Software Contest [Archive] - Horse Racing Forum - PaceAdvantage.Com

rmania

07-12-2003, 01:10 PM

I’m game......

Of course the rules would have to be tailored somewhat for me to participate.

Since RaceMaker currently focuses on SoCal sprints only, I’d be on the sidelines based on the initial proposition.

And, like a few other programs I’ve seen, mine doesn’t spit out winners. My “VirtualView” handicapping feature breaks a race down into 3 segments (half-mile pole, quarter pole, and finish). As mentioned in the tutorial, positions at the quarter pole are the strongest indicators. And, true to form, horses shown leading at the quarter pole have been winning at a rate of over 50% for the current Hollywood meet. Most have been sent off as favorites or second choice, but a few nice hits have kept the ROI above +25%.

Having said that, I’d have to pass on any contest that included exactas and trifectas since that would be testing my skills as a handicapper and not the validity of the software.

So like I said... I’m game if the results provide a fair evaluation of my software.

Speed Figure

07-12-2003, 01:17 PM

Rmania, You are one of the only software makers that post there programs picks.;)

rmania

07-12-2003, 01:23 PM

Originally posted by speed figure
Rmania, You are one of the only software makers that post there programs picks.;) Thanks fffast, but if you read my posts the picks are generally not "program picks". They are my picks and they're usually double-digit odds. I guess that's just the handicapper in me. Always trying to beat the favorite.:rolleyes:

Dave Schwartz

07-12-2003, 01:34 PM

Let me ask you a question... How should we represent HSH?

Which approach should we use?

PowerLine?
7-Button?
Synthetic Pace?
Prime3 Object?
Prime2 Objects?
Rating Objects?
NetPace w/suggested pacelines?
NetPace with some other custom paceline selection system?

How should we adjust the pacelines? For age? Penalize maidens stepping up?

Should we adjustour probabilities towards the public's (i.e. adjust for the toteboard)?

Should we use the Synthetic Pace (no paceline selection) combined with some other factor (such as class or form)?

My point is, that to have a contest that "takes the output of a program" is dependant upon the settings chosen by the user. And that is as it should be.

One of the biggest fallacies in racing is that a user is going to find ANY software that can do it all for them at the touch of a button right out of the box.

Now, that does not mean that a program cannot be MADE to "do it all for you." It just means that YOU've got to find the way to do it.

I'll tell you what I'd be up for...

1. Each software developer uses his own software...

2. Posts an EXACT system for doing the same thing in each race (i.e. a set of rules that his own users can follow along with).

3. We play EVERY RACE from 5f to 10f at each of (say) 8 tracks. Dirt, turf, maidens, FTS, everything. That should give us about 70 races per day.

4. We bet 2 horses in every race.

5. The picks MUST follow the rules set by the developer. That is, there is NO personal decision possible.

6. We send the picks via email to 2 contest monitors prior to (say) 2 minutes to post. (One contest monitor is simply a backup in case email fails.)

7. We also send along screen shots verifying that we made the correct picks from our system.

8. Miss a post deadline, lose the race.

9. If you don't follow the rules you set, you lose the race.

10. Results would be tallied every few minutes and posted on the PA board. (We'd need a volunteer to be keeping score in a spreadsheet and the spreadsheet is simply posted.)

11. The contest runs for 5 weeks... Starts on a Wednesday, 8 days later we do a Thursday, 8 days after that we do a Friday, 8 days after that we do a Saturday and the last is a Sunday. We should wind up with about 350 races, or 700 bets. (And possibly some scheduling adjustments would have to be made to accomodate particular vendor's personal needs.)

We could call it the Great Software Challenge.

I'd be honored to be the first entry. Any other takers?

Regards,
Dave Schwartz

Speed Figure

07-12-2003, 01:41 PM

Dave, That sounds great! Looks like HSH is ready to go.:cool:

Dave Schwartz

07-12-2003, 01:55 PM

HSH IS ready to go.

Tom

07-12-2003, 02:09 PM

Originally posted by Dave Schwartz
HSH IS ready to go.

Give me time to fund all my accounts! :eek:

rmania

07-12-2003, 02:47 PM

Originally posted by Dave Schwartz

Which approach should we (HSH) use?

PowerLine?
7-Button?
Synthetic Pace?
Prime3 Object?
Prime2 Objects?
Rating Objects?
NetPace w/suggested pacelines?
NetPace with some other custom paceline selection system?
As one can see, many of these programs contain multiple "sliced and diced" views of the same data.

The developer should have a feel as to which view produces the best overall "out-of-the-box" results. The developer should also know which method is the easiest to comprehend by the user.

In other words, what's the best and fastest way for a novice to use the program and make a profit.

The process should be simple and identified up front with no deviations.

If we limit the number of tracks to one per day, playing all applicable races, with win only on the top pick and a 3 horse exacta box on the top 3, then I'm in. Otherwise I'll be a most interested bystander.

Secretariat

07-12-2003, 04:06 PM

Curious how programs whicj rely heavily on modeling and datamining would be affected by your rules Dave.

My programs place a great deal of emphasis on data mining in making decisions. Or are you proposing no modeling, only using ratings.

Dave Schwartz

07-12-2003, 05:25 PM

Secretariat,

I plan to build a systematic approach using modeling on each individual race. Should take me about 90 seconds for each race.

In fact, I plan to use the same system I just demonstrated on our latest video, with one small modification (which I shall divulge, of course).

Dave

Dave Schwartz

07-12-2003, 05:39 PM

RMania,

To quote my previous message:

"One of the biggest fallacies in racing is that a user is going to find ANY software that can do it all for them at the touch of a button right out of the box."

You said: "In other words, what's the best and fastest way for a novice to use the program and make a profit."

In my opinion, novices are not supposed to be able to make a profit. It's just not that easy. There is a necessary level of commitment to an approach, a program, a discipline, if you will.

The important point is that it should be an approach that someone that does not have the skill of a Tom Brohamer or a Dick Schmidt should be able to accomplish without a lifetime commitment.

As for playing only one circuit in the contest... I would suggest that shows your product to be a "specialist" and personally, that would not be something I would be interested in being. Also, I think that it is not within the scope of this contest.

Personally, I do not do well with first-time starters... but I am not going to say "Let's not play any first-time starter races." Let me have to build a set of rules for playing those races.

Dave

Dave Schwartz

07-12-2003, 05:41 PM

PA? Would you care to weigh in here with an opion?

Is this something you'd care to have happen on your BBS?

Dave Schwartz

PaceAdvantage

07-12-2003, 07:11 PM

I would be very interested in seeing how various software programs stack up in a head-to-head contest.

Would be some great free publicity for those that do well.

It would be easy to score, since there would be only a handful of participants (can't be more than 20 programs out there right now, could there??? ;) )

I'd be up to hosting it if that's what happens...

Speed Figure

07-12-2003, 07:19 PM

Does it have to be the program maker or can it be a program user? I can only see 2 or 3 program makers doing this.

Dave Schwartz

07-12-2003, 08:34 PM

I suggest that it should be the "program maker's" place to do this. That way we all take responsibility for our software and cannot say "Well, he isn't using it correctly."

Regards,
Dave Schwartz

rmania

07-13-2003, 09:57 AM

Originally posted by Dave Schwartz
RMania,

To quote my previous message:

"One of the biggest fallacies in racing is that a user is going to find ANY software that can do it all for them at the touch of a button right out of the box."

I agree, no software can do it ALL...

You said: "In other words, what's the best and fastest way for a novice to use the program and make a profit."

In my opinion, novices are not supposed to be able to make a profit. It's just not that easy. There is a necessary level of commitment to an approach, a program, a discipline, if you will.

I disagree... If the developer has (over the years) found a way to continually make a modest return on investment and has properly packaged this knowledge into a program; and the user (novice or otherwise) displays a certain amount of discipline then, IMO, profitability is possible. And the profiability should increase proportionately based on the users knowledge of the game.

The important point is that it should be an approach that someone that does not have the skill of a Tom Brohamer or a Dick Schmidt should be able to accomplish without a lifetime commitment.

Again, I agree....

As for playing only one circuit in the contest... I would suggest that shows your product to be a "specialist" and personally, that would not be something I would be interested in being.

In most lines of work specialists make more $$$ than the "Jacks of All Trades".

Also, I think that it is not within the scope of this contest.

I didn't know that the "scope" of this contest had be finalized.
I guess I'll just watch.

VetScratch

07-13-2003, 09:59 AM

Dave,

My guess is that you will have a short field in this contest unless you find a way to let vendors set the rules by some sort of consensus mechanism. In the end, no vendor might get exactly what it wants, but more will get what they need to be competitive.

For example, your proposed rules may make it difficult for value-play systems to participate. In some fields, no horses will qualify as a system play, and your selection posting deadline will be cutting it pretty close for many value-play vendors.

What are the wagering rules and who wins? Maybe you need two contests: a bankrolled money contest for one group of vendors and a win percentage contest for a second group.

JoeG

07-13-2003, 10:38 AM

This post began as a wish to post the top 3 picks for a month for each software vendor. Dave chimed in with a first proposal (top 2 picks every race for 8 tracks). Then came posts about value, another about WP, exacta, trifecta.......then SoCal sprints only.

The problem is that there is no single way to bet. I tend to agree with Dave's initial proposal of Win bets 2 horses per race. I think (like the online contest) that when you include exactas, tris and pick3s you head more toward a "big score wins" contest. Win bets are more traditional in major contests.

But, is 2 bets per race fair? Just bacause I like it, doesn't mean that 1 bet per race is worse. Or 3 per race? and WP or WPS?

How about the "value" player?

I don't know the answer, but I like Dave's proposal of the software vendor sending in the picks and with screen shots.

Also using exotics can skew results. Last week for example I would have had a $1200 Harness exacta cold with some ratings I have (didn't bet it). That would have been a Great score for a contest, but these biggies don't come along often. And going down to the 4th rated horse would have gotten me a 5 figure tri. This one score would easily givin me a top finish in some contest, If all my other bets combined were not profitable I still could have possibly won. Would this prove anything?

rmania

07-13-2003, 11:24 AM

Albeit, the majority of products discussed on this board claim NOT to be “black boxes” (including mine), each program, more than likely, contains at least one profitable “angle” that requires little or no input/knowledge on the part of the user.

For the contest to be fair the developer should not be allowed to manipulate data in any way.

As far as posting 8 races at 10 tracks (or 80 races per day/challenge) I can’t see where that will prove anything unless the thought is that the more races the better chance of getting a desired percentage of wins.

Developers should have a feel as to which tracks are best suited for their program so I say pick your track of choice.

Identify one horse per race and score the contest based on win percentage – not bankroll.

If anyone has a problem with my program being “sprint” oriented then I'll take automatic loses on the routes.

betovernetcapper

07-13-2003, 11:29 AM

Why not score the contest in the same way a customer might evaluate a program. Most programs will produce a line on some sort or best 2nd best 3rd ect. Scores could be given top pick win%-roi-----2nd pick win% roi ect. If judges have too much time on their hands there could even be sub catagories like sprint-route-turf. :)

rmania

07-13-2003, 11:38 AM

Originally posted by betovernetcapper
Why not score the contest in the same way a customer might evaluate a program. Most programs will produce a line on some sort or best 2nd best 3rd ect. Scores could be given top pick win%-roi-----2nd pick win% roi ect. If judges have too much time on their hands there could even be sub catagories like sprint-route-turf. :) The only drawback I see is the ROI especially if one gets to choose their on track(s). Hopefully you can see my point on this.

Let's face it, if the win percentage is up there then the ROI will follow.

VetScratch

07-13-2003, 12:53 PM

Bystanders, who may become users of one or more software products, should logically prefer ROI results for this contest because:
(1) You are proposing a national cross-section contest based on many tracks and races (350 proposed, while a 1000 would be better yet). As opposed to all the ridiculous contests, like those at BRIS, which are based on a handful of races, ROI can finally be used to measure performance. Nothing else matters in the long run.
(2) Given the cost of software, potential users are looking to make commitments can't be undone and won't be changed with great frequency. Unless all the vendors agree to rebate license fees in proportion to periods of use, most users will stick with their selections for at least six months. This underscores the case for ROI as the most meaningful measure of success.

Dave Schwartz

07-13-2003, 01:50 PM

I weighed in on this thread because PenguinFan chimed in with a rather provocative post about software devleopers not willing to allow his product to be measured. So, if anything is accomplished, it must include addressing THAT particular post.

Now, how does one best do that?

In my opinion, the test must significant enough to provide a decisive answer to the question "How good is THIS software?"

Of course, that definition is very much open to discussion.

But, let's understand at the outset that this is not a "fun contest." The livlihood of each developer depends upon the reputation of his software (and himself). Any vendor that enters this contest is taking a great risk (and knows that, of course). But there should be a strong payoff.

Let's discuss the issues a bit...

"1. Playing top 2 (or 1 or 3 or 4, etc.) - This is a clear-cut benchmark. Many players like to play 2 horses in each race. I want people to know what they can reasonably expect of the top 2 horses for each program when this is finished.

Understand that most players will not actually play two horses in EVERY RACE on EVERY CARD. (Oh, they'd like to but the issue of value comes in to the picture and some races are simply not playable.)

2. Using value as the prime consideration - I have no problem with that on the surface, but value by its nature demands passing races. This needs to be a definitive test with no passes o a limited number of passes. And, if we are going to "demand" that the program plays (say) 8 races out of every 10 what set of rules can the developer create to handle a pass? You can't say "only play horses with a 10% profit ranked in the top 2" because that might only pick up 5 races on this card.

I am trying to create a level playing field. That is why I suggested that playing the top 2 was a reasonable number.

3. Playing Every Race This is a nice feature. It ultimately lets the program demonstrate that its top 2 selections win some percentage and return some $net. It is the combination of the two that is important.

If $net (ROI) is all that matters then we will turn this into a non-definitive contest like the PA handicapping tournament. (I am not knocking that tournament - it was FUN, but this is not about fun, remember?)

If it is based upon ROI then we will all be forced to play artificially like any tournament, chasing the long shots. Tell me... how many users would even want a product that won the contest with the highest ROI but only hit 5% winners?

If you think about it, the majority of us who develop software are developers first and players as an afterthought. (That doesn't mean we don't play or win.) That is because developing software, keeping it updated, dealing with customers, marketing (whatever that is) tech support, etc. are all full time jobs in themselves.

Look at the primary products that frequent this BBS... HSH, NetCapper, HTR, Equisim to name a few. I have first hand knowledge of all my competitors... they are all honorable guys, putting out good products, standing behind promises.

Take Ken Masa, for example. I have heard from many people that he has a lot of contact with his users, like I do with mine. That takes a LOT of time to do it right. You can't hustle people off the telephone just because there isn't any money to be made from the call. We all make a commitment to our software but mostly we make a commitment to our users.

Now, each of these products have a following. And among the following are some very successful users, some less successful ones, a few serious "non-winners" and a handful (I'd bet) of disgruntled customers that blame the software for their inability to win.

In my opinion, the secret to winning at the races is to match up the right approach and/or right software with the right user. This contest is simply about eliminating the pretenders from the contenders.

I have first-hand knowledge of 6 people who consistently make $100k+ at the races. And they all have one thing in common (beyond using my software): If they chose to use someone else's software they'd probably still come out winners.

I will, of course, be open to further discussions about the rules of the contest but I really believe the points that are critical are:

1. Fixed number of bets per race
2. Play every race
3. Play by a set of published rules to remove the artfulness of the player.

So, change the number of bets per race, allow a number of passes based upon a set of rules (i.e. Personally, I don't play races where FTS money makes up more than 28% of the pool), make the playing rules include oddsline instructions if you want, but do not allow simply playing the "perceived overlays."

We are not trying to test a spot play system here. We are trying to test our software for its general abilities.

And, RMania, as for building a contest around playing sprints in SoCal, we'd have to play 15 times longer to have a contest and then people would say, "But how does it do at the other tracks?"

Regards,
Dave Schwartz

VetScratch

07-13-2003, 02:28 PM

Dave,

Let's discuss the issues a bit...

"1. Playing top 2 (or 1 or 3 or 4, etc.) - This is a clear-cut benchmark. Many players like to play 2 horses in each race. I want people to know what they can reasonably expect of the top 2 horses for each program when this is finished.
Two horses per race sounds fine for the common sense reasons that you stated, but why the TOP 2? Some products offer alternative methods and would advise users to bet TWO TOP PICKS instead of the TOP 2 from one method. If the rules are TWO BETS, why not let the vendor set rules to designate which TWO?
3. Playing Every Race This is a nice feature.
If it is done this way with a constant wagering unit, then ROI or $2Net is still the best measurement. Which is better, 30% wins with $2Net=2.10 or 35% wins with $2Net=2.00? I would choose the higher $2Net.

Tom

07-13-2003, 03:05 PM

As a software users, my primary concern is what %winners do I get out of the top 2. Others may focus on roi, longshots, but personnally, I want to see what the program does at a basic level:pick winners. If I am sure about my program, I can apply whatever other criteria I want ot ensure I make money, but I want to know that the things picks the winner often enough for me be wrong frequently.
At HTR, we have contest going on pitting our own paceline picking skills against the prgrams' primary paceline picking mode-PL5. The criteria are pick your own lines and post the top 2 horse rated in velocity. One race a day. Same race for everyone. After 86 days, PL5 has won 41% of the time with a 16% roi. In fact, PL5 has outperformed everyone except one....ME! (Allright, I only played one day and won, so I am in at 100% wins and 770% winners with a $34 horse,but it sure looks good in the result thread...hehehe).
So far, NO NONE has made more money than the program in auto mode. Only one guy has a better win%...our own GR1!
But the point is, this test shows me how good the program is under a specific criteria. But there are many, many other things in HTR that are ignored here, and I suspect other programs also have features that will not come into play. I think everyone has to have an open mind about the results, and I think the developers should have the opportnity to voice concerns/opinions about there picks (before the race is run, of course).
Suppose we end up with the folowing results:

HSH 32% wins 12% roi
SHS 43% wins 6% roi
SSS 16% wins 46% roi
HHH 25% wins 20% roi

Which one wins? Which one do YOU want to use. I bet the choices are not unanimous, but here is a good way to measure what WE want in a program, eh???

MarylandPaul@HSH

07-13-2003, 03:09 PM

I would suggest including Place, and possibly even Show results from each vendors picks in the scoring. Any software that falls a bit short on the win end, but does well finding ITM runners, is certainly useful to an exotic player.

MP

JustRalph

07-13-2003, 03:26 PM

I don't think you will prove anything with this contest. I know that the software I use is tailored to allow me to use it in several different ways. I know that other users of the same software use it in very different ways than I do. It appears from earlier posts that a couple of others in here are written in the same manner. It depends on the user to effectively apply the software based on the features. I know that I use my favorite program much differently than the person who writes the software. If we allow the programmers to post the picks etc it would have no real reflection on how any one end user may use it. It's a pig in a poke guys. Some will probably turn out looking worse than what their software is capable of, and others might look fantastic based on the method of use. What really counts is how the end users would fair with the product. If you want to see how they use the program, go back to the results of the last contest and look at the software players. You already have the data.

Secretariat

07-13-2003, 07:38 PM

Dave Schwartz's comments below. He said :

"I will, of course, be open to further discussions about the rules of the contest but I really believe the points that are critical are:

1. Fixed number of bets per race
2. Play every race
3. Play by a set of published rules to remove the artfulness of the player."

Personally, I see no reason for a fixed number of bets per race. Each race offers a different challenge. Why play two horses when only one represents value, or why not play three at times, or show, or place? I do agree Exotics should be eliminated because one big hit distorts the totals.

Why not allow passes, and list passes as a component of recommended play? We should judge win percent, ROI, and pass percentage or money bet per race on average in the contest. This is one of the weaknesses of these contests, forcing a play in races where a good bet is not evident.

Dave, your number 3 is a little confusing to me. What do you mean by published rules of play. Can you provide an example?

Personally, I would rather see each software developer nominate a power user of his/her software. Primarily because they have a lot more time than I have.

I also beleive this power user should choose the track of his choice. Why not? This is closer to representing real conditions of usage. A very strong user of the program, playing in his own fashion, at the track of his choice. For example, there are some tracks I avoid, and like you some races I avoid.

I also think there should be an incentive to the contest. For example, free advertising on paceadvantage.com, for six months, and a free copy of everyone's software to the winning power user winner, and for those free software types with subscription prices, six months of free downloads from their sites.

Just some ideas.

JimG

07-13-2003, 08:13 PM

Originally posted by Dave Schwartz

I have first-hand knowledge of 6 people who consistently make $100k+ at the races. And they all have one thing in common (beyond using my software): If they chose to use someone else's software they'd probably still come out winners.

If six people are using your software and making over 100,000 a year doing so, I doubt that any other maker of software can boast to that.

Dave Schwartz

07-13-2003, 08:44 PM

Secretariat,

The original idea was to provide a contest that shows how well the software works in the hands of the typical user. This would demand that the performance not be based upon the ability of the user but the software itself.

That demands a rule-based approach.

Dave

Lefty

07-13-2003, 09:05 PM

I agree with Dave, must be rule-based contest so ea user of the particular software can duplicate the results.
Tom, I think the answer to your question who wins should be the software that makes the most profit based on a $2 unit.

Dave Schwartz

07-13-2003, 09:22 PM

Lefty,

I don't think there will be a winner... That is, just posting the results should be enough of a statement.

Dave

VetScratch

07-13-2003, 10:46 PM

JimG,
This is what you posted in response to Dave's statement that 6 people make $100K+ using HSH.
If six people are using your software and making over 100,000 a year doing so, I doubt that any other maker of software can boast to that.
I have seen plenty of boasts that top what Dave reported!

Everyone should remember that ALL handicapping performance claims are hollow boasts unless winnings can be objectively verified. Performance claims seem to be the common currency for marketing handicapping products. Dave would be operating under a severe handicap if he didn't do his share of boasting about the feedback he gets from users!

Here is how I look at it. For a claim to be taken seriously, it must be supported by gambling tax records, simply because ALL winnings MUST be reported to the IRS. Anyone who cannot do this is either a liar or a tax cheat, and I would never trust the claims of a tax cheat. Someone who cheats our country will not hesitate to cheat in other ways, including giving false feedback to Dave and other software vendors!

Dave Schwartz

07-13-2003, 10:59 PM

VetScratch,

I hope you did not view my statement as a "claim." Neither was it meant as a "boast." If that is what you got, then I'd suggest that you missed the intent of my message.

My point was that there are people doing it... I know of it because I know of people doing it that use my software... and that they could do it with other software as well.

Please resist the urge to turn this thread into another one of those "prove it or be convicted of lying" threads, I'd appreciate it. We're trying to accomplish something positive here.

Regards,
Dave Schwartz

VetScratch

07-13-2003, 11:22 PM

Dave,
There was no intent to impugn your credibility in the following exchange:

From JimG: "If six people are using your software and making over 100,000 a year doing so, I doubt that any other maker of software can boast to that."

From Me: "I have seen plenty of boasts that top what Dave reported!"

My references to "false boasting" were clearly attributed to users, and NOT to you (as a reporter).

I take your reply to mean that you can vouch for your users, and I accept that!

In fact, I am happy to surmise that you have a thriving tax accountancy to fall back on in case horseracing manages to self-destruct as an industry.
:) :) :)

Dave Schwartz

07-14-2003, 12:02 AM

LOL - Yeah, that would be me. Mr. Bean counter.

Dave

VetScratch

07-14-2003, 12:17 AM

Dave,

:) :) :)

I am currently facing an 8/15 deadline after filing an extension. Just in case, could you estimate what your firm might charge for a 1040 with 9 Schedule-Cs, 87 W2-Gs, 23 K-1s, and a 5-page Schedule-D itemization?

:D :D :D

Boasting???? Who's boasting????

Dave Schwartz

07-14-2003, 12:56 AM

VS,

So, now you are either teasing or goading... which is it?

Please let me know so that I may respond in kind.

Dave Schwartz

VetScratch

07-14-2003, 01:29 AM

I hope and pray that I was joking! 8/15 is almost upon us!:)

You can be sure, however, that I am equally enthusiastic about debating from a position on either side of most issues that pop up on this board.

Whenever I take a position, I always learn something from whatever ensues, either as a consequence of what I learn from research and contemplation about my adopted position, or from the results of similar efforts undertaken by those on the other side of an issue.

I also like to have fun along the way. :)

Hell, I've never been to a racetrack in my life! But after lurking and learning, then finally entering the fray, I feel like a reincarnated horseplayer from one of Damon Runyon's classic short stories!

Isn't that what cyberspace is all about?:D

Just Kidding???? Who's kidding????

Dave Schwartz

07-14-2003, 06:09 AM

So, which is it?

VetScratch

07-14-2003, 07:27 AM

Dave,

Like handicapping a race, the interpretation is up to you.

Without enigma and contradiction, horseracing wouldn't exist.

For example, how can this thread be aimed at a positive result?

After the contest is over, increased popularity will make failures out of whichever products performed the best in the contest.

Popularity spells Defeat.

That is the only immutable law that controls pari-mutuel wagering results in horseracing.

Is this bad? Not necessarily, as shown by the immense popularity of AllWays.

AllWays is the highest rated software ever reviewed by Philips, and has obviously been expertly crafted with good intentions and great care.

An ex-employee of BRIS has told friends in Cincinnati that AllWays downloads exceed $1.25-million per month.

If that is close to accurate, I would estimate that AllWays users bet at least $60-million per month.

Like any sophisticated system, most users do not become power users. They use AllWays like a black box.

As a result, AllWays has become an underlay generator, and so will your contest winners!

What a shame! But this contradiction between performance and results will allways exist.

rmania

07-14-2003, 08:45 AM

Well Dave,

After 2 days, 45 posts, and 1250 views it appears that you and I are the only developers willing to put it all on the line. And, since my "specialty" program doesn't qualify for such an event I guess its not going to happen. Too bad.

BTW, the quarter-pole view I described in the 5th post of this thread produced 4 out 4 winners yesterday at Hollywood and none were favorites. The payoffs were $13.00, $13.40, $21.40, and $11.00. But then who could be interested in a program that only gave them four plays.:confused:

Dave Schwartz

07-14-2003, 11:20 AM

Okay, I choose "goading." Go play with someone else.

We're trying to build a contest here.

rmania

07-14-2003, 11:45 AM

Originally posted by Dave Schwartz
I have first-hand knowledge of 6 people who consistently make $100k+ at the races. And they all have one thing in common (beyond using my software): If they chose to use someone else's software they'd probably still come out winners. Looks like you chose "goading" 20 or so posts back. But that's OK.. Doesn't bother me.

I'll go away so you can build YOUR contest but, from the way it looks, I doubt that it will ever happen.

I'll be around and up for any challenge should you change your mind.;)

Dave Schwartz

07-14-2003, 11:50 AM

RMania,

Sorry, got a telephone call before I could respond to your reply. My slight ws unintentional.

I am beginning to think that the idea of a contest isn't near as important as the concept of showing the public how our (notice use of the word "our") software works.

So, with or without other "contestants" the individual software developer can still do this. All that is needed is a place to post, some time to do the handicapping and a little courage. I'd think we can muster up all three.

Stay tuned for some more ideas on the subject.

Again, the slight was unintentional.

Dave

Secretariat

07-14-2003, 12:20 PM

rmania,

I don't think you and dave are the only two developers willing to participate, however, I strongly disagree with the rules Dave proposes for the contest.

I thought he was asking for discussion on the matter.

Personally, without a strong "passing poor races" aspect, and "without a playing your track of choice component" I would have serious reservations though about participating, as I don't use the software I develop that way, and would never advise a user to do so, so what would be the point.

VetScratch

07-14-2003, 12:25 PM

Rmania,
The payoffs were $13.00, $13.40, $21.40, and $11.00. But then who could be interested in a program that only gave them four plays.
Nice score! I find it incredible that you are willing to sell/share it!
Maybe you should just keep it for yourself.

Dave Schwartz

07-14-2003, 12:40 PM

Secretariat,

I am still open to discussion... after all, who elected me King? Sorry if I was not clear.

I simply put forth what seemed like a logical, "generalized" approach. Frankly, I would not expect to be profitable myself betting two horses blindly in every race... imagine a race where one had 3/5 and 7/5 as the top 2 selections... if ever there was a race that screamed to be passed it would be that.

But, if I can get to within (say) 4% of breaking even playing this way then a horse player should be able to carry it to the profit side from there by adding another factor, a contender selection process, some value rules or his own handicapping expertise.

I'd like to see a contest that focuses on the software and not the developer's handicapping skills.

But I am really open to whatever you have in mind. WHat are your specific thoughts?

Regards,
Dave Schwartz

rmania

07-14-2003, 12:40 PM

Originally posted by VetScratch
Rmania,

Nice score! I find it incredible that you are willing to sell/share it!
Maybe you should just keep it for yourself. Let me be the first to say that this doesn't happen every day.

But even if it did I question whether the odds could/would be affected by users of the program. After all, the program is designed for major cuircuts (currently SoCal only) and yesterday's handle at Hollywood Park nearly reached $18M.

Suff

07-14-2003, 12:55 PM

Originally posted by Secretariat

Personally, without a strong "passing poor races" aspect, and "without a playing your track of choice component" I would have serious reservations though about participating, as I don't use the software I develop that way, and would never advise a user to do so, so what would be the point.

Don't have a software....and watching from the back rows as this thing tries to get wings. I hope it does. Just a quick .02

To Much passing races, and track and race specific simply turn it in to a "Best bet" contest.

Are you absolutely sure the people who buy and use you sofware only use it for your "BEST BETS"

Your customers undoubtedly and hopefully use your product to compliment thier own experience.

And use it in races, and on circuits, and in ways that they handicapp and wager. Not in Robot formation with the 3 numbers it gives on one race at one track on two days a week.

Certainly a little tailoring to Balance the reality is called for.

Good luck.

pmd62ndst

07-14-2003, 02:01 PM

I have a database and homegrown software that shows an order of finish along with my own personal odds so if we did a win and place betting scheme, that would be fine with me. Here are some of my thoughts on the subject:

- I think we should play as many tracks and races as possible. Removing "artfullness" from the picks may be as easy as flooding the entries, therefore, making it difficult/not worth the effort to handicap hundreds of tracks 7 days a week.

- If anyone thinks believes for a second that they will see a positive ROI % in a contest like this is INSANE. The winner will probably be someone with a single digit negative ROI %.

- Any commercial software developer would be CRAZY to attempt this. I don't have anything to lose here. You do. I am curious as to how my software would compare to professionals. If I lose, that will make me work harder. If I win, I'll sell it.

PMD

VetScratch

07-14-2003, 02:25 PM

Rmania,
But even if it did I question whether the odds could/would be affected by users of the program. After all, the program is designed for major cuircuts (currently SoCal only) and yesterday's handle at Hollywood Park nearly reached $18M.
Odds will never be affected by your users so much as by your users' users!

Many rich owners and heavy hitters will never become any software vendor's user. However, they do "connect" with serious handicappers and use them as private touts. For example, if the "Stat Man" who has been keeping us posted on the AP pari-mutuel scandals is the same "Stat Man" that I remember when we used to run at AP, he will never have to put up a dime of his own money to come out a winner! All he needs is a phone.

Savvy users of programs like AllWays, HSH, and NetCapper often form parasitic relationships with clients who literally have money to burn. The result is like the "velocity of money" examples in Economics textbooks. Plenty of $2k-across-the-board bets are precipitated by spontaneous cell phone calls rather than serious handicapping.

Popular software programs generate far more handle than what seems logical for the size of their user bases. I would not be surprised to find that each licensed user shares information with at least four other parties.

lousycapper

07-14-2003, 02:33 PM

:D No point in this contest... :D cuz he ain't selling his program... no where, no how! Heh, heh!

-L.C.

penguinfan

07-14-2003, 02:50 PM

Originally posted by Dave Schwartz
I weighed in on this thread because PenguinFan chimed in with a rather provocative post about software devleopers not willing to allow his product to be measured. So, if anything is accomplished, it must include addressing THAT particular post.

Dave,

That comment was directed at a single individual who has been discussed probably more than enough. I have had no dealings with you either directly or indirectly so I apologize for it being a blanket statement and upsetting you. I will stand behind the reason I made the post though. Any software maker who touts his product as basically the Holy Grail and REFUSES to sample out even 1 card when I offered to pay for the data file has to send up a huge red flag. On top of that to claim the reason he can't do it is because he can't make the program send the results out as a file is rediculous, THEN to be told word for word how to do it and then decide he has made his last comment on the issue is enough for me to know I need nothing to do with him. I applaud the few of you who are willing to let the results tell the tale and those who are absent, well................

Again, I apologize for the blanket statement as it was directed toward a single person.

Penguinfan

Jake

07-14-2003, 03:30 PM

Why not make this simple and straightforward? Top 10 or 20 Best Bets for the day with a 20-1 odds cap, straight win and place bets. Any track and any type of race. Developer or his/her designated handicapper.

Forget about proofing the selection methods or systems...just demostrate what kind of bottom line the software can thrown when it is being used by a competent handicapper, chosen by the software developer. Win and place percentages and ROI will tell anyone with a brain whether or not the software is worth a damn.

Lefty

07-14-2003, 03:42 PM

Hmm, so far Dave and Rmania only commercial guys willing to put their software on the "firing line" Where are all these other guys?
Where's Gordon Pine, Ken Massa, Tom Console, Joe Zambuto, the Propace guy and
Handle? These are only, a few off top of my head.
Time to chime in, one way or another, guys.

Speed Figure

07-14-2003, 03:45 PM

I knew it would be only 2 or 3 "program maker's" that would do this.:(

rmania

07-14-2003, 03:49 PM

Originally posted by speed figure
I knew it would be only 2 or 3 "program maker's" that would do this.:( And what made you think that ? ;)

hdcper

07-14-2003, 03:58 PM

I have to agree with Jake on this one, keep it simple and straight forward.

Most important thing in my opinion is if you do decide to participate you agree on the time frame and everyone commits to it, no matter how things are going.

So use the KISS method, "Keep it simple stupid".

Bill

Myhorse1_X

07-14-2003, 04:06 PM

I would enter the contest, but I have a few questions first. There got to be rules.

Do we just put down the picks as the computer does without any handicapping from us, or do we pick the top 4 and handicap down to 1 or 2???.

Do we handicap to win, or do we do exacta's and trifecta's???

How many horses to we pick???1,2,3 or 4???

Do we do all races, or just the races we would normally play???

Let me know what the rules are, and I might give it a shot.

MyHorse1

rmania

07-14-2003, 04:54 PM

Originally posted by Myhorse1
Let me know what the rules are, and I might give it a shot.
If I were making the rules (which I'm not) they would go something like this:

1. Each programmer would have to submit, for approval, the method they intend to use for extracting their selections. The method would have to be one that limits the amount of user input. After all we are looking to evaluate the software, not the user.
Approval of any proposed method would come from the other contestants.

2. Each contestant would choose one track of their choice.

3. Contestants could enter 1 race up to the entire card.

4. Contestants could list up to 4 horses per race with "points" awarded for win only. Points would be scored and tallied using the follow method:

- a winner with 1 horse listed receives 100 points
- a winner with 2 horses listed receives 50 points
- a winner with 3 horses listed receives 25 points
- a winner with 4 horses listed receives 10 points

5. The one with the most points at the end of the contest wins.

Except for item 1 this looks pretty simple to me.

Dave Schwartz

07-14-2003, 05:00 PM

PenguinFan,

Apology accepted. I realize it was not directed specifically at me but, as I certainly fall into the category of "software maker" blanket statements to include me by default.

So, if, in the future you wish to vilify software vendors in general, please add a disclaimer like "does not include the ever-saintly Dave Schwartz" afterwards. <G>

Dave

penguinfan

07-14-2003, 05:13 PM

Originally posted by Dave Schwartz
PenguinFan,
So, if, in the future you wish to vilify software vendors in general, please add a disclaimer like "does not include the ever-saintly Dave Schwartz" afterwards. <G>

Dave

Thanks for replying, and it shall be so noted oh saintly one.:) :) :)

Penguinfan

penguinfan

07-14-2003, 05:15 PM

And FWIW, I would like to see how the top two picks only fare in such a contest. Show me the winners....

Penguinfan

VetScratch

07-14-2003, 05:50 PM

And who will be in charge of deriding, taunting, heckling, teasing, goading, debunking, and general button-pushing during and after the contest?
:) :)

Bruddah

07-14-2003, 06:13 PM

I have been reading this thread with interest. I am not knaive enough to think all software developers are actually doing so for the sake of mankind and horse players in general. I do believe most feel a need to search for knowledge of this game. They have abilities to create tools (software) to discover and organize what we pen and paper cappers have been doing for years. Obviously, some are better and further ahead than others. The individuals I have come to admire are the ones continually striving to improve their product. The individuals who continue to pursue excellence through all the criticisms, arguments and false claims of motivations by those less talented. To these individuals I take my hat off and say " Thank you for sharing your labor and creativity ". I happen to be one who is always behind the curve when it comes to handicapping software. Basically, I have found one which suits and works for me (Equisim). The dedication of the developer is unbelievable. Do I think his product is perfect? No and neither does he. Do I think maybe others have products as good? Maybe and so does he. But, here is where the real contest starts and ends. He keeps making it better and easier for us non tech nerds to use. User friendly is the key to any software program. JMHO from an old pen and paper capper.

Oh yes, I really do mean Thank you guys for sharing your work and creativity. You guys are tops in my book.

JimG

07-14-2003, 06:27 PM

Why don't the software developers do Calder-only and Karl can give his picks and see if he out handicaps the software developers.

Would be interesting to see.

Jim

rmania

07-14-2003, 07:09 PM

Handicapping software should not be confused with electronic versions of past performances (i.e., DRF, BRIS, TM, or ITS).

If written properly, handicapping software should provide the user with an automatic edge over conventional handicapping methods. Though the degree of success should increase with use and understanding, the software should contain some attribute that could lead to some success without ever getting past learning how to download the data.

To say a user cannot just push a button and be profitable is not necessarily true.

I would guess to say that most programs contain at least one attribute/screen/column of data which requires no input on the part of the user and, when used in its basic form, produces a certain percentage of winners. If not then maybe it’s time to go back to the drawing board.

IMO, these attributes are what should contested. In other words, how does the software perform on its own.

Secretariat

07-14-2003, 07:35 PM

To Dave Schwartz,

I will throw out my two cents on this

Rule 1.

100 plays will be made by a developer over one month. After the 100th play the handicapper in done. This way each handicapper makes the same number of plays, and passing is not an issue. To keep it simple exotic wagers are disallowed, although win, place, or show are all allowed. This means every develper is ideally betting 200.00 (2.00 wagers). I'd be willing to up this to 200 plays. That is approximately 20 cards.

Rule 2

The track or races chosen is at the discretion of each software developer, as is passing.

Rule 3.

The software developer must list the way/rule/feature the way the horse(s) are to be chosen by the program. For example, I might make my selection process as follows: Selections are chosen by clicking the Black Box option off the menu, and demand a Win% minimum of 20%, ROI minimum of 1.30, and a minimum database of at least 50 starts.

This way users can see how the software developer made the pick, and why he passed various races.

I feel this meets your issue of every handicapper playing an equal number of races. It tells users how selections were arrived at during the beginning of the contest. And it gives the developer the option to cherry pick the races they feel best about. One could argue that one should play just CRC since everyone is on an equal footing. Or someone else might say,play a variety of races such as 5 stakes races, 20 claimers, 15 allowances, three 2 year old races, statebreds, non-winners, ad infinitum, because it shows skill at all races. Or another might say play a variety of tracks from small tracks to large tracks, to east coast to west coast, etc.

I think the three rules I list above encompass enough variety for any developer. Basically, it gives them flexibility, and informs users PRE RACES how selections were arrived at. I'm OK with shipping picks to a neutral poster like Paceadavantage.com at least 10 minutes prior to race, or even night before play if necessary. Selections scratched would not have some weird assignment of play. They would simply be disregarded as plays.

Anyway, these are some of my ideas. I think it could happen if we try to arrive at an agreed set of rules, and leaving the bashing developer stuff out of this thread. Frankly, I am impressed with a lot of developers I've met, and the ones I meet do it because they love horse racing more than trying to rip people off.

Oh, and I still think adding the power player should be included. After all, how does one know the developer is really sending in the pick, rather than a power user anyway. I guess maybe the selection process rule above might cover that though.

Anyway, Dave, I had the original Thorobrain and have nothijng but much respect for your work. If we can all agree on the rules thing, it would be an honor to match my stuff against yours.

VetScratch

07-14-2003, 09:15 PM

If you actually expect potential users to follow this contest with keen interest, and treat the results as credible, you had better KISS up to the users instead of the software developers:

1) Play every race in the same racecards at the same tracks.
2) Play at least 12 tracks, representing minors/mids/majors with some night-time tracks included.
3) Play at least 5 cards per track (approx. 550 total races).
4) Preclude the taint of human bias by designating blackbox run-time options that any user can effortlessly duplicate.

This assumes that none of you include an indentured expert handicapper with each license. If you let the human element in, it is no longer a software contest.

Dave Schwartz

07-14-2003, 09:19 PM

Secretariat,

I'd say we are getting closer to the goal... Might I suggest we could perhaps reach a common consensus with a private telephone conversation?

And thank you for the kind words. They are appreciated.

Dave

Handle

07-14-2003, 11:06 PM

Originally posted by Lefty
Hmm, so far Dave and Rmania only commercial guys willing to put their software on the "firing line" Where are all these other guys?

Secretariat seems ready to go too there Chuck.

I think the contest would be great for all software vendors. The old adage - any publicity is good publicity. If you don't believe me, just ask RMania....

But, I have to say that I'm more than too busy right now to participate. Version 4.0 of EquiSim is in Beta now and I still have the User's Guide to finish up, then to fix all the bugs, and then to ship all the disks. Sigh. Yes, I could throw in the simulation winner, or top profiler, or top whatever picks if the rules weren't too byzantine, but EquiSim actually tracks how well these things do anyway -- so there you go. I'd consider entering such a challenge if:

A) The rules were dirt simple. I see Secretariat's point -- EquiSim has a lot of features designed to help you spot where the best plays are, but I'd rather approach this the easy way. Plus, Vet Scratch has a good point about this. What's the purpose? I presume to gain recognition with the handicapping masses. So design the competition so that it appeals to those "masses". This generally means KISS.

B) Someone other than the software vendors needs to create the rules for the competition. Vendors can vote to play or not to play (its a free country), but have no say in the rules. I think a guy like AnotherDave would be one good person to have on the comittee, maybe CJ, or Suff. They are generally level headed, certainly know the game, and they show no particular alliance (or abhorrence) to any (commercial) handicapping software. Kitts also comes to mind -- he's seen just about everying, VERY knowledgeable, but he doesn't jump on band wagons easily. JimG would be excellent. He always brings an unbiased, objective view point (trust me), is very involved in the game, and knows tons about handicapping software. He might be objectionable to some due to his past and present "relationships" with various developers. Anyway, my point is that there are hordes of folks on this board that are objective (and some that are objectionable...) and/or unaffiliated that could come up with some fine rules if they were willing to do so. I'm not going to sit around and bicker about the proper format for the contest though.

C) Forgot -- each deverlop picks the same number of tracks to play throughout the contest. That is, while the rules should be decided by a third party, it makes no sense to stick everyone on the same track, while it becomes more of a "handicapping" endeavor if you are finding your own track to play unless you are doing so with an algorithmic approach.

Of course, if someone wants to put up prizes for winning, I guess they then get to name the tune.

-Nathan

jackad

07-14-2003, 11:13 PM

Secretariat,
What is your stuff? What software?
Jack

cato

07-14-2003, 11:57 PM

Reality check: What is proposed will never happen and if by some miracle it did happen, there woudl be so many odd rules and caveats that it would be meaningless if someone is trying to find teh "best" software .
Plus it would be boring since it would be restricted to a handful of people who have delveloped a program.

If folks want a contest I suggest that we have a contest among teams of board members who use different programs. That is, the folks who use HTR vs HSH vs All-in One vs Equisim vs Handicaping Magic, etc etc.

The only qualification (and we would have to accept people's word on this) is that they use only their team's progam for the contest.

Just a thought.

Another fun contest would be similar to the Gulfstream "Survivor"contest where each participant is matched against another partcipant for a day or weekend in a single or double elimination contest.

Cato
My suggestion

Handle

07-15-2003, 12:03 AM

Cato,

What you describe was, in many respects, what the PA Inter Board contest was. We all formed teams and had a contest that turned out to be a whale hunt (though, I must add, we at ThoroTech do love whale meat <G>).

-N

Secretariat

07-15-2003, 12:14 AM

To Jackad,

In answer to your question, I've programmed for a variety of handicappers, but I personally use my own program Lightning Profits.

VetScratch

07-15-2003, 12:37 AM

Secretariat,
In answer to your question, I've programmed for a variety of handicappers, but I personally use my own program Lightning Profits.
No doubt about it, obscene profits can be a heavy and dangerous burden to take home safely. :) :)

cato

07-15-2003, 02:57 AM

Handle: What I described has a superficial apearance of a repeat of the Board handicapping cntest but, in fact hs little similarity.

The borad handicapping contest was just that--based on membership of boards. While some of the boards were at least superficially related to a program, there was no restriction or requirement tht people had to use specific handicapping programs in picking selections. And many of the boards had no connection to any program. So i think its completely different and would be fun...but that's just me.

Cheers, Cato

JustRalph

07-15-2003, 04:32 AM

Originally posted by cato
Handle: What I described has a superficial apearance of a repeat of the Board handicapping cntest but, in fact hs little similarity.

What? There won't be 78 Wingnuts from other boards coming in and calling us Shitheads and computer Geeks? Then running back to their board to gloat about it. That would be much different from the last contest.

Myhorse1_X

07-15-2003, 04:35 PM

I think that Dave should get to set the rules of the contest.
After all, we are all playing in his sandbox!!

MyHorse1

Lefty

07-15-2003, 05:09 PM

I like Dave's rules, but doesn't seem to be much interest among the "biggies" besides Dave and you Myhorse(Dynaform)
Of course since it's PA's board he might have some good idea about the rules. I think this could be a most interesting and informative contest so I hope it gets "off the ground."

Lefty

07-15-2003, 05:16 PM

Ooops, and include Rmania in that last comment too. Didn't mean to slight you Rmania.

Suff

07-15-2003, 06:04 PM

if the makers are just giving Thoroughbreds that the Software generates...why do THEY need to post the numbers.

Get a couple of users of each software to stand up and say..

I'll post "X" companies selections.

Then you'll see a run from the Majors to respresent themselves.

Little strong arming never hurt anyone........much.

canuck

07-15-2003, 06:10 PM

TEM Plays Wins WN% $ROI WP% Long--High
--------------------------------------------------
K1 02044 0644 32% 0.89 52% 0009 $42
K2 02044 0369 18% 0.73 37% 0026 $40
K3 02044 0317 16% 0.82 33% 0050 $61
K4 02044 0261 13% 0.84 26% 0074 $68
K5 02019 0181 09% 0.79 20% 0085 $62
K6 01903 0130 07% 0.86 16% 0090 $121
K7 01590 0064 04% 0.57 11% 0045 $86
K8 01156 0051 04% 0.92 10% 0043 $220
K9 01372 0029 02% 0.49 05% 0027 $125

K11O 00486 0190 39% 0.84 63% 0000

$ 02095 0260 12% 0.80 26% 0070 $68

Ev-1 02043 0375 18% 0.87 34% 0055 $76
Ev-2 02029 0344 17% 0.84 32% 0057 $69
Lv-1 02040 0342 17% 0.77 32% 0053 $51
Lv-2 02022 0347 17% 0.76 32% 0037 $79
VEL-1 02094 0482 23% 0.80 41% 0030 $64
VEL-2 02071 0388 19% 0.88 36% 0050 $101

C90! 00489 0124 25% 0.75 45% 0004 $25
C90* 01720 0420 24% 0.87 45% 0031 $42

Pscan Range Plays Won Wn% $ROI Long-High
--------------------------------------------------
0 to -1.99 04936 106 22% 0.79 0062 $125
-2 to -3.99 03700 493 13% 0.82 0111 $121
-4 to -5.99 03398 298 09% 0.84 0151 $101
-6 to -9.99 04182 193 05% 0.66 0125 $220

Races Tested: 2044

*Longs 22%
*HTRLongs 12%
*CHALK 57%

TRK: ALL
D/S: All
Type: Alw/Stk Clm Msw Mcl
Age: 2yr 3yr 3up/4up
Sex: Female Male
P/L: 5
Vi = Any

This is the output from the testing module from 8 tracks in the last 45 days-2044 races

You will notice that the top 2 rated horses win 50% of the time

So why go through all this melodrama??

Post the output of your top 2 rated strongest horses over all condtions distances surfaces and you will have a good starting point

Suff

07-15-2003, 06:12 PM

exactly.

canuck

07-15-2003, 06:33 PM

Same tracks--same time period-BUT--female allowance turfers going a route

ITEM Plays Wins WN% $ROI WP% Long--High
--------------------------------------------------
K1 00038 0016 42% 1.34 58% 0000
K2 00038 0006 16% 0.53 26% 0000
K3 00038 0006 16% 0.57 37% 0000
K4 00038 0005 13% 0.98 26% 0002 $28
K5 00036 0002 06% 0.33 19% 0000
K6 00033 0001 03% 0.57 06% 0001 $38
K7 00029 0001 03% 0.10 17% 0000
K8 00023 0000 00% 0.00 13% 0000
K9 00030 0001 03% 0.80 10% 0001 $48

K11O 00013 0005 38% 0.75 54% 0000

$ 00038 0008 21% 1.38 32% 0001 $28

Ev-1 00038 0003 08% 0.29 21% 0000
Ev-2 00039 0004 10% 0.89 23% 0001 $48
Lv-1 00038 0007 18% 0.63 42% 0000
Lv-2 00038 0006 16% 0.34 29% 0000
VEL-1 00039 0009 23% 0.84 41% 0000
VEL-2 00041 0005 12% 0.67 24% 0001 $28

C90! 00022 0010 45% 1.45 59% 0000
C90* 00019 0003 16% 0.56 21% 0000

Pscan Range Plays Won Wn% $ROI Long-High
--------------------------------------------------
0 to -1.99 00087 023 26% 0.95 0001 $28
-2 to -3.99 00081 010 12% 0.50 0000
-4 to -5.99 00062 004 06% 0.65 0002 $38
-6 to -9.99 00073 001 01% 0.33 0001 $48

Races Tested: 38

*Longs 11%
*HTRLongs 08%
*CHALK 61%

TRK: ALL
D/S: Turf Routes
Type: Alw/Stk
Age: 3up/4up
Sex: Female
P/L: 5
Vi = Any

canuck

07-15-2003, 06:40 PM

Maiden male claimers going a route

ITEM Plays Wins WN% $ROI WP% Long--High
--------------------------------------------------
K1 00031 0013 42% 1.22 65% 0000
K2 00031 0007 23% 0.81 35% 0000
K3 00031 0006 19% 1.05 35% 0001 $20
K4 00031 0003 10% 0.67 32% 0001 $23
K5 00031 0001 03% 0.22 16% 0000
K6 00031 0000 00% 0.00 03% 0000
K7 00029 0001 03% 0.49 07% 0001 $28
K8 00020 0000 00% 0.00 05% 0000
K9 00020 0000 00% 0.00 05% 0000

K11O 00004 0003 75% 1.48 100 0000

$ 00032 0003 09% 0.53 25% 0001 $20

Ev-1 00031 0010 32% 1.05 45% 0000
Ev-2 00031 0003 10% 0.66 26% 0001 $20
Lv-1 00031 0003 10% 0.28 26% 0000
Lv-2 00031 0013 42% 1.73 45% 0001 $20
VEL-1 00031 0006 19% 0.77 29% 0001 $20
VEL-2 00035 0008 23% 0.74 40% 0000

C90! 00000 0000 0000
C90* 00033 0010 30% 0.96 48% 0000

Pscan Range Plays Won Wn% $ROI Long-High
--------------------------------------------------
0 to -1.99 00060 017 28% 1.02 0000
-2 to -3.99 00048 007 15% 0.59 0000
-4 to -5.99 00057 005 09% 0.39 0001 $20
-6 to -9.99 00090 002 02% 0.28 0002 $28

Races Tested: 31

*Longs 10%
*HTRLongs 06%
*CHALK 65%

TRK: ALL
D/S: Fast Dirt Routes
Type: Mcl
Age: 3up/4up
Sex: Male
P/L: 5
Vi = Any

canuck

07-15-2003, 06:53 PM

2044 races tested--50% winners on top 2 rated but a flat bet loss of 11 cents on the dollar-(hey-it beat the takeout)

Go with top rated only female alw turfers going a route at the same tracks and time period and you get a profit of 34 cents on the dollar.

Go with top rated only male maiden claimers and you get a profit of 22 cents on the dollar.

Can you see the point I am about to make??

A contest where the developer is handcuffed into playing every race will prove NOTHING

A contest where the developer says--Hey--my data has been killing em in sprints for restricted male claimers less than 20k--lets post em up!!!

Now THAT is a contest I would be very interested in!

I am already interested in rmanias program and the contest has not even begun!!!!

(rmania--my email is canuck_steve@yahoo.com)

hurrikane

07-15-2003, 08:44 PM

You know Canuck...I was going to say that every month in every newsletter you can see results of different factors of HTR.
Go to their web site and you will constantly see questions, queries and answers to performance of the software.

What you propose here is done every day at HTR. On thing you have to give Massa...he throws his data out there..good and bad..you make the difference.

I'm not saying HTR is the best or better. Just that I don't think this thead is relevant.

It really comes down to what Dave said and was riddled about

Six people that use his software make over six figures a year but THEY WOULD PROBABLY DO THAT WITH ANY SOFTWARE!!!!!!!

that is the key. Not the software. How you use it!

canuck

07-15-2003, 09:00 PM

Kane

Not sure if you are agreeing with me or what...You said :I'm not saying HTR is the best or better. Just that I don't think this thead is relevant"

I would say if this thread is not relevant--where would it be? And relevant to what??

My point was not an endorsement of HTR per se...I was spitting out data to illustrate that HTR shows a loss--as would EVERY prog--if you bet 2044 races at all dist surf conditions etc. with its BEST single factor--the K1 BUT it shows a tidy profit if you know where to look.

This contest should be about the art of winning MONEY !!!!

Is that not what it is ultimately about???

Show Me the Wire

07-15-2003, 09:25 PM

Originally posted by hurrikane@HTR
what Dave said and was riddled about

Six people that use his software make over six figures a year but THEY WOULD PROBABLY DO THAT WITH ANY SOFTWARE!!!!!!!

that is the key. Not the software. How you use it!

I have trouble understanding this concept.

This conclusion does not make sense. If this conclusion is true then there are two ways to interpret the conclusion. One way is the same results would occur because different sw will give the same exact output the users are currently utilizing. The other scenario is the sw is totally meaningless to these select few and they could achieve the same results with pencil and paper. Since all sw does not produce the same output the conclusion has to be sw is either useless or valuable without factoring in the user.

I understand one user may interpret the data differently than another user and I am not talking about interpretation of data. I am talking about the type of information and the quality of the information to base the decision making process on. If these people could use any sw program as effectively as HTR or HSH then there is no difference in the type of information and quality of information provided by either program, therefore either program itself is not valuable. If I invest in sw I want a valuable product meaning it will improve my performance. Therefore, Dave’s original statement and your affirmation of this conclusion do not make sense to me.

Regards,
Show Me the Wire

Perception is reality

07-15-2003, 09:49 PM

I would think if software is good, it has to give information not readily available to player's not using it. A good player will learn how to use it, and a bad one won't. Different programs probably have different good information. The good player will figure out how to use it for profit, the bad one's won't.

Show Me the Wire

07-15-2003, 09:58 PM

cj:

I agree with you, not everyone will be as proficient with the same tool. But if you use a better tool it should make your life easier i.e. using a hammer instead of a rock. The hammer will improve my performance, but probably not to the level of a carpenter.

The conclusion I do not understand is the no difference in performance. If the people who are now making 100K using a particualr program switch, I would think their performance would suffer if the use a less valuable tool or increase if they are using a more valuable tool. However, to conclude their performances will remain the same indicates the tool itself is not valuable.

Regards,
Show Me the Wire

Perception is reality

Tom

07-15-2003, 10:00 PM

All the software vendors should give each of us a free copy of their programs, and if the require monthly downloads, they pay for those as well (after all, it was THEIR idea to require that, not ours!)
Then, we all use the programs for a year of two, and report back which one we like the best.
And of course, at the end, we return the programs.

(You can email me for my address where to send the discs)

:rolleyes: :rolleyes: :rolleyes: :rolleyes: :rolleyes: :rolleyes: :rolleyes:

VetScratch

07-15-2003, 10:55 PM

Show Me The Wire was right on target. This thread started by proposing a software contest. It now seems headed towards proposing a power users' contest.

If vendors can assemble their teams of 100K winners and then induce them to divulge all their secrets, I will abandon button-pushing and become a nun!

As a prospective software user, I don't care which vendor can divert a team of champions from their lucrative pursuits and induce them to volunteer for a handicapping contest.

Let the software compete, and trust that potential users already know they will have use their wits to join the ranks of 100K winners! As Show Me said, to do otherwise is practically an admission that your tools are indistinguishable and interchangeable. If that's true, commodity pricing should rule the marketplace, which is not what you want.

PS:
Dave, I know you pointed me towards the exit once already, but.... http://www.dlc.fi/~frog/wavs/sorydave.wav

Dave Schwartz

07-16-2003, 12:21 AM

SMTW,

>>>The conclusion I do not understand is the no difference in performance. If the people who are now making 100K using a particualr program switch, I would think their performance would suffer if the use a less valuable tool or increase if they are using a more valuable tool. However, to conclude their performances will remain the same indicates the tool itself is not valuable. <<<

I cannot explain it either, but... I have seen it over and over again. A player calls me up and tells me they are winning substantial sums of money with a particular approach that they are doing by hand in my software... combining the output of this screen and that screen, this field times that field, etc.

They want a custom screen that does precisely what they are doing by hand. I build it for them (i.e. it shows only in their serial numbered program).

Of course, this gives me access to EXACTLY what they are doing so I give it a good test. It makes 4% instead of the 16% they touted. The only difference is that I am using absolute final odds, while they are forced to play in the real world which means 2minute-odds. That just CAN'T be the difference, can it? So, I try it live for a couple of hundred races and see slightly worse performance. I think, "How can this be?"

I cannot explain it but I have seen it again and again.

For whatever reason, the individual that makes the system work, MAKES THE SYSTEM WORK.

That's why no two people get the same ROI using the seemingly-identical approach.

Somehow, this is beginning to actually make some degree of sense to me just because I have seen it so often.

Listen, we know that their are SOME people making significant money at the window. (I am not one of those, those I hope to change that.) Maybe you even know someone that does better than you and you really can't figure out why. You may even use the same program, play the same races but they consistently do better.

I contend that they have simply found a way to "make it work," and the differences can be very subtle.

And before the naysayers jump on me, consider that it would be to my advantage if I could tell this story with a different spin. It would be so much easier if I could simply say, "Click this button and you get a +12% ROI." I just can't promise that.

I'll tell you an interesting test... Have 5 people who don't know each other play the same exact approach at 5 tracks for 5 days. Something relatively simply like play any horse that ranks in the top 2 for ES Pts and last race Beyer that goes off at 3/1 or more. Have them compare results. I'd bet that they'll come up with different results. Similar, perhaps, but different nonetheless.

Now, they certainly won't if they use the same data and run a query but if they actually play them one race at a time...

I could be wrong, but it just SEEMS to be what I see.

Regards,
Dave SChwartz

Dave Schwartz

07-16-2003, 12:23 AM

BTW, I did not say that their performance remained the same... how would I know that? I said that they would find a way to win with just about any software.

In other words, the software does not MAKE them win. It ENABLES them to win.

There IS a difference.

Dave

Show Me the Wire

07-16-2003, 12:53 AM

Six people that use his software make over six figures a year but THEY WOULD PROBABLY DO THAT WITH ANY SOFTWARE!!!!!!!

Dave:

To me the above statement says they users will perform the same no matter what program they are using. If that is not what you meant that is fine, especially if you feel confident that any of your current big winners efficiency will decline if they suddenly switched to another program. I agree they may always be winners, but there should be a decline in performance if an inferior tool is used.

Also, you can say the DRF enables people to win that is not the point. The point is at least for me I want a better tool and by definition a better tool will improve my perfromance.

When you say people ask for customized screens that is a horse of a different color. You are providing a program to do a specific item and therefore, your customization makes the program valuable to the user. If this user changed programs and lost the ability to have the customized output his efficiency should decrease.

Enabling someone to win using your basic program is value if that person could not win without it. There is the subtlety that needs to be clarified.

Regards,
Show Me the Wire

Perception is reality

VetScratch

07-16-2003, 05:29 AM

Tom,
What about the free updates and annual renewals during the evaluation period!!! ;) ;)

hurrikane

07-16-2003, 09:03 AM

SMTW,

I'm not sure what the confusion is about but I can give you my opinion only.

Before HTR I was a winning horse player. Played a couple of tracks, DRF and some record keeping.

With HTR(and probably most software) I can organize the data, find pieces that are working and what is not working, build data, test, confirm(for confidence), and bet. I can handicap(if that's what you want to call it) every track in NA in about 15 minutes and find 15-20 plays a day. Before HTR I could do 3 tracks in about 4 hrs and maybe find 6 plays a day.

I don't know 2 people in HTR that use the software the same way.
There were about 8 of us at the Orleans playing across I believe 8 tracks for 3 days.
Although a couple of times about half of us would had the same horse, most of the time we were playing different horses. SAme software..different methods of attack.

It really is not entirely the software. I have data just like people using the DRF or Bris. How I chose to use that data is different than how everyone up here would use the DRF or Bris.

I don't believe any software with black box picks can produce a positive ROI but how you use the software will put you in the black.

Without plugging HTR I will say, Massas data is as accurate or more accureate than anything out there. He is also right out front about the data, what it produces and the effectiveness of the numbers. You don't see that a lot. Maybe not at all.

VetScratch

07-16-2003, 09:51 AM

Dave, Hurrikane, and other vendor voices have certainly made their case that results depend on the user. I think everyone knew that anyway. What puzzles me, in light of what has been proposed, is why persist in calling it a software contest instead of a users' contest? And if it's going to be users' contest, why bother to identify which product any user employs? That would seem consistent with arguments presented by the vendors. As SMTW has pointed out, you can't have it both ways without more clarification.

What is wrong with simply letting the software perform default or blackbox handicapping? Intelligent users are looking for a nice handicapping tool, not alchemy. As long as your results don't substantially underperform public handicapping results (which we already know are unprofitable), what is the risk?

Software contest or power users' contest? Why not take a poll?

Show Me the Wire

07-16-2003, 11:01 AM

hurrikane:

The source of my confusion is are the programs, at a certain level of sophistication, interchangeable or are the program themselves valueless?

If I understand your post, I take it to mean any program that allows you to build your db and test that db is valuable to you. So essentially you are saying the program you use is interchangeable with any other program that offers the same capabilities. Also, I understand the value of time and as you mentioned that is value to you.

In other words, if I understand you correctly a hammer is a hammer and it depends how the user holds the hammer. I will similar results with ABC's hammer as I would get from XYZ' hammer, depending how I hold it.

As a result of your response, I am interpreting your answer that the program is interchangeable with others and the program was not instrumental in making you a winning player. Additionally, I will concede the program is valuable if it increases your roi since using the program through either suggesting new plays or eliminating your prior losing plays.

As I pointed out before the program is valuable itself if the program is instrumental in creating a winning player, from a losing player. People buy software to become winners. They are willing to try XYZ software; because the player is not satisfied with his current results. If the software cannot improve the purchaser's pre- use results it is worthless to the purchasing player.

Additionally, this question does not apply to custom or home grown software, it only applies to the stock commercial programs

Does the above clarify why I was confused about the statement regarding that a group of specific players would achieve the same results regardless of the program?

Regards,
Show Me the Wire

Perception is reality

Dave Schwartz

07-16-2003, 11:55 AM

VetScratch,

The idea is that the software should create a "baseline" performance of some kind. WIll individual mileage vary? You bet. <G>

Dave Schwartz

Show Me the Wire

07-16-2003, 12:58 PM

Dave:

Yes, that is it. A commercial program should be measured by its "baseline" performance, and to be fair and objective baseline performance should be measured in contender selection, not roi. Roi is too subjective and is influenced by the reality of past posting (lol) and other exteranl factors. However, contender selection is very objective.

I propose if the software improves the user's contender selection it is valuable. For example if the user prior to using the program on the average has the actual winner in his top three (appropriate amount for field size) contenders only 50 % of the time and that same player experiences a significant increase of inclucing the winner 80% of the time due to the program, this makes the program valuable. This is the threshold test for value.

The question is can the selection process be improved by this specific program? If the program cannot increase the user's selection process the value lies not within the program, but the user.

Regards,
Show Me the Wire

Perception is reality

rmania

07-16-2003, 01:18 PM

“I would think if software is good, it has to give information not readily available to player's not using it.” Cjmilkowski

“Enabling someone to win using your basic program is value if that person could not win without it. There is the subtlety that needs to be clarified.” Show Me The Wire

“Since all sw does not produce the same output the conclusion has to be sw is either useless or valuable without factoring in the user. “ Show Me The Wire

“What is wrong with simply letting the software perform default or blackbox handicapping? Intelligent users are looking for a nice handicapping tool, not alchemy. As long as your results don't substantially underperform public handicapping results (which we already know are unprofitable), what is the risk?” VetScratch

“A contest where the developer is handcuffed into playing every race will prove NOTHING” canuck

“As a prospective software user, I don't care which vendor can divert a team of champions from their lucrative pursuits and induce them to volunteer for a handicapping contest.
Let the software compete, and trust that potential users already know they will have use their wits to join the ranks of 100K winners! As Show Me said, to do otherwise is practically an admission that your tools are indistinguishable and interchangeable. If that's true, commodity pricing should rule the marketplace, which is not what you want.” VetScratch

“As I pointed out before the program is valuable itself if the program is instrumental in creating a winning player, from a losing player. People buy software to become winners. They are willing to try XYZ software; because the player is not satisfied with his current results. If the software cannot improve the purchaser's pre- use results it is worthless to the purchasing player.” Show Me The Wire

And my favorite quote.....
“This contest should be about the art of winning MONEY !!!!
Is that not what it is ultimately about???” canuck

jackad

07-16-2003, 01:34 PM

Secretariat,
Is your program Lightning Profits for sale? If so, what is the URL?
Thanks.
Jack

VetScratch

07-16-2003, 02:06 PM

Dave,
The idea is that the software should create a "baseline" performance of some kind. WIll individual mileage vary? You bet. <G> Catch up with the rest of us. We all agreed with what you said ages ago. The question is: software contest, or user contest, or more waffling?

rmania

07-16-2003, 02:17 PM

Originally posted by Show Me the Wire
Dave:

Yes, that is it. A commercial program should be measured by its "baseline" performance, and to be fair and objective baseline performance should be measured in contender selection, not roi. Roi is too subjective and is influenced by the reality of past posting (lol) and other exteranl factors. However, contender selection is very objective.

I propose if the software improves the user's contender selection it is valuable. For example if the user prior to using the program on the average has the actual winner in his top three (appropriate amount for field size) contenders only 50 % of the time and that same player experiences a significant increase of inclucing the winner 80% of the time due to the program, this makes the program valuable. This is the threshold test for value.

The question is can the selection process be improved by this specific program? If the program cannot increase the user's selection process the value lies not within the program, but the user.

Regards,
Show Me the Wire

Perception is reality Just thought this was worth repeating.;)

JimG

07-16-2003, 03:06 PM

Just to scratch the surface of HTR, HSH, and Equisim you could do a simple thing like compare the top two picks using basic factors of each program and see how they perform without user intervention. Since I have used all 3 programs at one time or another over the last 4 years, I think you will find all 3 of them pretty strong right out of the box. If one were to do this, I would recommend KM Rating (HTR), CompRtg (HSH) and Simulation (Equisim) as good basic ratings of all 3.

What would the above prove, not much really, but I bet all 3 while showing a probable loss over several hundred races would probably beat the track take (-18%) of betting the fave to win in each race. This is significant why???.....well, I do not have alot of time to devote to playing and when I do I like to be able to get to the contention in a race quickly. All 3 of these programs do an excellent job in this area in my opinion, using just the basic ratings. If your a power user with any of these programs, I think the sky is the limit to what a sharp capper can achieve due to how quickly information can be processed over a myriad of tracks.

Jim

Show Me the Wire

07-16-2003, 03:40 PM

JimG:

It seems we are rehashing the same item. I proposed a baseline test using contender selection.

As VetScratch said are we going to have a software contest or what?

Regards,
Show Me the Wire

Perception is reality

Dave Schwartz

07-16-2003, 05:02 PM

VetScratch,

Waffling?

It would seem that you are simply looking to stir things up. I do not find it either enlightening or entertaining.

I will try to resist the urge to respond to your posts on this subject in the future.

Dave Schwartz

Dave Schwartz

07-16-2003, 05:05 PM

When I speak up I am criticized for trying to make it "my contest." When I don't take charge I am accused of "waffling."

I put forth ideas for rules and my rules set has been highly criticized. This is not MY contest.

Let someone come up with a set of rules and I'll decide if I am in or out.

Regards,
Dave Schwartz

Derek2U

07-16-2003, 05:22 PM

I think the software contest idea sounds cool. Why not do it for a
day or two --- as a warmup period then PAUSE & see if more
rules are needed? But the basic idea sounds Very Interesting.

Larry Hamilton

07-16-2003, 05:28 PM

Delman's contest starts next week...I think. In the past the contest was one race a day, every race day at Del Mar. All kinds of races. Why not "borrow" their contest? They do all the work, all PA has to do is post the scores. Hell, they even have a prize if you win.

One race a day, means plenty of time to discuss the race before and after from the various perspectives.

You could probably register as Horse Street Dave..thereby borrowing some advertisting too

Show Me the Wire

07-16-2003, 06:04 PM

Dave and others:

What do you think of my proposition regarding baseline testing for the contest? I think Mania seems okay with it. The only question would be what is the appropriate ratio of contenders to field size? I am sure that can be worked out.

After the contenders are posted everyone will have the opportunity to compare their contenders to the softwares' to see if some specific software will add value to the selection process.

Regards,
Show Me the Wire

Perception ie reality

Dave Schwartz

07-16-2003, 06:23 PM

My opinion is that 2 is the correct number, but 3 or 4 would work well enough.

3 in fields under 8 and 4 if more than that.

Show Me the Wire

07-16-2003, 06:29 PM

Dave:

Thanks for the response.

I vote for 2 contenders in fields of 7 or less and no more than three contenders in fields of 8 or more. This sounds reasonble to me.

Regards,
Show Me the Wire

Perception is reality

lousycapper

07-16-2003, 06:57 PM

:cool: SHOW ME ANY COMMERCIAL PROGRAM THAT HAS THE WINNER IN THE TOP 4-5 @ LEAST 85% OF THE TIME AND I WILL SHOW YOU A MONEY MAKER, REGARDLESS OF WHO USES IT. LET'S GET READY TO RUMBLE... HEH, HEH! :D

-L.H.

Tom

07-16-2003, 07:57 PM

Winner in top 4-5 85% of the time?
Her ya go...note several factors show flat bet profit after 500 races - all races, in a row,no throw outs, just the last 500 races at AP.

Tom

07-16-2003, 07:59 PM

Originally posted by VetScratch
Tom,
What about the free updates and annual renewals during the evaluation period!!! ;) ;)

...that goes wtihout saying....of course we get'em!

lousycapper

07-16-2003, 08:29 PM

:cool: GUESS WHO SAID THAT? YUP, IT'S GRAMPA, AGAIN! :cool:

GR1@HTR

07-16-2003, 08:53 PM

Hello race fans...Looks like an indepth discussion here...I'm sure this has been said before but this matchup wont really prove anything...Anybody just playing top computer picks in every race is DOOOMED...Gotta apply some elbow grease or study to maximize profits...

Good day.

rmania

07-16-2003, 09:48 PM

Sorry folks but I couldn’t wait any longer for people to make up their minds as to what this contest should be.

So today I started to post on the Selections board the out-of-the-box picks I related to in my earlier post.

At first I hesitated because it was a really tough card to tackle. All sprints were maiden events except one and there were a ton of 1st time starters. If fact, of the six sprints on the card I only provided “picks” in four of them. In the other two combined there was only one horse with an official start.

And as expected, the results were a bit below par (only 2 winners in four races). One winner was a single while the other was snagged from the top two. In the two I missed (got places in both) the winner was in the top three.

I realize this doesn’t sound all that impressive but this was accomplished without a single second of handicapping required.

VetScratch

07-16-2003, 10:11 PM

GR1@HTR
Hello race fans...Looks like an indepth discussion here...I'm sure this has been said before but this matchup wont really prove anything...Anybody just playing top computer picks in every race is DOOOMED...Gotta apply some elbow grease or study to maximize profits...

I don't think any users who have a hope of operating profitably expect any commerical package to show blackbox profits playing EVERY race at any track, let alone all races at all tracks.

Realistically, but probably not very feasible for this contest, the way to actually measure baseline performance is to compare several thousand complete software odds-lines with results, as in the Quirin example from the appendix in his first book (Winning...).

The best baseline program will perform closest to the probability-versus-results relationships that exist for track-odds-versus-results relationships (i.e., strong positive correlations). A negative ROI will be inevitable, but it will actually show which baseline handicapping package does the best overall job. What else is handicapping really about other than accurately predicting probabilities?

Dave Schwartz

07-16-2003, 11:26 PM

Okay, here is my idea...

This should be a "showcase" instead of a "contest." That is, it is an opportunity for each vendor to "show off" the performance of his product. The more mechanical the approach he chooses, the greater the liklihood that the public will react positively.

Each vendor creates a post in which his playing approach is defined. It should be such that a "reasonably" experienced user of the software could understand and (*within reason) duplicate.

The approach should be well-described so that there are NO gray areas; no areas of user interpretation.

If the vendor chooses to skip races, the rules for skipping should be stated ahead of time. (i.e. No maiden races, no turf, only SoCal Sprints, whatever.)

If the vendor added his own input to a race it should be marked with a symbol to indicate that personal decision was involved. This way the decision-makers are not left out in the cold. Again, I would think that the public will react more positively to the systems with fewer "no decision" races. Also, the public will probably be smart enough to notice if the big payoffs were mostly the result of "personal expertise."

The vendor can play whatever tracks he chooses, as many races per day that he chooses, however he wishes to bet, but according to the rules he posted.

Scores should be tallied showing the overall win percentage of races, bets, profit, etc. and should be broken out by:

Overall
No Personal Decision
With Personal Decision

Feedback?

Dave Schwartz

--------------------------
* - "within reason" because it is conceivably possible that the results of the output MIGHT be the result of the database of races on the vendor's system and this database might be different than any particular user's system.

VetScratch

07-17-2003, 12:10 AM

Overall
No Personal Decision
With Personal Decision

Are these the menu selections for the waffles or for the syrups?

Simple "verification of odds-line probability accuracy" needs no special sauces or complicated cooking instructions.

Dave Schwartz

07-17-2003, 12:23 AM

VetScratch,

So, if you have a 2/5 shot that you don't like and 7 overlays you bet them all? No rules needed? Just odds lines?

VetScratch

07-17-2003, 12:39 AM

Dave,

Handicapping is nothing more than a probabilities exercise.

After many handicapped odds-lines are analyzed, your best baseline odds should match up with results. Your 2/1 shots and your 20/1 shots should both win in proportion to their underlying probabilities. You must not have finished Quirin's book.

We really don't need 2001 ways to obscure the accuracy of baseline probabilities.
http://koti.mbnet.fi/badbee/wavs/stress.wav
:) :)

Dave Schwartz

07-17-2003, 12:45 AM

VS,

If it were only that easy.

Lefty

07-17-2003, 12:52 AM

vetscratch, i've seen a lot of computer generated oddslines and to think any of them are gonna win in accordance with their underlying probabilities while some of them are overlays is pure fantasy.

Handle

07-17-2003, 12:58 AM

Originally posted by rmania

And as expected, the results were a bit below par (only 2 winners in four races).

Wow, so 50% winners is "below par" for your program. Pretty sweet. This is all that EquiSim's Simulation can manage to do for good old holly park "out of the box". Sprints only. Goes back to 2002.

Version 3.5, Sprints only:
Dirt: 333 races, 101 wins (30.33%), - 0.69% ROI
Turf: 110 races, 24 wins (21.81%), - 1.5% ROI
Overall: 443 Races, 125 wins (28.2%) - 0.9% ROI

Version 4.0 (beta version), Sprints only
Dirt: 333 races, 31.53% wins, + 4.44% ROI
Turf: 111 races, 27.93% wins, + 12.79% ROI
Overall: 444 Races, (30.63%) +6.53 ROI

-Nathan

rmania

07-17-2003, 01:18 AM

Nathan,

Looks like you've made some great strides with your Version 4.

Keep up the good work. :)

Secretariat

07-17-2003, 01:26 AM

quote:

Hello race fans...Looks like an in depth discussion here...I'm sure this has been said before but this matchup wont really prove anything...Anybody just playing top computer picks in every race is DOOOMED...Gotta apply some elbow grease or study to maximize profits...
[B]

Congrats to Dave on the approach he is now advocating. I think it brings some flexibility to the process. I'm very close to coming on board and will be curious on the final showcase rules posted.

I do have a comment regarding above. (1) .....this won't prove anything? ... my question is then, why would any developer want to subject themselves to this...my answer, because it should be something enjoyable to the developer...I tested a rating this afternoon on the entire CRC 2001 season, the method showed a substantial profit over 5 mothns of that season ,and fell apart in the last few months. In a one month contest, this rating might have wiped the board. Every developer knows that, and that is why I kike Dave's idea of a showcase where it shows off the way the program works, and gives various ways the program will work, and discusses its features. Obviously, there is the risk of encountering a bad month or a streak month, but I can accept that. Just hope readers on the board keep that in mind. PaceAdvantage talked about advertising in one thread. Here is a way to advertise, and yet still allow posters to get information about the products.

Since most of our programs offer a variety of ways to personalize your own handicapping, I would like to see a few polls structured first on what users would most like to see emphasized.

Selection Process and/or Features
1. Oddsline approach connected to recommended odds and value
2. Black Box approach, playing top pick to win
3. Contenders Selection, narrowing a field down to contenders
4. Passing Races Parameters

Results Process
1. Win Percentage
2. ROI (Money Won/Money Bet)
3. Win Percentage times ROI
4. Percentage of Plays to Total Races on Cards

Types of Bets
1. Win Bets
2. Place or Show Bets
3. Exotics

VetScratch

07-17-2003, 01:28 AM

Lefty,
vetscratch, i've seen a lot of computer generated oddslines and to think any of them are gonna win in accordance with their underlying probabilities while some of them are overlays is pure fantasy.

If they don't average out that way, you have no basis for finding overlays when you compare software odds to track odds.

Track odds established by the betting public very closely line up with underlying probabilities (when many races are considered), so you need an equivalent degree of overall accuracy to detect single-race aberrations (overlays/underlays) on a consistent basis.

When software users change method options, select/weight variables, or override automatic paceline selections they are doing nothing more than trying to improve the accuracy of the software's internal probability calculations. If a program does nothing more than show and rearrange variables, you have a database viewer, and not a "handicapping" program.

Wagers are nothing more than probability propositions selected in accordance with each player's disagreement with the toteboard.

Programs that come closest to this criterion provide the best baseline starting point for tweaking by the expert players who become the 100k winners.

If this isn't true, haven't we come full-circle back to SMTW's query about whether there is any intrinsic baseline value to any of the programs?

Topcat

07-17-2003, 02:25 AM

Dave's compromise approach of a "showcase" makes a lot of sense. Of course i thought his original idea was fine.

GameTheory

07-17-2003, 04:56 AM

Originally posted by VetScratch

If they don't average out that way, you have no basis for finding overlays when you compare software odds to track odds.

Track odds established by the betting public very closely line up with underlying probabilities (when many races are considered), so you need an equivalent degree of overall accuracy to detect single-race aberrations (overlays/underlays) on a consistent basis.

Hmmm... I've posted a lot about this in my debates with formula_2002, so I won't go into heavy mathematical detail again, but consider this simplified scenario:

We have a card with 9 races.

We have three independent probability generators: A, B, & C. So for each race, we'll have an A line, a B line, & a C line.

Before we analyze what happens in our example, an aside:

I submit that the only true way to measure the accuracy of an odds (probability) line is by looking at the winner only. (No consideration will be given to "sorting" the field correctly by comparing the rank of the probabilities with the actual order of finish.) After all, the line is giving the supposed probability of winning the race, not of coming in 2nd or anywhere else. Therefore, to judge our line we'll be looking at the probability assigned to the actual winner, and that's all.

Back to our example:

Race #1:
Probability of winner on line A: 30%
Probability of winner on line B: 20%
Probability of winner on line C: 10%

Race #2:
Probability of winner on line A: 30%
Probability of winner on line B: 20%
Probability of winner on line C: 10%

Race #3:
Probability of winner on line A: 30%
Probability of winner on line B: 20%
Probability of winner on line C: 10%

Race #4:
Probability of winner on line A: 20%
Probability of winner on line B: 10%
Probability of winner on line C: 30%

Race #5:
Probability of winner on line A: 20%
Probability of winner on line B: 10%
Probability of winner on line C: 30%

Race #6:
Probability of winner on line A: 20%
Probability of winner on line B: 10%
Probability of winner on line C: 30%

Race #7:
Probability of winner on line A: 10%
Probability of winner on line B: 30%
Probability of winner on line C: 20%

Race #8:
Probability of winner on line A: 10%
Probability of winner on line B: 30%
Probability of winner on line C: 20%

Race #9:
Probability of winner on line A: 10%
Probability of winner on line B: 30%
Probability of winner on line C: 20%

Well, even without going into the details of how we best measure the accuracy of each line (which I've posted before), we can plainly see that each of these lines has exactly the same accuracy even though they "picked" (to varying degrees) different horses in each race. (Extrapolate this example and pretend this parity between lines holds up for 100s or 1000s of races.)

So, it is just a simple matter of betting on the overlays, right?

Sorry. You can't determine the ultimate value of a line-generator until you compare it to the public's prices. It's relationship to the public line is quite independent of its "accuracy". Let's say the mutuels went like this:

Race #1: $10.00
Race #2: $10.00
Race #3: $10.00
Race #4: $5.60
Race #5: $5.60
Race #6: $5.60
Race #7: $4.00
Race #8: $4.00
Race #9: $4.00

Which line would you rather be using to find your overlays?

What people don't seem to realize is that for a given set of races, and a given amount of "accuracy", there are an infinite number of odds lines that can achieve that accuracy. More precisely, an infinite amount of sets of oddslines (each set presumably generated by a single method) -- each set having one oddsline for each race in the set.

So how the probabilities average out is only part of the story -- you also want to maximize the discrepancy from the public odds because that is where the value comes from -- not from accuracy. You can easily imagine in the above example where line A was somewhat less accurate than the other two but was still the most derisable line to work with because it generated the most profit when betting the overlays.

VetScratch

07-17-2003, 06:51 AM

Game Theory,

Finally, a worthy adversary! (I couldn't resist that "aside" before putting on my game face to confront you.)

Let's step through this issue.

Game Theory:
I submit that the only true way to measure the accuracy of an odds (probability) line is by looking at the winner only. (No consideration will be given to "sorting" the field correctly by comparing the rank of the probabilities with the actual order of finish.) After all, the line is giving the supposed probability of winning the race, not of coming in 2nd or anywhere else. Therefore, to judge our line we'll be looking at the probability assigned to the actual winner, and that's all.

Reply:
Agreed. All we need to know is whether "generated 30% probabilities" are winning very close to 30% of the time, and likewise for 20%, 10% and other generated probabilities. Are they winning at the predicted rates when many races are considered, that is the question.

Game Theory & Reply:
Generator probability examples accepted as posted.

Game Theory:
Well, even without going into the details of how we best measure the accuracy of each line (which I've posted before), we can plainly see that each of these lines has exactly the same accuracy even though they "picked" (to varying degrees) different horses in each race. (Extrapolate this example and pretend this parity between lines holds up for 100s or 1000s of races.)

Reply:
Ok. All were equally accurate, but none of them were very accurate. I will concede that three equally adept odds generators could be found, and given enough races, their 10%, 20%, and 30% predictions might statistically vary from actual win percentages by the same margins. Let's say for argument that their 30% horses won 33%, 20% horses won 33%, and 10% horses won 33%, which is what the example shows. I would certainly agree they are "equally" accurate/inaccurate. However, I would keep searching with the goal of finding a generator where 30% horses won 30%, 20% horses won 20%, and 10% horses won 10%, which is desired objective when results are compiled for "100s or 1000s of races."

Game Theory & Reply:
Payoff examples accepted with the provision we scramble them daily to avoid the onus of backfitting.

GameTheory

07-17-2003, 11:42 AM

Well, it sounds like we agree then.

Basic point being is that you need to create a "value-added" line that is as accurate as possible while being as different from the public's as possible. Value doesn't come from accuracy -- it comes from discrepancy. It is very very tough to equal or better the accuracy of the public line (while not considering it in your own), and you generally will find that most increases in accuracy of your own line are simply moving you closer to the public's line and not really a helpful gain. That's because most people are basing their line one way or another on the same data that everyone else is looking at (or manipulations of said data).

So, in general, you will often need to sacrifice some accuracy to get more value, and that will often be the right thing to do.

We've also discussed in the past here how to create a hybrid line (combining your line & the public's to make a new line). If in that process, your line more or less disappears, you'll know your original line (made independently of the public's) isn't very useful. Depending on the character of your method, your line might be best used after hybridization, or it may be more valuable on its own.

Of course now I've brought into the discussion this business of making your line totally independently from the public's, but I'm just trying to show that equal gains in accuracy are not equal gains in value. If striving for the most accurate line possible without price considerations, you'll probably just end up with a mirror of the public, which is of course totally useless.

Show Me the Wire

07-17-2003, 11:57 AM

VS and GT:

All very interesting and esoteric. The bottom line is if the selctor does not have the eventual winner in his/her/its contenders there is no value. KISS, accuracy is important becuase without it you ain't got no value.

Regards,
Show Me the Wire

Perception is reality

VetScratch

07-17-2003, 01:10 PM

Game Theory,

I generally agree, but must point out that projected probabilities that fail to approach the accuracy of probabilities represented by public odds will have diminished value as an appraisal yardstick for value-based decisions in any single race.

If 5/1 projections are winning at the same rate as 5/1 horses, then a good handicapper will evaluate a horse projected at 5/1 that appears poised to go off at 9/1 (i.e., on the toteboard), and likewise will be apprehensive about fair value when a horse projected at 5/1 seems poised to go off at 2/1.

Your caution about mirrored probabilities in race after race should be mitigated in numerous instances where an imperfect market is created by a few extreme wagers, as when 20% of the players account for 80% of the handle, which happens more frequently at minor/middle ranked ovals than major tracks. However, imperfect markets (of meaningful significance) by statistical standards do occur at all levels.

lousycapper

07-17-2003, 01:21 PM

"Back-fitting and all the theories in the world aren't worth squat... the only thing that counts in this business is consistantly making money when bettin' on the hosses. Take that ya data-based geeks"

-L.C.

rmania

07-17-2003, 07:55 PM

This represents the 145th post to a thread which, when started, seemed as if it was going to pave the way to a software showdown contest. A contest not just for bragging rights, but to provide each developer a chance to demonstrate the validity and value of their program.

Sadly, with all of the posturing, double-talk, backtracking and bickering displayed, the idea is quickly becoming nothing more than a passing thought.

What’s worse is that what appeared to be interest from the rest of the board has turned into ridicule.

As for me, I’ve started my own contest and after two days of posting I’m holding my own. It would be nice though to have a little competition.

And since it’s my contest, I get to make the rules (or should I say rule). The rule is NO HANDICAPPING ALLOWED.

It anyone is interested in the challenge, you can post to the thread I started on the Selections board “HollyPark Out of the Box”. And no, you don’t have to play Hollywood Park.

Handle

07-18-2003, 12:46 AM

One thing though -- if there's no "handicapping" then what's the diffference between posting before the races and after them? Is it like a serialized novel printed in the newspaper where the authors know what's coming next, but its drawn out for the suspense? I'm not particularly down on it, I'm just pointing out that if its a mechanical system then it makes no difference whether the results come before or after the race, other than the psychological effect. The results would be the same irregardless of when you look at them.

Lefty

07-18-2003, 01:12 AM

Vets... No way a computer generated oddsline is gonna win at its probabilities. 2-1 the prob, is 33% now some of these 2-1 software lines gonna go at 3-1, 4-1, some at 10-1 and above. And you think they should win 33%? No way. But again, they don't have to. Say your 2-1 overlays avg 5-1 then if you can cash these 20% you're gonna be a winner.
If you ever find a software line that wins to its prob. i'll buy it and throw all the rest of my software away.

rmania

07-18-2003, 01:13 AM

Originally posted by Handle
One thing though -- if there's no "handicapping" then what's the diffference between posting before the races and after them? Is it like a serialized novel printed in the newspaper where the authors know what's coming next, but its drawn out for the suspense? I'm not particularly down on it, I'm just pointing out that if its a mechanical system then it makes no difference whether the results come before or after the race, other than the psychological effect. The results would be the same irregardless of when you look at them.
Nathan you come up with some of the darnest stuff. And of course your absolutley right. But for some reason it don't think that posting selections after the race would fly. This is a pretty tough group.;)

VetScratch

07-18-2003, 02:21 AM

Lefty,
Vets... No way a computer generated oddsline is gonna win at its probabilities. 2-1 the prob, is 33% now some of these 2-1 software lines gonna go at 3-1, 4-1, some at 10-1 and above. And you think they should win 33%? No way. But again, they don't have to. Say your 2-1 overlays avg 5-1 then if you can cash these 20% you're gonna be a winner.
If you ever find a software line that wins to its prob. i'll buy it and throw all the rest of my software away.
I agree with your observations but not with your conclusion.

Most of the programs that I have seen don't even make a serious attempt to accurately project probabilities (i.e., via odds-lines). However, this is not because it cannot be done.

Because so few true value players exist, most of the programs are content to put out odds-lines that cannot even be compared to the toteboard. How many times do we see 8/5, 3/1, 9/2, 10/1, 20/1, 20/1, 50/1, 99/1, 99/1, 99/1 on a toteboard?

My friend, who has been my mentor for applying software skills to handicapping, has surveyed a lot of programs, and says MOST odd-lines don't even predict one winner per race. Sometimes less than one winner is predicted; at other times more than one winner is predicted.

Many of the documented program examples, marketing brochures, and screen shots at vendor web sites actually SHOW bogus odds-lines of this sort! This is no better than Morning Lines, which are also bogus.

Since the "dumb" public always predicts one winner per race, some the the "brains" in the handicapping industry should be able to do likewise!

GameTheory

07-18-2003, 03:58 AM

Originally posted by VetScratch
My friend, who has been my mentor for applying software skills to handicapping, has surveyed a lot of programs, and says MOST odd-lines don't even predict one winner per race. Sometimes less than one winner is predicted; at other times more than one winner is predicted.

Can you clarify this? I don't get it. An oddsline is just an oddsline -- it estimates the probability of each horse in the race to win. Do you just mean the probabilities don't equal 100% as they should?

VetScratch

07-18-2003, 06:04 AM

Hi Game,
Do you just mean the probabilities don't equal 100% as they should?
Yes, how can you trust an odds-line that adds up to 70% or 130%, or even miss by more, as often happens in big-field races like recent Kentucky Derbies?

BTW, I discussed our mirrored odd-lines question with my friend, who is more into such stuff than I am. What I said previously needs some clarification. According to toteboard-industry sources who have occasionally volunteered data for academic studies, a very high percentage of win pool handle typically comes from a small percentage of wagers (not "players", as I stated). I guess it is not feasible to produce good player-percentage estimates unless you make sweeping assumptions about successive wagers placed through the same windows (or other points of origin).

So, when 80% of the handle comes from 20% of the wagers, how that relates to players is unclear. What it does indicate is a pari-mutuel market characterized by high volatility stemming from big wagers, and that argues against the chances of mirrored odds-lines race after race. In other words, when thousands of races are considered, a program "can" match the probability accuracy reflected by pari-mutuel studies, but it is unlikely to be a mirror of public of odds-lines because of the different way the pari-mutuel market functions.

In light of this, I would have to think over some of what you said in the tail end of your comprehensive post on this subject. I have a hard time accepting that maximizing discrepancy (i.e., diminished accuracy) is the way to go. Are you basing this on new studies that refute widely accepted pari-mutuel studies?

hurrikane

07-18-2003, 06:31 AM

vet,

what studies? where are these studies and who did them?

GameTheory

07-18-2003, 06:46 AM

Originally posted by VetScratch
In light of this, I would have to think over some of what you said in the tail end of your comprehensive post on this subject. I have a hard time accepting that maximizing discrepancy (i.e., diminished accuracy) is the way to go. Are you basing this on new studies that refute widely accepted pari-mutuel studies?

My point was mainly theoretical, but what I said was to maximize discrepancy and accuracy. It is a dual and (possibly) competing goal. If you only consider accuracy, you are going to be headed for that mirror most of the time, and your moments of discrepancy will be less useful than if you explicitly considered it in your method. But you certainly don't just want to maximize difference either without regard to accuracy.

Maybe your friend is familiar with the concept known as "Pareto optimal", which provides a framework for balancing goals that are competing. The goals in this case aren't always competing, but the more you rely on the same information as everyone else they will naturally tend to conflict.

VetScratch

07-18-2003, 07:43 AM

On the mirroring question, I see (or maybe just sense) a reason to stick with my contention that zealous pursuit of accuracy won't be rewarded by useless repetitive mirroring.

A program that calculates probabilities stop there. It does not suggest what to wager or how much to wager. Accurate probabilities are the only goal.

Meanwhile, the betting public has historically demonstrated remarkable long-term accuracy with respect to underlying probabilities, but the betting public (like the stock market) follows a volatile path. If all players were limited to one $2 wager, this volatility would decline.

Many players vote in relation to their bankrolls, and their convictions are not weighted equally. Whether most players bet $10 or $100 has more to do with bankroll size than handicapping, and this is the source of volatility.

A program-generated odds-line is simply a tool for analyzing a volatile market. Is the horse that I like trading over or under fair market value (i.e., should I buy or sell?). Isn't that the essential that issue must be quantified in order to successfully play pari-mutuels? (i.e., because of takeout).

When five-thousand 8/5 horses no longer win very close to 31% of the time (subject to adjustment for takeout variations), I will be more open to other ideas.

BTW, I will ask about your last question to see if my friend knows anything about it.

Are you sure Pareto is not a condiment or a sauce (joke)?:) :) :)

VetScratch

07-18-2003, 08:40 AM

Hurrikane,
what studies? where are these studies and who did them?
I think some old-timers preceded Quirin (1979), but his book popularized this kind of study.

Most every region/circuit used to have local a expert/publisher who kept cumulative statistics and published annual updates showing results (often showing cumulative and previous season).

These guys began to disappear (like the local variant sheets) when the information business became a national enterprise dominated by big on-line providers.

I haven't seen one of these booklets for several years, but my OCD/statistics friend has evaluated his own database without revealing a drastic change to the assertion that horses tend to win in accordance with the probabilities expressed by their odds.

If you compare pari-mutuel probabilities to win percentages, you should produce a ladder that reflects substantial agreement. (See page 296 of Quirin to see such a ladder).

This is why I asked Game Theory if he had new studies that no longer support this thesis.

hurrikane

07-18-2003, 10:16 AM

vet,
I believe and know about these studies. When someone says my friend did this and that study I just like to know what where and when.

I'm guessing your friend is Joe. Our own formula. :D

VetScratch

07-18-2003, 11:02 AM

Hurrikane,
I'm guessing your friend is Joe. Our own formula.
No, my friend is someone I grew up with from grade school days. When our parents made us go off to college instead of ending up on the backside, we lost track of one another. Years later, we accidently got reuinited working on software in Cleveland (the Avis Wizard, not the BRIS Wizard:):) ).

Where can I find Joe's study and compare his formula to Quirin's original formula? I am clueless about Joe unless you mean Takech out in California. We can't get his stuff from BRIS or at newsstands in the Midwest and South.

If you are with Takech, I would advise folks to pay keen attention to the only thing I know about him. He is dead right by saying you need privileged information to really beat the horses. This is what we combine with our handicapping. In other words, start out with a horse for which you know something positive that is very privileged, then use computer handicapping to decide whether you are likely to have a "good thing" when you get the good effort that you have reason to expect. Unless you have a privileged way to predict whether a good effort is highly probable, handicapping the history of horses falls short of really predicting how a horse will run today.

rmania

07-18-2003, 11:30 AM

Originally posted by VetScratch
No, my friend is someone I grew up with from grade school days. When our parents made us go off to college instead of ending up on the backside, we lost track of one another.
So that it explains it. You went to college and got yourself a phD. And now you spend all of our time on message boards p_iling it as h_igh and as D_eep as possible. Tell me, do your parents know that this is how you're using the education they paid for.:)

VetScratch

07-18-2003, 01:36 PM

Rmania,
So that it explains it. You went to college and got yourself a phD. And now you spend all of our time on message boards p_iling it as h_igh and as D_eep as possible. Tell me, do your parents know that this is how you're using the education they paid for.
My parents were only intent upon forcing me to see other options outside the racing subculture. Same for my friend. As a result, we can live with one foot in and one foot out. Most of the other kids jumped in with both feet at age 16 or 17, with both wonderful and tragic consequences.

So what I did with a BA was never important to my folks. They just wanted me to see that the outside world offered opportunities. When they are naturally gifted, many young people simply grab the first ring because success comes so easy for them.

What's your story?
:) :) :)

rmania

07-20-2003, 12:56 AM

Originally posted by rmania

As for me, I’ve started my own contest and after two days of posting I’m holding my own. It would be nice though to have a little competition.

And since it’s my contest, I get to make the rules (or should I say rule). The rule is NO HANDICAPPING ALLOWED.
Well, tomorrow is closing day at Hollywood Park and it will also be the last day of my contest.

I don't know what will happen tommorow, but since I'm the only contestant, I think I've got a pretty good lock at winning. And with a 53% win rate and a +72% ROI going in, I'm pretty much guaranteed that I'll at least break even.:D

Remember Gino...Never say never.;)

VetScratch

07-20-2003, 01:31 AM

Based on the success of your HOL posts, a good final day should enable you to wrap up a fine year and spend the remaining months roaming the world's playgrounds for the rich and famous!

If this is your first taste of excess, I'm sure Gino would be glad to help you shop for the appropriate wardrobe, accessories, and toys for Livin' Large!
:) :) :) :)

Bobby

07-20-2003, 02:12 AM

Vetscratch,

are you a man or a woman? it matters to me

VetScratch

07-20-2003, 02:20 AM

ROFL -:D :D :D :D

Bobby (as opposed to Bobbie),

Did you mean to say, "You show me your's, and then I'll show you mine!

Sorry pal, but I've already got two ex-husbands and Gino to put up with, so take a number and get back in line!

:D :D :D :D :D

Bobby

07-20-2003, 02:49 AM

Thanks for your response.

To be honest, it really doesn't matter to me what sex you are.
I think your a smart ass. You need to bridle your ego. Check this link out
http://www.spiritualityofsuccess.com/SiteData/Docs/LEAVING%20EGO%20OUT%20OF%20LEADERSHIP%20reprint.do c

VetScratch

07-20-2003, 03:06 AM

Bobby,

Did you think I was smarting-off to Rmania? Nothing could be further from the truth. My congratulations and wishes for good luck with the final card at HOL were SINCERE as driven snow!

At the same time, do I envy Rmania's fantastic success? Sure, don't we all, based on how he can "party" while the rest of us have to work through December?

While Rmania is Livin' Large, I'll probably have to grind away at the pari-mutuels and continue to churn out lame material for Michael Surferman, who bills himself at the "Great White Hope On The Def Jam Comedy Circuit".

BTW, members of this board are "too smart" to fall for a DOC link.
Instead of playing amateur tricks with macros and scripts, you had better cozy up to a real hacker! Don't come back until you can ambush me with a PDF document!
:eek: :eek: :eek: :eek:

Amazin

07-20-2003, 12:30 PM

Posted by Handle:

Sprints only. Goes back to 2002.

Version 3.5, Sprints only:
Dirt: 333 races, 101 wins (30.33%), - 0.69% ROI
Turf: 110 races, 24 wins (21.81%), - 1.5% ROI
Overall: 443 Races, 125 wins (28.2%) - 0.9% ROI

Version 4.0 (beta version), Sprints only
Dirt: 333 races, 31.53% wins, + 4.44% ROI
Turf: 111 races, 27.93% wins, + 12.79% ROI
Overall: 444 Races, (30.63%) +6.53 ROI

-Nathan

One year ago,I started a thread about what was the ROI of your software.Just like in the off topic section,it was me v.about 20 pa members including Handle.I kept getting ridiculed for trying to get a statement of roi of any software.Looking at those figures Nathan posted,I now know why.Just as I had suspected.Nice to see your honest Nathan,eventually.

Vet scratch:

Since Dave alluded to his temporary retirement as a result of this thread,I have read most of it and conclude that that Waffle and Syrup remark was not taken well. It's just my observation but I'm not criticiizing you for it.That would be a compliment for me considering what I get here.

Rmania:

I like your confidence in your program(no waffles).I do have a black box homemade program that allmost prompted me to enter this thing because it does really well,but it's still in infancy.I am in the process of a web site that will have picks from my black box,archive them and have a running total of ROI.Right now it's ROI is hovering too high for believability.

Suff

07-20-2003, 12:51 PM

[QUOTE]Originally posted by Handle
Wow, so 50% winners is "below par" for your program. Pretty sweet. This is all that EquiSim's Simulation can manage to do for good old holly park "out of the box". Sprints only. Goes back to 2002.

This thread still about a contest? I was in the selections forum telling everyone why they shoulda bet an 85-1 shot that finsihed 5th. meanwhile the equisim program had the 155 Ex and 430 tri with top three piks. But what impressed me more about that Box (i;m not even sure what it is) is that the 12 was 1-1 and the box had not only had him "not " winning , it properly placed him Right off the whole ticket. And you have to figure 12 horse field he was 1-1... to the naked eye he looked unbeatable. Thats pretty fair work. Isolated incident I know...and all box's do this. Its just I wanted to point it out. because it was pretty impressive.

Handle

07-20-2003, 07:22 PM

Amazin,

You must be confused as I don't know what you are talking about.

EquiSim TRACKS THE ROI FOR ITS SIMULATION WINNER. The new version tracks the simulation winner and a number of other "factors". I've said this how many times??? Its PART OF THE PROGRAM. Ok? Its no wonder why you feel that its "you against everyone else". Quite frankly, you come off as a raging asshole.

-Nathan

Amazin

07-21-2003, 12:22 AM

Nathan:

Just as you are mistaken about me being "a raging asshole",you are mistaken that you did not discuss ROI with me last year and did your best not to give me a straight answer.Interesting isn't it that someone you consider an asshole has a much better memory than you do.So what does that make you?
Here's the thread:

http://www.paceadvantage.com/forum/showthread.php?threadid=2703&perpage=15&pagenumber=1

Handle

07-21-2003, 12:37 AM

Amazin,

Actually, I said that you come off as a raging asshole. And this, without a doubt, is true. I understand that inbreeding can cause mental defects -- but do note that I don't consider this an excuse and, therefore, I have no sympathy for you.

You wrote something last year that was couched in inflamatory remarks. I wrote back to say that I didn't think that ROI was the be all and end all of handicapping software. Now, a year later, you burst upon the scene in typical pustule fashion accusing me of dishonesty, of waffling, of not giving you a straight answer (as if there were a straight question to begin with), when all of the time my software does exactly what you were talking about in the first place -- tells you were the ROI is. Maybe I would have "sided" with you last year, but since I honestly dislike everything I've ever read of yours because you sound so negative, nagging, and just generally red and inflamed, I can't bring myself to do so under any circumstances.

You can have the last word, I'm not going to enage in this further, except to say that I wish you'd go away -- you remind me of a raging hemorrhoid.

lousycapper

07-21-2003, 01:25 AM

:D "This contest is much ado about nothing. Very funny stuff, however. Reminds me of the committee that decided to design a cow and wound up with a camel." Grampa, thanks for that one! :cool:

-L.C.

rmania

07-21-2003, 09:33 AM

Originally posted by lousycapper
:D "This contest is much ado about nothing....
L.C.

I couldn’t agree with you more....

That’s why I made the decision to break away and start my own contest which, after 5 days, concluded yesterday.

I was somewhat disappointed that no other developers joined in, but I wasn’t all that surprised. It probably had something to do with my one and only contest rule “no handicapping allowed”. And though this may have seemed like a silly rule, I see this as the only way to independently judge any program.

I’ve said it before and I’ll say it again... IMO handicapping software should give ANY user a decisive edge just my turning it on.

Well, the results are in and the winner is RMania with a win rate of 52.17% and a positive return on investment (+ROI) of 47.35%. The percentages were based on a $2.00 win wager on each horse posted. And, according to the “rule”, this was accomplished with no handicapping what-so-ever.

Of course, my program does allow for handicapping and I still insist that it is a tool – not a “black box”. So, who knows, maybe by combining the “inherent edge” with one’s handicapping knowledge and skill the percentages could have been much higher.;)

SAL

07-21-2003, 09:57 AM

You know, you said that the selections made with no handicapping, but I disagree.
On the first day of your selections you stated that horses that were on the lead at the half mile mark in HOL sprints showed a decent win percentage and roi. THAT IS HANDICAPPING.
You didn't use the best finishers as your picks, you used the best half mile leaders. Will you do the same when Del Mar opens? Maybe, maybe not. But making the decision to use the the half mile leaders as your selection is handicapping as far as I'm concerned.
Nonetheless, excellent showing.

rmania

07-21-2003, 10:10 AM

Originally posted by SAL
You know, you said that the selections made with no handicapping, but I disagree.
On the first day of your selections you stated that horses that were on the lead at the half mile mark in HOL sprints showed a decent win percentage and roi. THAT IS HANDICAPPING.
You didn't use the best finishers as your picks, you used the best half mile leaders. Will you do the same when Del Mar opens? Maybe, maybe not. But making the decision to use the the half mile leaders as your selection is handicapping as far as I'm concerned.
Nonetheless, excellent showing.
SAL,
No disrespect intended, but you need to re-read what I posted on the first day (and every day after).

I stated that selections were those shown leading after a half-mile in RaceMaker's "VirtualView" screen. I also stated that two horses would be listed in the event of a tie. The truth is that a vast majority of those selected were NOT actually leading after a half-mile in their previous race. In fact, more than a few were as many as 8 lengths back after a half-mile in their previous race.

And I was true to my word regarding "no handicapping". All I was doing was down loading the past performances and going directly to the quarter-pole view and picking the horse(s) shown leading.

SAL

07-21-2003, 10:26 AM

I think you need to reread what I said. I said you selected the horses shown leading at the half mile mark in your program. I never said anything about leading at the half in their last race.
And your decision to use the quarter pole view instead of the finish to make your picks is HANDICAPPING.
Why did you use that point of the race instead of the finish? Why not the 1st quarter? The decision to use the half is handicapping, in my view.

VetScratch

07-21-2003, 10:40 AM

Rmania,
Impressed by your posted results, I looked at the 10 race results for HOL 7/20/03.

You elected to make a pick in only 5 races:
#1 - 5 horses
#2 - 6 horses
#4 - 5 horses
#6 - 5 horses
#8 - 4 horses

You skipped the only sprint with a large field, #10 with 10 horses.

You also skipped the four routes with 6, 7, 8, and 8 horses, respectively.

I don't call this blackbox automatic handicapping!

What kind of a contest do you call this... Shooting Fish In A Barrel?

What happens when you post your selections for all 10 races, or play a racetrack that others might actually enjoy playing instead of the joke card at HOL?

rmania

07-21-2003, 11:25 AM

Dearest VetScratch,

I have to hand it to you.. You are one tough nut to crack.

Since you are a runaway winner with posts to this thread (33 – mostly unsolicited) you, more than anyone else, should have an idea as to my approach and the limitations (or should I say the “specialty nature”) of my software.

But, for those that don’t, I’ll be more than happy to explain my choice of plays yesterday.

First of all my program is designed for sprints only. More specifically, SoCal sprints only (though I working on adding additional circuits).

Second, the number of horses you listed in each race (though accurate) was the result of numerous scratches on the card. In fact, I was “scratched out” of two of the races I posted.

Which leads to the following statement....

With six sprints on the card, I initially opted not to post race 10. Yeah, there were 10 horses entered, yet 7 were first time starters. HOWEVER..... after I was scratched out of the two races I mentioned I went ahead and posted a single selection for race 10. And, as luck would have it (?) that horse won.

So you ask “What happens when you post your selections for all 10 races, or play a racetrack that others might actually enjoy playing instead of the joke card at HOL?”

Well, I’m sorry but it ain’t gonna happen.

Anyway, I thought the premise here was to make a profit...

Suff

07-21-2003, 11:40 AM

Originally posted by rmania
SAL,
view and picking the horse(s) shown leading.

Rmania.

I respect you for laying out your numbers. But agree with me under the circumstances its a Little salt in the wounds of some hurt ego's and feelings coming out of this thread.

You put up your numbers. You wanted the Numbers to do the talking. They did. Let it go. the numbers are there for anyone to see. I had a guy I know and trust tell me he's buying your program because he's struggling at SOCAL and is willing to give you a try. Mission accomplished on BOTH FRONTS.

You numbers are there and people are buying.

For peace and Harmony. End this please.

jackad

07-21-2003, 11:59 AM

Do 23 races prove anything? Why not continue to post for at least another 23 races?
I've been using Thorobrain5 in a particular way for the last 10 days. I show a 37% win frequency (betting one horse to win) and an ROI of 65%. And I still don't trust that the approach will prove profitable over a longer term.
Am I being realistic or overly cautious and pessimistic?
Jack

jackad

07-21-2003, 12:02 PM

I should have included mention that these data cover 73 bets over the 10 day period.
Jack

Amazin

07-21-2003, 01:00 PM

Handle;

You are one BORING dude.At least others who want to put me down have some kind of lame reason.You don't have any reason why I'm an asshole to you.

My analysis of your posts shows you to be an extremely INSECURE person especially regarding your software.I notice that anytime anyone says anything remotely critcal of your beloved EQUISIM.Personally I have downloaded your free version and found that allthough it has it's moments,just like you did in the PA contest,the majority of the time it is such a poor performer it makes the user lose interest..I do think the cartoon deal is the best part of your program.But that's for amusement only.

It's quite clear to me that the reason you dodged giving me a straightfoward answer last year on ROI of your program and do give one this year was because last year it was embarrassing to post a negative ROI as a standolaone figure whereas this year,you have something positive to compliment it.Not so embarrassing or damaging to your product.

I can tell that Rmania has confidence in his program because he is willing to put up rather than shut up.

What is so obvious is that if I have a product that has a positive ROI,I WANT people to see how well it does with ROI because that is a major seller of the product.When someone WAFFLES about it,the red flags are a very good indication that the product is shaky at best.Care for some syrup?

lousycapper

07-21-2003, 01:26 PM

Originally posted by rmania
L.C.

I couldn’t agree with you more....

That’s why I made the decision to break away and start my own contest which, after 5 days, concluded yesterday.

=============================

Mr. RMania,

Grampa thanks you!

-L.C.

lousycapper

07-21-2003, 01:32 PM

Originally posted by Amazin

I can tell that Rmania has confidence in his program because he is willing to put up rather than shut up.

What is so obvious is that if I have a product that has a positive ROI,I WANT people to see how well it does with ROI because that is a major seller of the product.When someone WAFFLES about it,the red flags are a very good indication that the product is shaky at best.Care for some syrup?

=============================

"Never, never forget that this business is all about Return On Investment. Only DUMMIES don't care and they eventually go broke."

-L.C.

jackad

07-21-2003, 02:26 PM

But I repeat: Do 5 days and 23 bets prove anything?
Jack

Show Me the Wire

07-21-2003, 03:07 PM

Okay here is my take on this commercial software and hurt feelings. I started a thread regarding paceline selection citing Brohamer theories and specifically asked how is this applied or not applied while using software. There were over 50 responses and not one seller of software voluntarily posted a response or attempted to answer a valid question. Nathan responded but his response was to a direct question about the value of his software graphically displaying a race.

I do not know maybe my question was perceived as an attack on the commercial software sellers' products (notice plural possessive use). However, it is a threshold question raised by an authority on the subject.

I was disappointed, but not surprised that none of the commercial sellers wanted to address the pros and cons of mechanical paceline selection, i.e. last race, etc versus contender selection prior to selecting today's representative paceline.

To sum it up I am tired of seeing substantive discussions ignored, while hearing claims about rois and such. The bottom line is roi is important in any business, but the business plan must be feasible.

The only provider on this thread that has defined the feasibility of his product has been Mania. If I purchase his software, I understand the concept and what he believes is important to select contenders for a specific circuit and distance race. If his selection method is better than average the roi will reflect the user's ability to discen when value is present or not for the size of his wager. Yes ladies and gentlemen roi does not equate to income. I can have an roi of 25%, however that does not mean I will have a substantial increase in my income. A positive roi only means I did not lose money.

As far as it concerns me I do not want to hear about rois, but the concepts and I do not mean your secret recipe.

Regards,
Show Me the Wire

Perception is reality

Show Me the Wire

07-21-2003, 03:16 PM

Originally posted by jackad
But I repeat: Do 5 days and 23 bets prove anything?
Jack

Yes, it proves what he did in a five day period using his selection method and that is all anyone can prove about any method. Past success does not insure future success.

What is the point in asking proof of something that cannot be proved?

Regards,
Show Me the Wire

Perception is reality

Lefty

07-21-2003, 04:00 PM

Sal, I have to agree with Rmania about him not handicapping. He made a hard and fast rule that is used for every race he plays.
In my mind, handicapping is when you use several factors and juggle them around somewhat differently depending on the situation, the odds, and sometimes just the mood of the moment/

VetScratch

07-21-2003, 09:01 PM

Rmania,
I missed the point that you were only playing sprints in your contest. Only looked at your posts for the last day and the card at HOL, which did end on a miserable note. Why did they bother to race?

For what you set out to do, you did well.

SAL

07-21-2003, 09:57 PM

Sal, I have to agree with Rmania about him not handicapping. He made a hard and fast rule that is used for every race he plays.

Lefty, I see your point about using the same rules for every race. BUT- he made the decision to use the leaders from the half mile mark as his selections.

Why did he do that?

He did it because he saw that horses who were leading at the half mile mark did very well at Hollywood park. Coming to that conclusion IS handicapping. It would be no different if I decided to use the best closers at Belmont, or the best 1st quarter mile horses at Emerald Downs.

Will he use the same call to make his selections for Del Mar? He may or may not, depending on how the track is playing. But whatever decision he makes will be a HANDICAPPING decision.

Lefty

07-21-2003, 10:45 PM

Sal, under your guidelines then every decision is handicapping decision because rules have to emanate from a source and not just the imagination. Guess we'll just disagree a little here but that's what makes a hoss race.

Tom

07-21-2003, 10:54 PM

But, whould a new buyer know to use the second call as the descison maker usin gthe program right out of the box?

rmania

07-21-2003, 11:35 PM

Originally posted by rmania
...And, like a few other programs I’ve seen, mine doesn’t spit out winners. My “VirtualView” handicapping feature breaks a race down into 3 segments (half-mile pole, quarter pole, and finish). As mentioned in the tutorial, positions at the quarter pole are the strongest indicators. And, true to form, horses shown leading at the quarter pole have been winning at a rate of over 50% for the current Hollywood meet. Most have been sent off as favorites or second choice, but a few nice hits have kept the ROI above +25%.
OK guys.... This quote of mine is from the 5th post to this thread. Not only does it describe the approach I planned to use, but it revealed that this approach is also documented in the software tutorial.

SAL

07-22-2003, 12:21 AM

OK RMania, this will be my last reply on this thread.

If you continue to post selections using the half mile call at Del Mar and go 0 for 15, but see that horses in the final call are winning 40%, are you going to continue posting half mile call leaders? I don't think so. THAT is what I'm talking about. Handicapping, user intervention, your gut feeling, whatever you want to call it. It's not "out of the box".

Granted, I know you didn't design it to be that way, you've made that clear. But this may be why a handicapping contest with ironclad rules for software makers may not work. Just my .02

Later

rmania

07-22-2003, 12:35 AM

SAL....

The quote from my last post contained the sentence "As mentioned in the tutorial, positions at the quarter pole are the strongest indicators."

The tutorial is written into the software. This theory (if you choose to call it that) could be considered the cornerstone of the entire program. The tutorial goes on the say that horses shown to be in contention at the quarter should rarley be disregarded.

And yes, I will continue with this approach at Del Mar.

Amazin

07-22-2003, 01:45 PM

Rmania:

What paceline selection method do you use?In my own software program,I actually have to manually look at the race I want the program to use so I guess with Sal's rules my program wouldn't qualify either as a blackbox because I am partially handicapping (which is true).

So are your pacelines that your program uses automatically selected or user selected?

Also does your program use a single pace line or an average of multiple pace lines?

rmania

07-23-2003, 09:31 AM

Originally posted by Amazin
Rmania:

What paceline selection method do you use?In my own software program,I actually have to manually look at the race I want the program to use so I guess with Sal's rules my program wouldn't qualify either as a blackbox because I am partially handicapping (which is true).

So are your pacelines that your program uses automatically selected or user selected?

Also does your program use a single pace line or an average of multiple pace lines?
Initial record creation is performed using each horses most current race/paceline.

Paceline substitution is an option with up to 10 pacelines to choose from. Personally, I wouldn't recommend going any further than 2-back unless that race was a "troubled" race.

Another option is to modify the final quarter time only thus preserving the quarter and half-mile fractions of the selected (or default) paceline. I find this method to be quite valuable since most horses (at least in sprints) will come close to duplicating their previous effort for the first half-mile. It's usually that final quarter where the difference is seen.

As for you final question, my program uses a single paceline only. IMO, averaging multiple pacelines produces useless information.

jk3521

07-23-2003, 11:50 AM

rmania,

Until your program uses BRIS or TSN data and can handicap other tracks besides So.Cal, I do not see much of a market for your software. Who needs So.Cal racing with 5 horse fields. And of course with small fields any program would show a good winning percentage. What is your winning percentage with fields of 8 or more? How many people use ITS data? Not me or anyone else I know . And lay off of Handle!

Amazin

07-23-2003, 12:17 PM

Rmania

I also prefer the single paceline method to represent the horse's ability.
In picking a paceline I do have the option of having the program pick out the paceline.In automatic mode it picks the paceline that represents today's #1surface and #2 distance or as close as possible.However the results are not as good as me picking the paceline manually,but it is a great time saver.
You say you don't recommend going back more than 2 back unless there was "trouble".Let's say today's race is a dirt sprint and and the last dirt sprint a horse in question had is 5 back and at a different track.The last 4 were dirt or turf routes at present track.It sounds like you would be using a turf or dirt route to represent the horse's ability in a dirt sprint.Personally I would find that inaccurate and would use the 5 back sprint. Would you agree?

rmania

07-23-2003, 12:37 PM

Originally posted by Amazin
You say you don't recommend going back more than 2 back unless there was "trouble". Let's say today's race is a dirt sprint and and the last dirt sprint a horse in question had is 5 back and at a different track.The last 4 were dirt or turf routes at present track.It sounds like you would be using a turf or dirt route to represent the horse's ability in a dirt sprint. Personally I would find that inaccurate and would use the 5 back sprint. Would you agree?
Of course I would agree. But I should clarify my statement. When I said no more than 2-back I was thinking in RaceMaker terms. Since the program reconfigures fractional times based on surface and distance and the conversions are meant to represent a 6f dirt sprint, I have a certain amount of confidence using that 2-back race if necessary.

IMO, the problem with going say 5 back is that race may have been 2 or 3 months ago and it unlikely that the horse is in the same physical condition today (better or worse) as it was 5 races ago.

But hey, that's just my opinion.;)

Amazin

07-23-2003, 12:52 PM

Using 5 back certainly raises the "present form"issue but I have found it works both ways in that horses returning to their last good race based on surface or distance can and do pay very well.

I'm still a bit nebulous about the way your program works with distance and surface changes,Are you saying that if it's a dirt sprint,and the paceline used is a dirt route,your program will take the 6 furlong part of the route and convert it to sprint fractions for comparison to the other sprinters in the race?

rmania

07-23-2003, 01:05 PM

Originally posted by Amazin
I'm still a bit nebulous about the way your program works with distance and surface changes,Are you saying that if it's a dirt sprint,and the paceline used is a dirt route,your program will take the 6 furlong part of the route and convert it to sprint fractions for comparison to the other sprinters in the race? Basically speaking, you've identified the concept. Of course, there is a lot more involed with the conversion process than just taking the time for the first 6fs.

As for returning to a distance and surface that was profitable in the past, the horse still has to be (to some degree) fit enough to win such a race. This is why I try to focus on races no further back than 2. And the converted pacelines generally indicated whether the horse is fit or not.

Amazin

07-23-2003, 02:07 PM

Rmania:

I've come to the conclusion that verifying fitness is not allways obvious from the form.

One last question.Another difficult situation is turf.I personally pass the race in my program if there are more than 2 entries with no turf experience and the rest do have turf experience.What do you do?If you do rate these non turf horses,how can you rate them comparably with the turf horses?

rmania

07-23-2003, 02:32 PM

Originally posted by Amazin
Rmania:

I've come to the conclusion that verifying fitness is not allways obvious from the form.

I came to that conclusion years ago.

One last question.Another difficult situation is turf.I personally pass the race in my program if there are more than 2 entries with no turf experience and the rest do have turf experience.What do you do?If you do rate these non turf horses,how can you rate them comparably with the turf horses?

One of the basic elements to handicapping is the need to make assumptions (please no jokes). Its fairly obvious that turf sprints are run at a faster pace than dirt sprints. And its somewhat obvious that horses moving from dirt to turf generally produce quicker fractions. So, I have to assume this to be true for any horse making that transition.

So, with a little bit of adjustment to the turf fractional times (i.e., slow them down a bit) you can make a better decision as to which horses should be competitive. Pretty simple... right?

Dave Schwartz

07-23-2003, 10:01 PM

Okay, so here is what I am doing about this.

I am creating a section on my website called "Showcases." A showcase is a particular systematic approach, fully explained and documented.

It is still under development, but anyone interested can look at the first showcase at this link:

http://www.horsestreet.com/Showcases/01PaceHcp/

I began on July 1st and tested one week of CRC. Next comes HOL, BEL and CD.

It is a lot of work to document it like this, but it absolutely shows that there can be no fudging (otherwise I would get challenged by my own group).

Opinions?

Regards,
Dave Schwartz

PS: It appears I am back. (Guess I wasn't such an oak after all. <G>)

VetScratch

07-24-2003, 05:32 AM

Believe me, Dave, if I thought you were merely an oak, you wouldn't catch my attention! Welcome back, Mighty Redwood!

Dave Schwartz

07-24-2003, 11:15 AM

I have updated with more test results.

http://www.horsestreet.com/Showcases/01PaceHcp/

If anyone needs an explanation, or if something is unclear, please let me know.

Dave Schwartz

Amazin

07-24-2003, 11:51 AM

Rmania quote:"So, with a little bit of adjustment to the turf fractional times (i.e., slow them down a bit) you can make a better decision as to which horses should be competitive. Pretty simple... right?"

I wish it was that simple.I find that from dirt to turf there is a qualitative difference in performance rather than a quantitative difference in performance.For example a horse on Turf for the first time with sprint speed who suddenly has no speed on turf.Or a first time turf horse who beat me last year with horrible dirt PP's winning at 50-1 because he was bred by Danzing.

Another problem is like yesterday's feature at Del-Mar.The winner was an import @9-1.There are no pace lines to handicap imports and they are usually off a period of time before coming here(your fitness question comes in).But by manually capping I saw this horse was a definite contender.Julie Krone-Kathy Walsh(nice connection),7 for 8(first or second) lifetime on turf.Excellent fast works.This was only possible to manually handicap,and I do not allow my program to try handicap these kinds of races.It's futile with a program relying on a mathematical comparison.

azmike

07-24-2003, 08:46 PM

Originally posted by Dave Schwartz
I have updated with more test results.

http://www.horsestreet.com/Showcases/01PaceHcp/

If anyone needs an explanation, or if something is unclear, please let me know.

Dave Schwartz

I am very impressed with your work to date. I have not previously looked at your program but it certainly appears very well documented and very powerful. How long did it take to run the cards per track on average. If I downloaded all tracks for a day and decided to run this program for all tracks how much later would I have the results? Also, does your program permit you to print results and your program selections on the same page as you show at this web link?

I have frankly been cautious about checking out your program because I have read how complex and time consuming it is to learn. I am not a computer whiz so I was concerned if it would be too much. But this--if I am reading it correctly--looks like a great system to produce contenders with little manual effort. Looks good. Thanks.

Dave Schwartz

07-24-2003, 09:40 PM

Mike,

I am working on an updated summary and should have it up by the weekend.

The great majority of my workload in this project was creating the pages. The actual handicapping time was about 90 seconds per race, or a little less.

As for printing things out, that is always tough in a program that is extremely flexible. (What I want and what you want will probably be entirely different.) We have addressed this problem by using a "notes" file for each race. Almost every screen has a "notes" button that, when clicked, adds the current screen to the notepad file for the open race. You can even edit the screens BEFORE clicking the notes button. Thus, if I am only interested in showing the top 4 horses, I can delete the others, add a few hand typed comments of my own, and click "notes" to append it to the file for that race.

There is a function that allows the user to consolidate the notes for one card into a "card note file," and another function that adds all the "card notes" into a "daily notes" file. Some of our users then open this file in a Wordpad or some other word processor (where they can easily change the font and orientation) and print away.

If this fails to meet your needs you can always export the factors to a CSV file and import them into a spreadsheet or database.

Dave Schwartz

azmike

07-24-2003, 11:17 PM

Originally posted by Dave Schwartz
Mike,

As for printing things out, that is always tough in a program that is extremely flexible. (What I want and what you want will probably be entirely different.)

Dave Schwartz

Dave: Maybe my question wasn't phrased in the correct manner: When I viewed your pages on your current project I noticed the race results and the program choices where on the same page. Is that something I could do with the program "automatically" after races or would I have to "cut and paste" or spend a great deal of effort to combine print outs?

Also, I tried to read this entire thread but if I missed something I apologize for asking something that has been covered earlier in this thread--is this "project" an attempt to show what your program can do in "automatic" mode (black box ?) for purposes of the software contest? Is this project limited to one aspect of your program? I mean are there other ways to do this "automatically" with other factors? If yes to the last question, then did you pick this "pace" test because it has been the best method as far as ROI?

Thanks.

Dave Schwartz

07-25-2003, 12:49 AM

Mike,

There is no best method. It is whatever works for you.

The goal here was to produce a relatively simple to apply approach that would do a "good" job of selecting prime contenders. I just totalled the results of the test... Amazingly, averaging 3.66 horses per race, totally ignoring obviously unbettable races (i.e. there was a race with 3 horses at 2/1 or less... why play that one?), it showed a $2.06 $net (i.e. +3% ROI)! Only 629 bets, but that is pretty astounding.

What you put on the paper is up to you. Remember that this is but one of 50+ "systems" we have built. Some are simpler, most are more complex. Some use a database, some don't.

It just depends on what you are trying to accomplish.

Please forgive me if I am answering your questions vaguely. I am not trying to be difficult. Perhaps you should write to me privately instead of on this forum. (This is a bit more commercial than I try to be here at PA.)

Dave

Dave

lousycapper

07-25-2003, 01:11 AM

:D Let's put a fork in this baby... It's done, overdown, cremated already, dead and buried. There's not going to be a contest. Wake up and smell the coffee....ummmmm! :D

-L.C.

azmike

07-25-2003, 01:28 AM

Originally posted by lousycapper
:D Let's put a fork in this baby... It's done, overdown, cremated already, dead and buried. There's not going to be a contest. Wake up and smell the coffee....ummmmm! :D

-L.C.

I agree.

Although it would have been interesting to see some of the software programs go head to head.

I agree with some of the posters who say you can't really compare software programs in this manner. There are so many variables involved in how you set it up, how you use the data and output, and how you would bet it, if at all.

Having said that, it would interesting to see how well the various programs would do in picking the horses straight up. But I would expect they would all be similar. My experience with many other programs is most all get the favorites and others sometimes get the higher priced horses. I haven't seen one that stands out above the rest. It fluctuates--in my experience--track to track and day to day.

It would have been interesting though :)

hdcper

07-25-2003, 02:48 AM

Dave,

I have a question about the preparation of the filter(dynamic filter). If I am reading the setup right, the dates used for creating the filter is 6/1/01 thru 12/13/10 which seems to include the races you tested the filter on. Seems this would cause an error factor when you include the races in your model that you then handicap in the test.

Why not build your filter and then handicap races for say the next week and post them each day prior to the running of the races. Eliminates any question of what tracks you might leave out because of poor performance.

Again, will the dynamic filter based on the past predict the future?

Bill

rmania

07-25-2003, 08:40 AM

Derek2U

07-25-2003, 08:47 AM

Of course it's a Doable thing. And very interesting. I'm surprised
at how some peeps obsessions prevent them from tackling an
otherwise easy contest. Pre-Selection jitters? PS: Just DO IT.
As Joey Ramone sang, Stop Thinking About It.

rmania

07-25-2003, 09:16 AM

Well...

After two days, 10 sprints, and using the same selection criteria as the final week at HOL, the program is 6 for 10 (or 60% win rate) with an ROI of +71%. And these ain't no 5 hoss fields!

The only negitive was that I had to use 1.1 horses per race.

Yeah, yeah... I know it's only been 2 days but there seems to be a pattern forming here.

Dave Schwartz

07-25-2003, 10:50 AM

Hdcper,

>>>
You said:

I have a question about the preparation of the filter(dynamic filter). If I am reading the setup right, the dates used for creating the filter is 6/1/01 thru 12/13/10 which seems to include the races you tested the filter on. Seems this would cause an error factor when you include the races in your model that you then handicap in the test.

Why not build your filter and then handicap races for say the next week and post them each day prior to the running of the races. Eliminates any question of what tracks you might leave out because of poor performance.

Again, will the dynamic filter based on the past predict the future?<<<

When the dynamic filter is apllied, it resets the date to one day before the current race date. That is programmable by the user because some people don't get the results every day so it can be set to date -2 or whatever. But the default is race date minus 1. That is actually shown on the dynamic filter screenshot and discussed briefly. I think I will make it more obvious.

Thanks for the feedback.

Dave Schwartz

rmania

07-25-2003, 01:17 PM

Originally posted by jk3521
rmania,

Until your program uses BRIS or TSN data and can handicap other tracks besides So.Cal, I do not see much of a market for your software. Who needs So.Cal racing with 5 horse fields. And of course with small fields any program would show a good winning percentage. What is your winning percentage with fields of 8 or more? How many people use ITS data? Not me or anyone else I know . And lay off of Handle!
Sorry I didn’t respond earlier to your post. I guess I was just thinking/hoping that if you were indeed seeking the answers to your questions then you would/could find them on your own.

But alas, it appears that you are just one more in a long line of posters who find pleasure in this sort of dialog.

So here goes... I hope you don’t mind that I’ve broken your post up so that can attempt to answer each question.

“Until your program uses BRIS or TSN data and can handicap other tracks besides So.Cal, I do not see much of a market for your software.”

This may be true. But then again, BRIS and TSN are only more popular because of price. And, if one is playing 10 tracks a day it doesn’t take a genius to see why they are more popular. Of course, I only play one track so I don’t mind paying the extra buck.

“Who needs So.Cal racing with 5 horse fields. And of course with small fields any program would show a good winning percentage. What is your winning percentage with fields of 8 or more?”

Looks like you got me there! I revisited the results of my contest, pulling out only those races with 8 or more entries (more than half of the races posted) and my worst fears were realized. There was indeed a drop in the win percentage (46.15% vs 52.17%). But I was only using an average of 1.5 horses per race and I did have an equal amount of place finishers (not that that counts). And, I hate to say it but there was a drop in ROI (+45.71 vs +47.35). So I apologize if my original numbers were seen as misleading.

“How many people use ITS data?”

I know of at least one person.

“Not me or anyone else I know .”

Ok, so that’s 1 against what 2, 3 ?

“And lay off of Handle!”

Sorry, but I’m not sure of which handle you speak of.

BTW, are you the same person from the ES board that posted you were having so much fun with the program that you were forgetting to bet? Hmmm... Sounds like you’ve come up with a sure-fire system for at least breaking even at this game.

canuck

07-25-2003, 03:24 PM

rmania:D

jk3521

07-25-2003, 03:35 PM

rmania,

Nobody likes a wise guy! I just stated a few facts. Take your software and shove it.....

rmania

07-25-2003, 03:45 PM

Originally posted by jk3521
...I just stated a few facts.
Facts?

I don’t recall any facts mentioned in your post.

Plenty of accusations... but no facts.

Here’s something you might want to remember... if you start shooting at people, don’t be surprised if they shoot back.;)

jk3521

07-25-2003, 03:53 PM

Goodbye.... I hate egomaniacs !!

Topcat

07-26-2003, 12:13 AM

<<<Okay, so here is what I am doing about this.

I am creating a section on my website called "Showcases." A showcase is a particular systematic approach, fully explained and documented.

It is still under development, but anyone interested can look at the first showcase at this link:

http://www.horsestreet.com/Showcases/01PaceHcp/

I began on July 1st and tested one week of CRC. Next comes HOL, BEL and CD.

It is a lot of work to document it like this, but it absolutely shows that there can be no fudging (otherwise I would get challenged by my own group).

Opinions?

Regards,
Dave Schwartz>>>

Just out standing Dave!

If every software develper did a similar show case it would really give potential users an opportunity to see the ability of the program.

Hey- it might even cut down a bit on ythe number of people whose hoibby is not horseracing but criticizing.

lousycapper

07-26-2003, 12:04 PM

:D Will they work in my Grampa's "Mr. Coffee" coffee maker. :D

-L.C.

rmania

07-27-2003, 02:02 PM

Originally posted by Dave Schwartz
I just totalled the results of the test... Amazingly, averaging 3.66 horses per race, totally ignoring obviously unbettable races (i.e. there was a race with 3 horses at 2/1 or less... why play that one?), it showed a $2.06 $net (i.e. +3% ROI)! Only 629 bets, but that is pretty astounding.
Dave,

Been waiting three days now for somebody to ask this question:

Assuming all of your bets were win only, what percentage of those +/- 170 races did you win?

Dave Schwartz

07-27-2003, 05:01 PM

71%

Go to this link, click Introduction and find "results."

http://www.horsestreet.com/Showcases/01PaceHcp/

rmania

08-01-2003, 02:27 PM

After eight days, 38 sprints, and using the same selection criteria as the final week at HOL, the program is 19 for 38 (or 50% win rate) with an ROI of +176%.

This was accomplished betting win only using an average of 1.4 horses per race.

PaceAdvantage

08-01-2003, 04:36 PM

Are these picks posted? If not, this doesn't do me much good. I can, however, offer you this advice:

If your success rate maintains itself over a longer period of time, you're much better off just taking the product off the market, and making yourself a boatload of money by going to the mutuel windows with both fists....

rmania

08-01-2003, 05:37 PM

Originally posted by PaceAdvantage
Are these picks posted? If not, this doesn't do me much good. I can, however, offer you this advice:

If your success rate maintains itself over a longer period of time, you're much better off just taking the product off the market, and making yourself a boatload of money by going to the mutuel windows with both fists....
Regarding the picks, no their not posted. If I had my own website I would post them. But let me ask this... did the picks I posted during my contest help you? Probably not.

I realize that the are many skeptics when it comes to this kind of thing so I did try to establish some means for getting independent verification. I gave a copy of my program to one of the members at PA with the hope that he would, on occasion, chine in and support my claims.

The members name is Stonepony and I haven't heard from him since, though he still posts on occasion. I've tried e-mailing him but he doesn't reply.

If I or somebody else can't get Stonepony to talk then maybe you'd like to provide the verification PA. It can be arranged.

As for the boatload of money you mentioned... well let me just that I'm doing pretty good at the track. ;)

If ever there was a way to "have your cake and eat it too" this is it.

Lefty

08-01-2003, 08:08 PM

Why not just post the picks ahead of time in the selection forum and avoid any controversey? Isn't that what you did for Hol?

rmania

08-19-2003, 09:48 AM

After 4 complete weeks, 110 sprints, and using the same selection criteria as the final week at HOL, the program is 55 for 110 (or 50% win rate) with an ROI of +79%.

This was accomplished betting win only using an average of 1.45 horses per race.

Here's a breakdown of the winner's odds:

< 1.00 = 8 (15%)
1.00 - 1.90 = 12 (22%)
2.00 - 3.90 = 20 (36%)
4.00 & > = 15 (27%)

rmania

08-27-2003, 09:34 AM

After 5 complete weeks (only 2 left), 137 sprints, and using the same selection criteria as the final week at HOL, the program is 72 for 137 (or 53% win rate) with an ROI of +83%.

This was accomplished betting win only using an average of 1.44 horses per race.

Here's a breakdown of the winner's odds:

< 1.00 = 11 (15%)
1.00 - 1.90 = 16 (22%)
2.00 - 3.90 = 25 (35%)
4.00 & > = 20 (28%)

Here's some new stats...

Of the 137 sprints 77 (or 56%) were "single selection" races.
In those 77 races the selection won 36 times for a win percentage of 47%.

kitts

08-27-2003, 02:45 PM

rmania-

When are you going to add another track?

A) Very soon
B) Not very soon
C) Prefer to continue to ignore the question

Speed Figure

08-27-2003, 04:42 PM

Kitts,

Your talking about ignoring questions! :rolleyes:

kitts

08-27-2003, 06:36 PM

Speed figure-

I try very hard not to ignore questions. Sometimes the thread dies before I get in. If I have missed any questions, it wasn't because I was ignoring them. I'll gladly answer any questions you have. Either on this board or direct at kittszoo@cox.net

rmania

08-28-2003, 12:33 AM

Originally posted by kitts
rmania-

When are you going to add another track?

A) Very soon
B) Not very soon
C) Prefer to continue to ignore the question
kitts,

This may sound like a cliché but believe me if it was that easy, or if RaceMaker was your typical handicapping program, I would have had multiple tracks/circuits a long time ago.

This is going to take a lot of work and I continue to plug away. But I guess it won’t be soon enough for most of you guys. And I’ll tell you right now... I won’t take shortcuts at the risk of turning out an inferior product.

But you know what... I took a look at the addresses of current users and I’m seeing more east coast users than west coast users (unless you count Las Vegas as west coast).

I know I didn’t answer you question to your satisfaction but.....

I’ll try and keep you posted on my progress.

Speed Figure

08-28-2003, 12:44 AM

Are you saying that Las Vegas is on the East Coast!:eek:

rmania

08-28-2003, 12:47 AM

Originally posted by speed figure
Are you saying that Las Vegas is on the East Coast!:eek:
Last time I looked LV wasn't on any coast :confused:

panda1

08-28-2003, 06:29 AM

Hi,

Just wanted to let you all know that i purchased Racemaker a week ago ............ It is a great piece of software and does what he claims.........I know 5 days is only a small sample but i have a positive ROI for each and everyday just betting win bets...........If you like to play California tracks this is a must have...........All you guys complaing about the software just doing the California circuit............I'm from Delaware and live 5 minutes from delaware park...........But i'm no fool........if Racemaker can help me make $$$$ i don't care what track it handicaps ...........I don't know about you guys but $$$ is $$$ and i don't care what track it comes from:D ..............just my 2 cents........Have a nice day.............:cool:

Larry Hamilton

08-28-2003, 09:30 AM

I have a comment about all software for sale on the net, not just the one under discussion.

We have only four ways to determine if we should purchase the product. 1) Buy it blind without regard to our lack of knowledge about the author and the product. 2) Get a free test drive from the author. 3) As best as is possible with a keyboard, know the author. 4) Buy it on word-of-mouth from a trusted source.

Those are facts, now here are conclusions: I would never buy anything from someone whose claims are all red-board and whose name is anonymous (method 1 above). There are several reputable salesmen here with reputable products (see methods 2,3, and 4 above), why does anyone have to buy phantom software from phantom humans? If you do, you deserve what you get.

My favorite line with respect to this point: "Hey buddy, you want to buy this Rolex watch? Only $5"

rmania

08-28-2003, 09:21 PM

My favorite line with respect to the previous post is "Build a better mouse trap and the Larry Hamiltons of the world will line up to shoot holes in it."

Actually I just made that up but it's soooo true.:rolleyes:

kitts

08-28-2003, 09:53 PM

rmania-

I appreciate your position on tracks covered. I live in Nevada and am able to play multiple tracks. I tried your program when you were offering a test version and I have no major issues with it. Certainly no issues with limiting it to sprints. But I do need multiple tracks and will try yours when that happens. I hope you let us all know about it when that fine day comes along.

Larry Hamilton

08-29-2003, 09:34 AM

I wrote many angry replies, threw them all away when I realized, nobody promoted me to guardian of the horse player.

Bruddah

08-30-2003, 03:34 AM

I have used method #3 very successfully over the years. In doing so, I have found two individuals to be of the highest integrity and honesty over the years. They are Larry and Nathan. If either of these two said they had developed a washing machine, which could handicap horses, I would be the first in line to buy. They also are two of the sharpest knives in the drawer. They have insight into handicapping which surpasses the average poster on the subject.

The above in no way takes away from the other intelligencia on this board. Certainly, there are others of this calibre. Then again, there are those who "wannabe" viewed in this manner and they post their drival ad nauseum. ;)

rmania

08-30-2003, 07:47 AM

Originally posted by panda1
Hi,

Just wanted to let you all know that....
Before you get too excited there's something you should consider.
Here's a quote from one of the more admired and respected members of this board:

"One of the more important facets of software is value. This facet has been ignored in the above conversations. The concept of value is rather simple, based on one question. Can I profit using it? If the answer is yes, then it's value is infinite. If the answer is no, it's value is nothing.

Of course, the one question leads to another--How much profit? The answer to the second question is the constraint of the first question."

rmania

09-14-2003, 12:46 PM

After the entire Del Mar meet, 193 sprints, and using the same selection criteria as the final week at HOL, the program is 98 for 193 (or 51% win rate) with an ROI of +66%.

This was accomplished betting win only using an average of 1.41 horses per race.

Of the 193 sprints 114 (or 59%) were "single selection" races.
In those 114 races the selection won 50 times for a win percentage of 44%.

Depending on individual success at Del Mar this year, these numbers may or may not seem impressive. Of course, these are just "blind" plays that required no additional handicapping.

So how have things gone for those who actually made their own decisions while using the software?

Well, you guys remember Panda1, right. The one (and only) PA member to post about his success. The one who was immediately labeled as an anonymous shill. Well, it turns out that another member, one who attended the Toga Party, knows Panda1 personally and actually talked to him about his experience. And, since this board is somewhat light on success stories, I've decided to post a portion of an e-mail I received from Panda1.

9/9/03: Yesterday i had a good day as well...........Winners paying $ 28.60 in race 2 and a second i believe in the 5 or 6 race the 1 horse paid $ 17.80 to place.......on the simulator he was leading a long way before getting caught.........and i hammered the 2 horse in the last race.................I've been using this software for about a month now and all i do is make money.........Can't get any better than this.............Thanks again