What’s in YOUR software?

[rmania]

As I see it, there are basically two types of handicapping software.

One attempts to handicap the races for you while the other serves as a tool to assist in the handicapping process.

Of the two, the most popular (and most prevalent) seems to be the type that attempts to do the work for you. Of course, these types will ever master the art of handicapping as there are just too many variables and too many scenarios to wade through.

Most (if not all) of these systems rely on statistical calculations and manipulation which, over the course of time, fails to produce the desired results. And, because the calculations and manipulations are performed internally without any explanation as to how the results were arrived at, there are no “lessons learned” for the user. One continues to be at the mercy of the system, never knowing when it is vulnerable.

Needless to say, systems such as these are ideal for those who approach this game in the same manner as playing Keno. Play as many games as possible using numbers that seem to keep coming up and hope for the best.

[rmania]

In my initial post I meant to say "these types will never...." instead of "these types will ever....". Guess it serves me right for trying to communicate a thought at 7:30 in the morning.

To be honest, I've never used a such a program (e,g,. one that handicaps for you) on a live race. I've only messed around with the test data supplied by most demos. So maybe my statements were a bit out of line.

[Derek2U]

Yeah i agree with ur post 100% .... yet i still think its possible 2 put the art & science in the right mix. hehe I dunno how it could
be done .... my father plays the races is a math prof and he pours over the DRF and only now thinks a puter could help things. He's got a lotta art down firm and showed me how he sees a race.
since i've come into PA I've met some ReallY Good Cappers who
just rely on the puter & I cant see that either. Of course, I only
play for the excitement of it & I could NEVER think/act like a pro ...
but isnt it just the greatest game in the world??

[rmania]

Originally posted by Derek2U
Yeah i agree with ur post 100% .... yet i still think its possible 2 put the art & science in the right mix. hehe I dunno how it could
be done .... my father plays the races is a math prof and he pours over the DRF and only now thinks a puter could help things. He's got a lotta art down firm and showed me how he sees a race.
since i've come into PA I've met some ReallY Good Cappers who
just rely on the puter & I cant see that either. Of course, I only
play for the excitement of it & I could NEVER think/act like a pro ...
but isnt it just the greatest game in the world??

You’re right Derek, it is the greatest game in the world!

And, for myself (and probably most handicappers), it’s the ultimate challenge.

This is the main reason why I have never used computer generated picks without knowing the logic behind them or being able to influence the results using my own knowledge of the game. It’s kind of a self-gratification thing.

Maybe I’m weird, but I get more satisfaction from solving the puzzle than I do from cashing a ticket.

I’m not surprised that your father, being a math professor, is also a horse player. It kind of supports my ultimate challenge statement. And, I am not surprised in the least that he has (so far) avoided computer programs. It’s probably a self-gratification thing.

I have my own ideas as to what handicapping software should do to make one successful at this greatest game in the world. However, it kind of bucks the trend. But then I have a history of bucking the trend.

You read my original post so you know how I feel about “pick” software. As a handicapper AND a computer programmer I can tell you that the weakest link and the “Achilles heel” of “pick software” are they scenarios. You can compile all of the statistics you want and manipulate them however you want, but if the software is unable to blend all of these ingredients into a single scenario (or race) then, in my opinion, you’re back to playing Keno.

Computers aren’t too keen on blending. It’s basically a black and white environment. Yet, the human mind is the ultimate blender.

[betovernetcapper]

I use 3 software programs-At the Races-Netcapper and Equisim. None of these programs are mindless-play the top number-check your brain at the door of the type you are refering to. People spend hours-days struggling with arcaine matters such as setups-matchups readouts-interpatation of various ratings. When I view the forums at HS and HTR I see people ingaged in the same pursuits.
I have tried your software and while it's beautifuly written and a pleasure to run, it pales in usefulness to Equisim ,the ORIGINAL SIM program.
just my 2$

[rmania]

Originally posted by betovernetcapper
I use 3 software programs-At the Races-Netcapper and Equisim. None of these programs are mindless-play the top number-check your brain at the door of the type you are refering to. People spend hours-days struggling with arcaine matters such as setups-matchups readouts-interpatation of various ratings. When I view the forums at HS and HTR I see people ingaged in the same pursuits.
I have tried your software and while it's beautifuly written and a pleasure to run, it pales in usefulness to Equisim ,the ORIGINAL SIM program.
just my 2$

Has to be a sign of the times where as what used to be 2 cents has become 2 bucks :)

If the programs you mentioned as using don’t fall into the “mindless picks” category, then I am in favor of them. And if they work, then that’s even better.

So, you use 3 programs. Do you integrate the results or do you use them for different scenarios (i.e., maidens vs claimers vs etc.,)?

I first downloaded the Equisim demo years ago. I was more interested in the sims than how they were arrived at. I wasn’t all that impressed (too cartoon-ish), so I deleted it. I recently downloaded it again to see if any improvements had been made and, to my surprise, nothing had changed. I guess it’s evolved into a cult thing.

Another problem with the sim (and this is a biggie) is that it doesn’t really tell me anything. In other words, I can’t look at it and get a feel for how the race might be run. And, in my opinion, that’a what a simulation is supposed to do. I don’t know, maybe I’m lacking the necessary software to make these sims a little more like-like.

[betovernetcapper]

In answer to your question ATR and Netcapper will be integrated (hopefully) this week.
Perhaps this is a personal thing but I can look at the Equisim sims and get a feel for how a race may develop but maybe I just have a superior imagination. :)

[Handle]

RMania -

Ok, let me get this straight - horses run around a virtual track that is represented to scale in true 3D. By and large, they attempt to get as close as possible to the rail while running around the turns -- and you see the horses running _around_ the turns. Horses on the outside are effectively at a disadvantage because they must run further on the turns. The simulation, which, again, is to scale, takes this into account. Now, as these horses run they accelerate and de-accelerate. Further, you can put the camera on any horse you want - you can even "ride" the horses with jockey cam. You can zoom in and out, and pan around the field. The MPH of the leader, distance traveled, and time is displayed on screen along with the current top 4 (the way a simulcast feed typically looks).... And you want something that is more "life-like" in order to get something out of it?

Here's something to try: watch the simulation -- are there a bunch of horses all on the lead? Gee, wonder what that might mean. How about one lone horse running ahead of the pack? How about a 46/1 shot named Magic Weisner screaming up from behind in the Preakness.. he might be useful in an exacta, or the ticket to scoring a 16K superfecta.

You must have one hell of a simulation. How about sending me a copy:

thorotech@thorotech.com

Nothing has changed in a couple of years? Right, my cult following isn't as interested in the software wasteing CPU cycles rendering the snot flying out of the horses noses -- so little has been done to improve the 3D rendering of the simulated race. They'd much rather I spend my time developing new secret handshakes and finding young virgins for sacrifice at our monthly rituals. Oh, that and developing features that can help them handicap horse races; things like the Profiler feature and the ability that EquiSim has to track how well its simulated races perform against actual race results built into the software. The simulations have evolved over the years so that they can make effective use of any pace line to simulate a race, irregardless of the distance of the pace line and the race. In the next version it will be possible to simulate performance based on multiple pace lines. And... ES simulates any race from any track out of the box.

Nathan - Cult Daddy, EquiSim

[Speed Figure]

WOW! It's your turn RMania:D

[rmania]

Hey Nathan...

I think I covered my tracks by suggesting that I may not have the proper software on my machine to do the sims justice.

I think that the 3D rendering is quite cool.

I'll be glad to show you mine as you have allowed me to see yours. :)

But you have to promise that you'll give me your honest opinion of it.

[Speed Figure]

RMania, The sim on your program looks like the sim on Quarter Pole Plus. Have you ever played Quarter Pole Plus?

[Handle]

Ok, I've seen the RaceMaker demo. Thanks RMania for sending it along. Now, my honest opinion of the "simulation of a race" portion of the program....

The simulation looks alright, but uses old sliding panel (80's arcade style game) technology for displaying the race. I didn't see any pause buttons so I could stop the movie and have a good luck at what was going on - nor did I see a way to fire off all the simulations so I could then review the data that they generated (charts). These would be nice.

These issues limit the usefulness of the program to handicapping:

A) It only does sprints. This makes me wonder why. Is it because of the lack of turns (see below) - or is there another reason related to the use of the data - how is the data used to generate the races? If its a simulation, then it should be more or less straightforward, so why not routes too?

B) The user lacks any sort of control over the simulation save for inputing your own fractional times for each horse (and I'm just guessing that this would alter the simulations, I didn't try it). I have one cult member who made a whole lot of cash last year. His primary modus operandi was to simulate races multiple times, each time using a different set of pace lines for selected horses to get "what if" scenarios. Because I can't see the effects of different pace lines on a simulation, I have less of an understanding of what the simulations truly represent. Its the "what ifs" that are very important in horse racing....

C) It doesn't display the horses running around the turns. If I'm going to _watch_ a simulation of a race in order to get something out of it, then I want to see what's happening throughout the race. Then I can make a judgement - that horse is on the lead on the far turn -- can he hold on, will he tire? How far ahead is he of the others? Is there anyone that looks close to him today? RaceMaker's representation, with the fade-out, fade-in rather than a display of the turn makes it much harder to get this view of the race.

D) I don't know if I'd consider it a simulation. A simulation, by definition, is an attempt to re-create actual conditions in a virtual space. Perhaps the program is taking all of these factors into account, and then displaying the results as a run down the backstretch and a charge up the stretch - I don't know. But, given the way that RaceMaker displays the races, my guess is that it does not take into account such things as position around the turns, or whether it is to scale in any meaningful way.

Rather than a simulation, it looks like an "animated" graph. That is, it takes some measure of each horses speed and graphs them. Each sprite (horse) on the screen represents a point on the graph. Add time to the graph, animate the points over time. Except, I'm not sure if this "graph" has a uniform scale. Is 1 length equal to 1 length throughout the race?

To me, a simulation is a way to understand how a set of factors interact with each other. Those factors may be each horse's likely velocity throughout the race, the distance each horse must run, and how each horse's running style might effect the way the race is run (the pace). RaceMaker certainly fits this to a degree -- you can see where each horse is in relation to the other based on their speed.

Or does it? To what degree is the position of each horse based on its past performance data, and based on the running time of the race? Is the lead horse on the lead by X Lengths because his past performance data put him X Lengths on the lead, or is he simply some amount of "screen space" on the lead because of this data. This is not a rhetorical question - I don't know. But the way the horses often move up and move back makes me think that the "simulation" is loosely based on the data and the sudden movements are more or less random.


-Nathan

[Handle]

Addendum - I did find a way to subsitute pace lines for each horse. This option, which I was surprised to not find at first, alleviates some of the misgivings I mentioned above under lack of control.

[rmania]

Now Nathan...

You were suppose to be looking at mine – not yours.

JUST KIDDING :) :D ;) :p

If you look at your copy, there’s help on the Sim splash screen that instructs you to press “P” to pause and “R” to resume. It also details the “Print Results” feature (i.e., charts).

The decision to use sprints only was mine. The sims are merely a by-product of the handicapping and the handicapping methods work exceptionally on sprints. Too many variables with routes. But then, this was reveal from the start.

The animation is driven by the data. And, though paceline substitution is a feature, a better feature, in my opinion, is one that allows the user to modify a paceline and use that data. Makes for even better “what if's”.

And you’re right, it doesn’t show the horses running around turn. This would require a 3D effect (such as yours) and I don’t know that it’s really necessary for my application. The layout is to scale (try timing a race by qtr.. it’s pretty close) and the fade out represents the mid-point of the race.

The rest of what you saw may have been a result of the machine you’re running on. It’s pretty smooth on mine.

Thanks for the opportunity to share this with you.

[JustRalph]

Originally posted by Handle
Nothing has changed in a couple of years? Right, my cult following isn't as interested in the software wasteing CPU cycles rendering the snot flying out of the horses noses -- so little has been done to improve the 3D rendering of the simulated race. They'd much rather I spend my time developing new secret handshakes and finding young virgins for sacrifice at our monthly rituals. Oh, that and developing features that can help them handicap horse races; things like the Profiler feature and the ability that EquiSim has to track how well its simulated races perform against actual race results built into the software.

rmania

Equsim users rarely even run the sim in real time. They run it automatically and just refer to the chart that it produces. The program has evolved several times in the last couple of years. It's just like any other good software, you get out what you put in. ES is a damn fine product. I have been in the computer support business for years. I have yet to find a product that was more reliable on multiple platforms etc. There are some that work on a lower level so as not to have the inherent problems that can crop up. They are less intuitive etc. Handle is a solid programmer who is very reponsive to users. Skip the computer crap and it's a damn fine platform for handicapping. If you find one that is better, buy it! I can attest to it. I have only been playing the horses a couple of years. But I made a ton of money last year. Now a ton is a relative term. But for me and several of my friends it was a really neat year. Some of them called or emailed me before the triple crown races and we had a good time. That 46-1 horse in the Preakness was straight out of a cold Super that ES picked. It matched my handicapping except I had a question on the 46-1 shot. ES answered that. Big day for me. The program is very versatile and the latest version does much more than it did two years ago. The problem you have is when you disagree with the program. It takes a bunch for me to disagree with the program. But that makes you a better capper. The way you use the program is the key. Back to what you put in. I suggest you try it again. You totally missed the boat the first time. :cool:

[Handle]

Originally posted by rmania
The animation is driven by the data. And, though paceline substitution is a feature, a better feature, in my opinion, is one that allows the user to modify a paceline and use that data. Makes for even better “what if's”.


An earlier version of EquiSim (before development ceased, of course and it became a cult classic ;) ) allowed for direct manipulation of the the pace lines that controlled simulation. You could also add "relative umph" to a horses performance (add kick, fast early). I took this feature set out because nobody used it. Manually manipulating the figures to watch the simulation is too high a price to pay to get that "what if".

ES does have the ability, however, to do simulation tuning. This is where it runs the same set of races over and over again, each time changing the parameters that goven simulation as a whole, to find the best values for these parameters. The concept is a sound one, but the results have been mixed. Some have good success with tuning, especially when finding significant "biases" in how ES sees the effects of early and late speed (correlations exist to how a specific track plays), others chiefly ignore tuning altogether.

-N

[JustRalph]

Originally posted by JustRalph
rmania
I suggest you try it again. You totally missed the boat the first time. :cool:

Maybe a more accurate statement was that you should have stayed on the boat for a longer ride. Then you could have witnessed all the changes. The future looks very bright in the upcoming versions. :cool:

[rmania]

Originally posted by JustRalph
Maybe a more accurate statement was that you should have stayed on the boat for a longer ride. Then you could have witnessed all the changes. The future looks very bright in the upcoming versions. :cool:

Sorry Ralph, missed your previous post first time around.

My comments about the sims pertained to those found in the demo and, more specifically, the visual representation.

What I downloaded last week looked the same (at least to me) as what I downloaded 2 years.

My "cult" comment was not meant to be offense. I was just offering a reason as to why the visual representation had not been upgraded.

Once more, my comments were based solely on the demo. If the official version is different then maybe you can explain how.

[Lefty]

When I was a salesman, many eons ago, one of the first tenets of salesmanship was never never knock another product or even the appearance of knocking a similar product. Just concentrate on the good points of your own product and let the buyer form his own conclusions.

[rmania]

Originally posted by Lefty
When I was a salesman, many eons ago, one of the first tenets of salesmanship was never never knock another product or even the appearance of knocking a similar product. Just concentrate on the good points of your own product and let the buyer form his own conclusions.

You are so correct in what you say Lefty.

I apologize as I should have never ventured down that road.

Sorry Nathan......

In an earlier reply I addressed one of Nathan's questions, but I wasn't really clear in the response.

The question was:
"A) It only does sprints. This makes me wonder why. Is it because of the lack of turns (see below) - or is there another reason related to the use of the data - how is the data used to generate the races? If its a simulation, then it should be more or less straightforward, so why not routes too?"

In the reply I stated something to the fact that there were too many variables for the method to be effective with routes. And this is so true.

In case you're not familiar with the method, the program attempts to convert paceline infomation into a common distance for each horse in a race. I like to think of it as changing apples into oranges. And, in my opinion, converting sprints into routes would just be too unreliable (again, because of the variables).

The converted information is then displayed visually (by qtr), giving the user an almost instant idea as to how the race shapes up.

At this point, the user can handicap the race using a variety of methods including paceline substitution.

However, what I have found to be more effective is using a horse's most recent paceline and modifying the final quarter only to reflect how I believe the horse will improve or regress based on the overall effort in its last race.

I do this because I have found that (in most cases) a horse is likely to display, in the upcoming race, its current half-mile form with the only difference being that (all important) final quarter.

[JimG]

Originally posted by Lefty
When I was a salesman, many eons ago, one of the first tenets of salesmanship was never never knock another product or even the appearance of knocking a similar product. Just concentrate on the good points of your own product and let the buyer form his own conclusions.

That's what I call hitting the nail on the head. Great post!

Jim

[Derek2U]

any of U puter guys have a piece of software to keep a Running
Tally of the Contest $$$ ? .... raise ur hands

[BillW]

Originally posted by Derek2U
any of U puter guys have a piece of software to keep a Running
Tally of the Contest $$$ ? .... raise ur hands

Derek,

How about a spreadsheet? Probably won't find any software specifically for contest scoring. Would have to home brew it.

Bill

[Handle]

Its silly that there's nothing to score handicapping contests available. Obviously some places have the software (BRIS for example). It would make managing the contests a lot less work.

With a spreadsheet you still have to enter players' selections and results by hand (ugh).

I've thought about creating such a thing, and it could have some other cool features too, and may well do it when I get a chance (?). But, alas, nothing yet.

-N

[BillW]

Nathan,

Yea, Spreadsheet like using a baseball bat to drive in a nail ... unfortunately, it beats a word processor :).

I'm working on some bet tracking SW now, which this would have vague similarities to, but not even close enough to use it as a core. Definitely a unique application.


The ultimate would have to have a feature to allow customizing the rules and would understand exotic win calculations. Just enter team names and D/L results from one of multiple sources (Bris instant charts/Bris exotic results/Equibase charts etc.). My teeth itch thinking about it. Not impossible, but I think the effort : utility ratio would be high.

Bill

[GameTheory]

Writing something to keep score would be pretty easy, but you'd still have to enter the selections by hand (but not the results). For an online contest, you could create a form that the participants enter their picks into (on a web page), which could then be automatically entered into a database or output file in a consistent format that your scorekeeper program could easily parse.

To make the bones of such a system is only a day's work I think.

To make a generalized, flexible system for any contest with different rules, etc. would be more complicated...

[BillW]

GT,

Agree ... I think the trouble of handling the custom rules would be more effort than it would be worth (good example ... look at the unique rules for this contest). Building something for each contest would be the way to go.

Bill

[Handle]

GT,

I think you're underestimating the time requirement here, although a lot would depend on what you were shooting for.
A day for something that simply accepts selections (program numbers) and then passes those numbers off to the more complex "back end", perhaps would be a day job.

But I think there's quite a bit more work if you want something that reads in entries (better yet, goes and gets them automatically from other web sites) and presents these to the user for him/her to make selections; understands exotic wageriing format; reads in results (could go get them); compares results; maintains a database of results; does all of this one the web with little or no administration.

Of course, I'm talking from scratch here. Writing the parser to parse the entires, the results file, the user interface, etc..

[betovernetcapper]

Nathan if you don't have time to show the horse's snot on the sims, I don't think you should engage in speculations on new software.
This is a tough contest-I've got a double and a pic3 and am not likely to finish in the money today.

[GameTheory]

Originally posted by Handle
GT,

I think you're underestimating the time requirement here, although a lot would depend on what you were shooting for.
A day for something that simply accepts selections (program numbers) and then passes those numbers off to the more complex "back end", perhaps would be a day job.

But I think there's quite a bit more work if you want something that reads in entries (better yet, goes and gets them automatically from other web sites) and presents these to the user for him/her to make selections; understands exotic wageriing format; reads in results (could go get them); compares results; maintains a database of results; does all of this one the web with little or no administration.

Of course, I'm talking from scratch here. Writing the parser to parse the entires, the results file, the user interface, etc..

Ok, correction -- it would take *me* a day because I've already got a downloader, a parser, etc. I would just need a web form for selections (it doesn't have to do anything fancier than email the adminstrator in a consistent format or append to a text file on the server) and some "glue" code to hold it all together. I would also assume that the adminstrator would have to run the program on his home machine once a day to update things and create a report or something that he would then upload to a contest results page somewhere. There would probably also be a program to take the entries and turn them into a multiple-choice web form. Like I said, it wouldn't be flexible - you'd have to hand code in the scoring rules for the contest, but workable. All the basic components I've got practically ready made...

[Derek2U]

I dont know what you mean by those technical things, but
could such a program HELP reduce the workload in our contest?
I'm sure it would help PA even if it helped 25% or so. But that
stuff is too heady for me.