com) and it's API.

...I plan to do the code for the analysis of the database, the analysis of the current days data and structuring of the plays for the day. I can also do the wager tracking and reports...Great - I'd love to hear your ideas.

...Where I need the help is in automatically acquiring the PP and results files and preparing the data for load into the database...I could probably help you here.

...I also need help in producing streaming tote data that I can use for betting decisions...Don't know much about where such data is available. Producing streaming data feeds though is no problem.

...Last, I would like the capability to push those bets automatically into one or more online betting sites at x minutes to post...Should be fine.

...If necessary, I could probably cobble all of these tools together myself, but I would prefer to concentrate my efforts on the research, and daily wagering decision software.

If these tools are readily available, would someone please point them out to me. If they are not commercially available, would any of you be willing to share concepts/techniques (not necessarily code) with me?...Yeah, I could lend a hand and I'd be interested. Publically though. We've had enough of secretive work - it's time to open up and share our knowledge.

Anyway, back to the original topic and question:

What's missing? The social aspect of handicapping. We as handicappers tend to work alone, duplicating our efforts, wasting time looking at the same data. The software so far doesn't encourage collaboration. Crowd source. Team. Ally.

What's next? More collaboration, more analysis, more data (initially manually generated).

[JustRalph]

Now what may work or should I say at least gives a handicapper a chance is software that is very customizable, can access any of the 1,500+ data fields available, has some decent proprierity formulas built in it and can query large size databases. While these are also products that may be available to the general public cause of their "wide difference in use" and having querying capabilities I believe they the ONLY PRODUCTS any new or old handicapper should invest in. And of course also investing in the data files this software accepts. Now I'm not saying you can even win with any of these software programs. What I'm saying is I think it gives you the best "slight chance" over all of these other sold racing products I mentioned.

For those of you who don't think that there are good software apps out there for horse racing, please read the above and introduce yourself to some good software that meets the above standards.

It's out there................and just because you don't use it.......or you don't get it........... doesn't mean somebody else isn't kicking your ass at the windows with it. The sad part is you don't even know it.

Dad used to say........smart guys know what they know.
Really Smart guys know.........what they don't know...........

both types populate this thread............. :lol:

[InFront]

InFront,
That is the most intelligent post I have seen in this thread.
Thank you.
Regards,
Dave Schwartz

Thanks Dave. While I may not always agree with everything you post about handicapping, database research, wagering, etc. one thing I always do is at least read and listen carefully to all your posts. And you are one of the very few members I do this to.

Like I said I have tried many, many software programs over decades and while there may have been some things I liked in each there were also many things I didn't like or should I say not agreed with. This is why I now use only a customized database software that has "most" of the exact TOOLS to my specifications I want to extract and query all kinds of algorithms, ideas, angles, concepts, etc. But feel if any handicapper is gonna spend hard earn money at least spend it on a product that may be useful which in my opinion IS ONLY some kind of database customizable software program. This is for two important reasons:

1. As long as they are very customizable programs dozens of people could still buy them and because they can all use them in so many ways their final handicapped picks based on their own "hard-core" research can be completely different from each user. This is not the case with so many other types of racing products.

2. If something can't be backtested over a large database what the hec good is it. Sure you can read a book or paper method and try it's theory out over a few hundred races manually but that won't mean crap. It seems you still need querying capable database programs regardless what you buy or read. If not you'll never really know what works, what doesn't, what factors or combinations of factors have at least some value, what things the public way overbets, etc.

If the very first ten years of my handicapping I was one of those newbies who bought everything, manually tested everything use the old paper Form, tried my own ideas the paper/pencil way, etc. When I look back at it now it was a waste of ten years. But once I joined the computer database world so many things became much clearer on how to really dissect all these factors looking for value by learning and testing so much more about horse racing in lightening speed compared to the old days. How anyone can win without the use of a "computer and serious database software" is beyond me unless they have some "solid inside behind the scenes information" which I don't believe even exist anyway.

[Dave Schwartz]

InFront,

My experience over the years has been that the majority of people who are capable of writing their own software have a bone in their head that requires them to do so. Nobody else's software will do exactly (or maybe even "remotely close") to what they want.

Simply put, they need to "roll their own." (Which is a phrase that takes me way back to my youth - but I digress...)



1. As long as they are very customizable programs dozens of people could still buy them and because they can all use them in so many ways their final handicapped picks based on their own "hard-core" research can be completely different from each user. This is not the case with so many other types of racing products.

This is what we have tried to accomplish in the HorseStreet Handicapper. The price the user pays for that is in the level of complexity in the product. Alas, we have determined years ago that very few people actually want this. Or perhaps it is better to say that they want it simple and flexible, two adjectives that usually do not go well together.

Of course, most horse players are unrealistic about what they expect from a software product anyway. I once had a user who told me that he wanted 40% winners at an $8.00 average mutuel and a play in every race. Oh, and he thought the program should cost less than $100 and come with free upgrades and tech support for life.

Another time a guy on Prodigy (remember that?), said that any software vendor who charged more than $5 for their software was a crook. After all, the disk only costs $0.50 and getting 10:1 on the money should be plenty. (In fairness, that person later apologized for this when he was educated on the matter, but both Dick Mitchell and I took strong issue with him.)

Perhaps the piece Pièce de résistance of this is when I started a thread here on PA many years ago. The question I asked was, "If I sold a program that would absolutely earn you an income of $20,000 per month, what would you be willing to pay for it?"

I emphasized in the explanation that imagine you were fully convinced that it would do just what I said it would do.

I got a wide range of responses, but the highest was $1,000! Not $1,000 per month mind you, but $1,000 total! And that was the highest offer by double or triple!

So, it is no surprise that players have wild ideas about what software can or should do for them.



2. If something can't be backtested over a large database what the hec good is it. Sure you can read a book or paper method and try it's theory out over a few hundred races manually but that won't mean crap. It seems you still need querying capable database programs regardless what you buy or read. If not you'll never really know what works, what doesn't, what factors or combinations of factors have at least some value, what things the public way overbets, etc.


In our software, we have two ways of going: static or dynamic. In a static handicapping system you have a rigid, systematic approach - a single set of non-changing weights. This can easily be back-tested against your database, even if it is highly complex, encompassing (perhaps) dozens of IV tables or objects.

A very common approach is to have several systems, or as we call them, "analysts." Each analyst specilizes in a particular type of race. As an example, the software ships with a "Fred Davis-type" approach which contains seven analysts:

claiming sprints
claiming routes
allowance sprints (non-claiming)
allowance routes
maiden older
2 yr olds
turf

Now, these are rather high-level systems (analysts). That is, they are "global" in the sense that they are used for all north American tracks. In the real world, if you hired an "analyst" to handicap races for you, logically, one might want some "specialists" to "help" him with the handicapping. These specialists would be experts in a variety of things - perhaps 1st time starters, early speed, jockeys, trainers, form, whatever. They might be specialists on a global basis as the analyst is or perhaps they might be specialists at a particular track or even track-surface-distance.

During the handicapping process, the specialists, or "generic analysts" as we call them, chime in with their area of expertise whenever the situation calls for them to do so. It is a very robust system

One of our professional users actually has over 1,200 analysts which are automatically controlled by the software to "fire" at the appropriate time.

Think of it as 1,200 "models," some at the "primary" level and some at the "generic" level. All this is integrated together, even in the back-testing.



In a dynamic approach, the systems are created dynamically. That is, at the time of the race, a "filtering process" is fired that fetches races "like this one" from the database and the system is built "on-the-fly." Depending upon the complexity of the approach (and the speed of the computer) this process will take from a few seconds to several minutes.

It is simply impossible to test a dynamic system against the database. It would be the equivilent of saying, "Go to the database, and get some races... for each of these races, go to the database and get some races... and for each of these races... Well, you get the idea.

(Okay - it could be done but the coding of such a process would be numbered in the thousands of hours.)


The biggest problem we face is that dynamic approaches are significantly stronger than static approaches, but the only way to test them is one race at a time. Or, more accurately, one race at a time until you have amassed a thousand races or more.


I don't really know why I wrote this... I guess I just got caught up in the moment and wanted you to know that there really is sophisticated software available. (And after I spent 20 minutes writing this I will be danged if I am going to flush it. <G>)

If you'd care to see it explained a little more graphically, here is a PowerPoint that will do so:
http://www.practicalhandicapping.com/desktop/learning/h4upgrade/lesson09/H4Up009_files/frame.htm


Regards,
Dave Schwartz

[Warren Henry]

Dave,

I haven't heard the bone in the head phrase since a Marine Corps Major explained to me that all Majors have a bone in their head that tells them that they don't have to justify their instructions to Sergeants. Never mind that his instructions didn't make sense and that there was no way to implement them in the programs that we were attempting to write. This was about 1968.

We must have had a customer in common back in the early days of personal computers. I too had a guy who thought a 10 to 1 markup on raw media was a reasonable profit. He later bought a tool from me for about a hundred bucks and I bet I put at least a hundred hours into supporting him as he tried to use it. I was never sure if he was yanking my chain or if he really needed that much help.

And, BTW, you are wrong. When they get old and impatient, those who maybe could do it themselves would rather have someone else do it for them (as long as it was done EXACTLY to their specifications) <grin>.

[Dave Schwartz]

I haven't heard the bone in the head phrase since a Marine Corps Major explained to me that all Majors have a bone in their head that tells them that they don't have to justify their instructions to Sergeants.

Col. Sherman T. Potter, 4077th M*A*S*H

[traynor]

<snip>Respect and agreement are not the same thing, although, to be honest, I haven't actually disected his post - I am not really sure what he said on this topic.</snip> Dave Schwartz

Totally off the topic. You use the word "dissect." Hidden behind the term "textual deconstruction" is a wealth of information you might find interesting. You could call it dissecting on steroids.

In essence, it is that the form and format of communication often carries more (and frequently unintended) information than the content. Everyone is aware that "it ain't what you say, it's the way that you say it." Textual deconstruction goes way, way beyond that. I first got into it in a graduate level psychology class that required "coding' conversations; essentially breaking the language down into discrete units, categorizing them, and interpreting the context in conjunction with the content. It was an eye-opening experience.
Thanks

[InFront]

Dave,
Excellent post and great link to read on database research. I have read over your website several times in the past and enjoyed reading much about your own software. But I do admit that is one commercially sold program I never tried mainly cause it didn't use BRIS/TSN data. But just by reading about it I can see it is one of the more powerful but maybe more complicated programs available.

I understand the problem any programmer can run into when developing a program that will be intended to be sold on the market and NOT for their own private use. You can't make it too complicated or no one else on Earth would ever understand on how to use it tons of features. You would be answering calls and emails 24/7. This is why you as a programmer must ALWAYS limit any software to how complicated and powerful it can really be. As with the software I use it would need a manual six inches thick just to go over "most" of it's many features and options and that wouldn't even cover everything.

Now as far as someone offering a software or method that can guarantee $20,000 per year month one will need more detailed info about such method as roi%, win%, type and number of plays, etc. and of course a free trial test. But if all looks solid and for real don't see why not offer that person at least $5,000 per month user fee. For someone to offer you $1,000 as a "one time fee" is ridculous and that was the highest bidder. That just goes to show the mentality of handicappers/horse players in general how they think. And because they been scammed and ripped off so many times in the past is another reason most now process with caution.

[traynor]

<snip>
Where I need the help is in automatically acquiring the PP and results files and preparing the data for load into the database.

I also need help in producing streaming tote data that I can use for betting decisions.

Last, I would like the capability to push those bets
automatically into one or more online betting sites at x minutes to post.

If necessary, I could probably cobble all of these tools together myself, but I would prefer to concentrate my efforts on the research, and daily wagering decision software.

If these tools are readily available, would someone please point them out to me. If they are not commercially available, would any of you be willing to share concepts/techniques (not necessarily code) with me?</snip>

Last first. A gentleman in Seattle hired an MIT computer science wiz to write a program that input a set of selections and acceptable odds and automated the wagering process. It was written in Python, and he might be willing to part with it, or copy it. I have no connection to the person, or interest in the matter. (That means if you can work a deal with him, I am happy for you both.)

Automated data entry into a database should be easy. The data format is relatively static, and comes as a stream. You create functions to extract the various elements of interest and squirrel them into your database.

Creating the data models may be a bit more rigorous. While "multiple regression analysis" seems impressive (at least in name), it is often little more than a fishing expedition in which a few large mutuels create the impression of profitability where none exists. You might look into the truncating mutuels to the interquartile mean to avoid data corruption. In essence, average the range from 26% to 75% of mutuel prices in a given category, and truncate any outliers to 1.5 times that average. It is an easy way to avoid the curse of outliers corrupting projections.
Thanks

[DJofSD]

Totally off the topic. You use the word "dissect." Hidden behind the term "textual deconstruction" is a wealth of information you might find interesting. You could call it dissecting on steroids.

In essence, it is that the form and format of communication often carries more (and frequently unintended) information than the content. Everyone is aware that "it ain't what you say, it's the way that you say it." Textual deconstruction goes way, way beyond that. I first got into it in a graduate level psychology class that required "coding' conversations; essentially breaking the language down into discrete units, categorizing them, and interpreting the context in conjunction with the content. It was an eye-opening experience.
ThanksAnd to take it further off coarse, I believe this is work pioneered by Gottman in his studies of why marriages fail.

[Dave Schwartz]

I'm thinking, "Thread drift."

[DJofSD]

Ya, perhaps. Who needs that psychological crap any ways.

But I can't say that a lot of the recent postings is answering the oringal question either.

[Dave Schwartz]

Ya, perhaps. Who needs that psychological crap any ways.

But I can't say that a lot of the recent postings is answering the oringal question either.

I was actually including my pown posts in that.

You know, the original topic of this thread was quite inspiring to me. It casued me to sit down and design a whole bunch of new things for the future. Before long I realized that I have the answer to "What's next?"

At least, I have an answer.

One must think in terms of magnitudes. That is, one does not just ask, "What will improve my performance?" Rather, one should ask, "What will raise my performance a full magnitude above where it is now?"

I really think that I have some answers. When I add up the things I haven't done in the past, things that I have talked about doing but never implemented... even some things I have written some code for... What if I implement all of them... in some fashion? Or at least what I deem the important ones?

I am just plain excited.


Thanks, Ron Tiller, for starting this thread.


Dave

[Nitro]

Well it seems this thread has gone from “What’s Missing” to “What I’ve Got”. I admit, it must sound great to those who might prefer this approach and even enticing those contemplating its use. In the world of para-mutual betting all are welcome to use whatever method they believe will satisfy either their short or long term goals. Unfortunately, only a small percentage will succeed. So, if you want to believe that within that small percentage the majority are softwhere users then nothing that anyone says will make any difference.

Now what may work or should I say at least gives a handicapper a chance is software that is very customizable, can access any of the 1,500+ data fields available, has some decent proprierity formulas built in it and can query large size databases. While these are also products that may be available to the general public cause of their "wide difference in use" and having querying capabilities I believe they the ONLY PRODUCTS any new or old handicapper should invest in. And of course also investing in the data files this software accepts. Now I'm not saying you can even win with any of these software programs. What I'm saying is I think it gives you the best "slight chance" over all of these other sold racing products I mentioned. Response....

For those of you who don't think that there are good software apps out there for horse racing, please read the above and introduce yourself to some good software that meets the above standards.

It's out there................and just because you don't use it.......or you don't get it........... doesn't mean somebody else isn't kicking your ass at the windows with it. The sad part is you don't even know it.

Dad used to say........smart guys know what they know.
Really Smart guys know.........what they don't know...........

both types populate this thread.............For those of you that believe that softwhere is the answer that’s going to put you permanently into the black then go for it!

From my point of view, in the final analysis everything that evolves from it is nothing more then a subjective response to a wealth of data from the past. You know the theory about how history is supposed to repeat itself? Unfortunately, in the horseracing game historical data is not only insufficient it’s incomplete. Without the objective information from the present, it will time and again fall short of expectations.

If it really was “kicking ass” the percentage of 1st & 2nd choice betting favorites would be at an all time high across the board. So I’m personally glad not “sad” that some think it’s the ultimate tool for playing this game.

By the way, your Dad’s well intentioned and very apropos comments should really be directed to the outsiders using machines to search for information that they think they know. Because the really smart guys know that they’ll never know as much as those on the inside and neither will their machines.
And you’re right, “both types populate this thread”.

I’ve been in this game going on 40 years, and I’ve never regretted my divorce from the soft stuff.

But before I excuse myself from this thread, I would like to express a couple of things to set the record straight. All of my comments were directed toward the use of what I believe to be a substandard approach. This was based on not only my personal experiences, but my knowledge of “What’s Missing”.
I think its unfortunate that when differences arise that some take it personally. As illustrated below:

Every once in awhile there is a supposedly successful player that likes to beat his chest and say, "MY way is the only way." When they do, they show their ignorance and lose credibility. Generally, I just chalk it up to a "need-to-lead" and forget it. (Which is what I did with you.)I’m a bit mystified by the logic here. Where in this entire thread did I discuss anything what-so-ever related to my personal success. Where did I mention anything about doing things “My way and Only my way”? No sir! The only ignorance that’s demonstrated here is by those who read into things that don’t exist. So I would imagine that it’s very easy to forget things that were never there to begin with. So who actually looses the “credibility”?

Incidentally, I’ve always taken up the same philosophy I used when I went into softwhere many years ago: “My way” is for me to use and no one else. If I’m going to exploit anyone it’s going to be in the betting pools where everyone has a place and an equal chance to put their money where their mouth is.

I appreciate the opportunity to have mentioned some apparently non-conforming commentary about a very popular topic.

Good luck to all no matter which method you decide on using.

[DanG]

So, if you want to believe that within that small percentage the majority are softwhere users then nothing that anyone says will make any difference.

BTW: Speaking of “software” your spell check is betraying you…

These threads are always interesting and often incredibly rigid. (from many angles) I’ll read several intelligent posts (Nitro as an example) and it is the polar opposite of what my experience is. All I can say is…thank goodness that we can look at similar tools and draw such different conclusions and that is not a knock on either belief system.

Speaking of Warren Henry’s wish list in an earlier post; while I admit to speed reading it / most of what I saw is available now with HTR and I would imagine many of HDW’s clients. To take it to the next level I will admit knowledge of a DB app (Access etc) can pay dividends; but there are stock tools out of the box that address most of the “what’s missing” the way I read it.

Wishing out loud; two of my all time favorite PA posters get involved in this thread; “Game Theory” and / or "Rick" from Nevada.

[ryesteve]

One thing that I don't think has come up yet in this thread: didn't the guy who's probably bet (and won) more on horses than anyone else ever, Bill Benter, use software to do so? Isn't that enough to refute the premise that "software can't work"?

[Dave Schwartz]

One thing that I don't think has come up yet in this thread: didn't the guy who's probably bet (and won) more on horses than anyone else ever, Bill Benter, use software to do so? Isn't that enough to refute the premise that "software can't work"?

Steve,

Good point.

[acorn54]

One thing that I don't think has come up yet in this thread: didn't the guy who's probably bet (and won) more on horses than anyone else ever, Bill Benter, use software to do so? Isn't that enough to refute the premise that "software can't work"?

i think that bill benter also used people who watched the races and noted the trips that the horses took in each race.
also i believe bill benter travelled half way around the world to hong kong to bet the horses. evidently the hong kong racing industry gives a better environment for the horse bettor to make a profit.

[ryesteve]

i think that bill benter also used people who watched the races and noted the trips that the horses took in each race.I'd expect he would... I also suspect that the information gleaned from these trip notes were among the inputs to his program, rather than used as a subjective add-on... and it's information available to anyone who wanted to do the work. No "inside information" required...

[DJofSD]

OK, Dave, you can zing me again <g>.

Who is Bill Benter? Yes, I've picked up on the prior messages that he's is/was a professional gambler and bet the horses. Is there any other information available about him?

[Dave Schwartz]

DJ,

Didn't mean to zing you... Did you by chance see Two and a half men last night? ("Zing" was the topic of sorts.)


The following article talks about Benter's exploits.
http://www.wired.com/wired/archive/10.03/betting_pr.html


Dave

[DJofSD]

Didn't see it but I believe my son did. I'll ask him about zing.

Thanks for the link!

[acorn54]

I'd expect he would... I also suspect that the information gleaned from these trip notes were among the inputs to his program, rather than used as a subjective add-on... and it's information available to anyone who wanted to do the work. No "inside information" required...

i guess my point is, that there is more than what software can do for you to make significant amounts of money from betting the horses, otherwise why would peter benter go half way around the world to hong kong to set up shop to bet the horses?

[ryesteve]

why would peter benter go half way around the world to hong kong to set up shop to bet the horses?Because the pools sizes were bigger, and perhaps also because he felt he had a better edge because of the quality of the competition he faced. I wasn't disagreeing with you. The only reason I brought up Benter was as a counterpoint to the "you have to be naive to think you can win with software" argument.

[acorn54]

Because the pools sizes were bigger, and perhaps also because he felt he had a better edge because of the quality of the competition he faced. I wasn't disagreeing with you. The only reason I brought up Benter was as a counterpoint to the "you have to be naive to think you can win with software" argument.

yes i read an article on peter benter's operation and it mentioned the betting pools were much larger in hong kong and also there was alot more dumb money in the betting pools from people who like to bet lucky numbers and so on.

[keith70]

Years ago (1979-2000), I was in the horse software business. For various reasons, I burned out and left the game. I am retired now and would like to pursue some theories that were marginally profitable in 2000. With a
little improvement and the current rebate availability, I believe that I can develop a series of plays that I can play pretty much automatically and make decent money. I want to do this for my own personal amusement and
profit. I have no desire to get back into the commercial software (or tout) game.

When I left, I quit following the game completely. My friends from the racing days tried to keep me abreast of some of the changes, but ended up giving up on me as "I didn't even read The Form."

After an absence of nearly ten years, I had expected to see that the software sellers would be selling most of the tools that I wanted back then. I am shocked at what I am NOT finding.

I am seeing massive "do all, end all" types of systems, both black box and tweakable. but I don't see the small tools that would allow a player easily automate some of his handicapping tasks.

Where is the tool that will automatically acquire multiple data files for a racing day?

Where is the tool that will parse/extract the data from these files and load a database?

Where is the template tool that will allow a novice to build his own database using the above tools?

Where is the conditional betting engine that will acquire streaming tote data and automatically post bets with an ADW site according to the users specifications? (The conditional betting capability at the ADW sites is not complex enough for the types of bets that I want to make.)

I am really surprised that these capabilities are only available as part of larger suites of software or to those who develop them themselves . Am I the only guy who wants to do my own analysis and my own selections, but wants to automate the process?


While I WAS a programmer, my skill set is not currently what is needed to build the tools that I seek.

In an ideal world, here is what I want to end up with ---

A) Software that will automatically acquire past performance and results data from any/all tracks running at the time. Also software that will push this data into a database.

B) A complete past performance database including lifetime PPs for all horses currently running in North America.

C) Software to analyze the database looking for profitable patterns

D) Software that will analyze the current past performances and automatically select plays based on parameters identified in prior analysis.

E) Software that will monitor the tote and structure bets close to post time and automatically submit those bets to PTC or other ADW sites.

F) Software that will monitor the results of the wagers and report profit and loss statements.

I plan to do the code for the analysis of the database, the analysis of the current days data and structuring of the plays for the day. I can also do the wager tracking and reports.

Where I need the help is in automatically acquiring the PP and results files and preparing the data for load into the database.

I also need help in producing streaming tote data that I can use for betting decisions.

Last, I would like the capability to push those bets
automatically into one or more online betting sites at x minutes to post.

If necessary, I could probably cobble all of these tools together myself, but I would prefer to concentrate my efforts on the research, and daily wagering decision software.

If these tools are readily available, would someone please point them out to me. If they are not commercially available, would any of you be willing to share concepts/techniques (not necessarily code) with me?

Warren, please send me a message if your interested in chatting sometime. I have developed software that will do most of what you say in A and B. Havn't got to the other steps yet. The platform I am using is mysql, php, and vb.net. I am still refining my database structures and open to help, ideas, suggestions. I have always leaned towards class in my projections and I truly like the system "Hidden Profit". See my post:
http://www.paceadvantage.com/forum/showthread.php?t=54229&page=2&pp=15
http://www.paceadvantage.com/forum/showthread.php?t=54733&page=2

I have use a lot of software from ADPA, Multistrats, Thorobrain to Racecom. I have learned a lot from all that software, but none of it did what I wanted to do personally.

[keith70]

Whats Next...Whats missing.

I really like the idea behind jcapper. I think more advance databases and ways to run querys on the data. Also implementing different types of technology into handicapping and wagering. Would like to know who else out there is using minitab, six sigma principles, and neural networks? This what I see is next and missing.

The key elements I find behind neural networks is feeding the learning system with very quality data, which brings us to six sigma principles.

[Maxspa]

Ron,
I believe there is a market out there for a software product that could meet the needs for the most discriminating horseplayer. IMHO, let me list a few positives from software that I have acquired and could provide ideas to formulate a new product. They are as follows in no particular order:
1. Dave Schwarz- HSH -no paceline selection process
2. Dave Schwarz- Use of artificial Intelligence with a new wrinkle that
eliminates the non productive race!
3. Michael Pizzola-Black Magic- Identification of race types- ex:(Lone Early),
(Heavy Pressure) etc.
4. Michael Pizzola- Horse patterns- Pattern Gaps etc.
5. Ken Massa - The Robot
6. Gordon Pine- Modelling techniques.
7. CJ - Speed numbers
8. Jim Cramer - His Data but also his power rating predictor.
9. AllWays - Ability to collect some data based on your preference!
10. AllWays - Smooth software with very few error messages!
11. My Way - Vermont Bob's Innovative Ideas-ex: (workouts)
12. Thoropredictor - Use of your own formulas.
13. Formulator - Video replays.
14. Synergism - Excellent Modelling!
I believe that most of all, handicappers want answers to questions that come up during the handicapping process, from layoffs to speed figures to
trainer patterns etc. In addition, these answers have to be answered by a click of a button because of time constraints and late scratches. There should be a core of handicapping information that you can bring up on a screen with a click of your mouse that is relevant to today's race. Since many handicappers use a variety of racing information in different ways perhaps a main screen that correlates with your pre-requisites could be utilized.
In constructing my list of what I believe to be important aspects in the handicapping process, does not mean you can ever have a software that contains any or all of these ideas but it shows there's a lot of room for new ideas and approaches to handicapping!
Maxspa

[dutchboy]

http://www.cigaraficionado.com/Cigar/CA_Archives/CA_Show_Article/0,2322,1413,00.htmlyes i read an article on peter benter's operation and it mentioned the betting pools were much larger in hong kong and also there was alot more dumb money in the betting pools from people who like to bet lucky numbers and so on.

[acorn54]

good article on hongkong bettors
what i got from article was that even peter benter's software fares better when the betting pools are filled with alot of dumb money. i just wonder, if even with the best software if too many people use it then there is no money to be made. seems one of the key reasons benter and others that make alot from betting the horses go to hong kong is because there is less smart money being bet in the hongkong betting pools.

[Dave Schwartz]

i just wonder, if even with the best software if too many people use it then there is no money to be made. seems one of the key reasons benter and others that make alot from betting the horses go to hong kong is because there is less smart money being bet in the hongkong betting pools.

I have consulted for several "asian teams" over the years. One of the teams actually wanted me to re-write Bill Benter's program and gave me a copy of it (which I have long ago lost).

It really wasn't all that exciting... mostly mundane stuff, like time-decay finish position and the like.

I mention this because (apparently) anyone who ever worked for Benter took a copy of his program with them. After several years, with his reting approach at work on computers all over Hong Kong, the avdantage dwindled to nothing and even negative.


BTW, Benter's software was written in Foxpro. A by-product of the Benter-program-proliferation is that the majority of teams in the asian arena still use Foxpro as the language of choice.
:lol:


If you are not a programmer, you may not know how laughable this is.


Regards,
Dave Schwartz

[DJofSD]

Dave, I've never used FoxPro. I know it's long in the tooth, however, if it still works low these many years and newer paradynes, I won't sneeze at it. (Heck, I can't, I still have a couple of VB for DOS programs I use on an occasional basis.)

[Maxspa]

All,
Check out the Selling Price on this Software venture!
Maxspa
SIM-Horse ™
Horse Handicapping Software using Simulated Annealing.
BETTER than Neural Networks!
BETTER than Multiple Regression!
An entirely NEW approach!
NEW TECHNOLOGY for Professional Handicappers. Uses $1 NRX files.
LIMITED EDITION - ONLY 100 AVAILABLE
To order send $2,495 (check/m.o.) to:
Brainchild Software Incorporated
3643 Grand Ave, San Marcos, CA 92078
Questions? Call us at (619) 581-0848
Copy protected. Req. 3GHz Pentium, 40GB disk space, Windows XP/Vista/2K, USB port.
GET $500 OFF WITH THIS AD - OFFER EXPIRES 4/25/2009 *
*Total price with ad discount is $1,995.
10-DAY NO RISK RETURN POLICY

[Maxspa]

All,
I was so amazed by the software advertisement, I forgot to mention that you can find the ad in the Daily Racing Form classified section!
Maxspa

[fast4522]

Tom,
Was there a specific part of Kgen that you liked or draws you to it on occasion other than a big race. And Energy being Phase VI does anyone have the formula for the ranking PR and LX or the manual? I was thinking LX was late pace and PR might be something like average pace. Its my general feeling these programs still work as well as they ever did, but few take the time to apply them as they were originally intended and instead go for the short cuts.

[Tom]

Check your email.

[pandy]

As soon as I saw this thread I too thought, Trakus. Any program that takes information gleamed from pps is too prone to human error (chartcalling) but Trakus isn't. I'm sure that a program could be created for Trakus where it creates a performance rating based on trip (ground loss, bias, etc.) plus a lot more.

[46zilzal]

Inter-track variants need to be updated. As good as the current Sartin software is, one always has to remember that certain tracks, (Mtn, PHa, Canterbury, Presque Island and a few others) look too good when compared to others.

Major problem I see with wishful thinking is that most people never really learn a software inside and out before they switch.

Read Malcolm Gladwell's Outliers and his chapter on 10,000 hours. The minimum time to master most things one has shown an aptitude for.

[46zilzal]

As soon as I saw this thread I too thought, Trakus. Any program that takes information gleamed from pps is too prone to human error (chartcalling) but Trakus isn't. I'm sure that a program could be created for Trakus where it creates a performance rating based on trip (ground loss, bias, etc.) plus a lot more.
Are you aware of the costs? Now that the NA economy is in decline few IF ANY are going to invest in that huge outlay of cash.