Battle of the Handicapping Softwares

[DarkDream]

Since being on this forum, I've been aware of a lot of different handicapping software out there.

Wouldn't it be a great idea to make all the different softwares talk the talk. That is, to see how a certain set of handicapping softwares do against each other.

A competition of sorts would be organized and each software would be given the goal of handicapping a set of races over a period time. Each user representing the software would then post the software picks on the forum prior to the race.

There would be a tally done to see which software has the highest win%, money made and so on. For such a competition, there would only be theoretical $2 win bets made.

I understand that there may have to be some human skill involved (such as picking a pace-line), but if a mechanical rule was set out that each user of the software obeyed, it would at least be a consistent way of attempting to measure a software's ability to make money.

What annoys me with so many of the software out there, is that all of them claim to be the ultimate best and so on. None of them present any statistics (or at least lame made up stats such as it picks winners 90% of the time) of the ability of the program.

Personally, if I produced a program. I would test it extensively and present my statistics, and tell the person that they can expect this % of winners and so on, otherwise I'll give them their money back.

How does this idea sound?

--DarkDream

[ryesteve]

I understand that there may have to be some human skill involved
Well, yeah... most software isn't meant to be a blackbox, which makes it hard to do the sort of objective comparison you have in mind.

[andicap]

Agree with Steve (and I never agree with someone from Rye)

In a 10 horse race, 10 users of HTR could come up with 5 or more horses in a lot of races. It's a useless and irrelevant test. I don't use the software because it "picks" good horses. I use it because

a) the numbers are good, but lot of people have good numbers
b) the program is very easy to use for MY purposes and has excellent features and tools not found on other programs that I feel comfortable using.
c) the developer, Ken Massa, is honest, smart, generous, and accessible.
d) There are constant upgrades that keep making it better
e) Public bulletin board to ask questions
f) monthly newsletter
g) Ability for Ken to develop custom versions of software for moderate prices so I have adapted HTR around my play instead of vice versa as you need to do on many other programs.
h) no hype. no "70% winners." Just a great tool that enables me to handicap a race in 5-10 minutes and I'm aiming to get it down even more.

There are other programs that can boast many of these attributes as well, and I'll never knock a program unless the developer makes outlandish boasts (a la Tom Console).

So a contest would prove nothing about the software, only my ability to use it well. And having everyone pick the same paceline seems silly as well.
What's the output you would use?
-AP?
-Speed?
--some "power" number, which HTR does not have unless you count the "K" rating which is not designed to be used by itself.

I use about 8 figures many of which were custom made for me by Ken. In a heavily pressured route race I'll rely on a late speed or energy rating for example.

[joeyspicks]

ALL the legitimate software mentioned on this site are most defintly NOT black box "fraudware". What your question implies is that you just turn on the software, process the race and the software cranks out the BEST horse. This could be measured in the way you have described.......however not many (if any) people on this board are using software this way (at least I dont think so!)

The software mentioned on this site and reviewed by many many users over the years are really "information processing machines" which should then be interacted with by the USER. Each user could easily come up with different picks for different reasons (as it should be........most races have NO BEST HORSE). The user must supply his knowledge and experience to the information provided by the software. Even if you had a contest.....with each user declaring WHAT SOFTWARE they are using to make their picks.....you could easily have multiple users of the SAME software picking different horses and finishing far apart in the standings. YOU just cannot get away from the human element and PERSONAL RESPONSIBILTY for your OWN selections. Get over the black box type of thinking asap for your own benefit!

[JustRalph]

http://www.paceadvantage.com/forum/showthread.php?t=10694&highlight=software+contest

check out this thread from a while back.................

[ryesteve]

Agree with Steve (and I never agree with someone from Rye)
You can think of me as being from Queens if you want to maintain that :D

http://www.paceadvantage.com/forum/showthread.php?t=10694&highlight=software+contest

check out this thread from a while back.................
Wow, I'm surprised it took that long to come to the conclusion that it wouldn't be a helpful contest

[JustRalph]

Wow, I'm surprised it took that long to come to the conclusion that it wouldn't be a helpful contest

That isn't all..........there was a very long thread before that one.......that was a little rougher if I remember it right. But it was very informative because many of the software developers participated in some philosophical discussions. It may still be readable.......not sure though.

Do a lookup/Search for "Team Thorotech" and you might find some stuff on the Team Software contest that played out here a couple years ago. It was pretty fun. There were some wingnuts from a few other boards who came in and shot their mouths off a little........but overall the contest was fun.

[NoDayJob]

:lol: Good 6213119146 luck. This has been hashed and re-hashed so many times--- enough already. :lol:

NDJ

[kingfin66]

It would be much more fun to just have a handicapping contest on the board. Bubbles runs the one using the free DRF pp's and there is a P4 contest as well. What I'm talking about would be more of daily selection posting contest, where people have $10 to spend to either pick 5 horses $2 each, 1 @ $10 or any $2 incremental combos in between. WIN BETS ONLY. There could be a cap on the number of daily selection, or just set a weekly limit... Run it on Wed-Monday when Saratoga starts and end it when both Saratoga and Del Mar have ended. This type of contest would bring out the real cream of the handicapping crop as it would reward value bettors. When registering for the contest, people could declare what software, if any, they use as their primary handicapping tool. At the end, the software would not be declared the winner. The handicapper would.

[kev]

Or one way to see if there good or not is to see how many user have enter contest around the country and fin. well in those. Then again I don't know if everyone that wins shares what he/she uses after winning a contest, it would be a good plug for the company if they could get some free data out of it. For ex. over on the thorograph board the other day this....
"A big tip of the hat to TG customer Paul Shurman, who won the Bradley contest over the weekend, which sends him back to the NTRA finals.

Also congratulations to Paul Shurman, who won the Youbet contest the same day.

That's right, Paul won both. Since by rule he can only have one seat in Vegas, Paul gave up his Youbet title.

Which made the winner of that contest TG customers Michelle and Charlie Thompson, who will join Paul in Vegas."

Also the guy who won the big daddy contest last year was a Thorograph user along with the DRF, even when they had the show on ESPN, it had the guy using TG figures while capping a race.

[Buckeye]

let's put it this way, who's got a profitable black box and selling it?

I didn't think so!

I think it's necessary to "get ones feet wet" in order to find the software (if any) that best fits your feet.

A contest would prove nothing as far as which software per se is better, it would only prove which user of a given software (if they're using one) had the most success within a limited timeframe.

[Buckeye]

One more quick question for the paceline pickers,

what do you do it you can't reasonably pick a paceline . . .

That's right, you pass the race. Those are the races that must be looked at differently. (I'm always stating the obvious because sometimes it gets overlooked.)

How does a program (or user of the program) deal with these types of races other than "I pass" ?

The best answer - as I see it - is to relinquish control to another thought process that may not even be your own.

Now we may be "getting somewhere" :)

[PaceAdvantage]

A battle of handicapping software should involve the software and the creator, nothing more. For instance, Dave Schwartz would make the selections using HorseStreet (if he were to participate).

[Overlay]

:lol: Good 6213119146 luck. This has been hashed and re-hashed so many times--- enough already. :lol:

NDJ

I assume you meant 6213119147. :D

[Lasix1]

A battle of handicapping software should involve the software and the creator, nothing more. For instance, Dave Schwartz would make the selections using HorseStreet (if he were to participate).
Actually, PA, I'm not sure this would prove anything either. Just because Dave or Ken or whomever can or can't make money off their own program, doesn't mean that you or I couldn't make money off it, or that we wouldn't lose our shirts trying even if they didn't.

Dick Mitchell and his genius programmer at Cynthia Publishing created the All-in-One Program, but at the track Ron Ambrose used it very differently than Dick did, and with different results.

Moreover, some programs such as those with automatic pace line selections could be used as black boxes, but others don't really have that capability.

As others have said, computers are handy tools, but most of them require human judgement and I would be suspicious of any that claimed it didn't.

[ratpack]

[QUOTE=Lasix1]Dick Mitchell and his genius programmer at Cynthia Publishing created the All-in-One Program, but at the track Ron Ambrose used it very differently than Dick did, and with different results.

QUOTE]

Correct, I know a guy using All in One #6 that found a way in the reports of AIO that he could see when the top choice would win or place but not Show and he would exploite that in certain distance and class situations to win trifectas.

I am sure he used AIO in other ways but in that one situation he told me about out of 78 races that top choice either won or placed 35 times and showed 3 times and out the rest.

It is how you use the program

[DarkDream]

ALL the legitimate software mentioned on this site are most defintly NOT black box "fraudware". What your question implies is that you just turn on the software, process the race and the software cranks out the BEST horse. This could be measured in the way you have described.......however not many (if any) people on this board are using software this way (at least I dont think so!)

Just because people don't use software in a black box fashion, doesn't mean you can't make up rules to go ahead and treat it as such.

I did recognize the fact that some software may be difficult to present its picks, because of the need of human interaction, but with appropriate "rules" and guidelines in place one could do so.

The software mentioned on this site and reviewed by many many users over the years are really "information processing machines" which should then be interacted with by the USER. Each user could easily come up with different picks for different reasons (as it should be........most races have NO BEST HORSE). The user must supply his knowledge and experience to the information provided by the software. Even if you had a contest.....with each user declaring WHAT SOFTWARE they are using to make their picks.....you could easily have multiple users of the SAME software picking different horses and finishing far apart in the standings.

What you are saying is borderline ridiculous. Your argument is that many of the handicapping software out there is as such, that give it to a hundred different people and you'll get a 100 different picks. Then what's the use of having the software in the first place? At the very least, an experienced group of users would most likely pick in general the same top two horses. I would also imagine these top two horses would most likely have a fair chance of winning. That's because the tool is actually helping them pick better. If not, why are they using it in the first place?

Just because you have a tool, and a hundred different people may *use* it in a different way, that is not to say they can't get the same result.

To help eliminate the problem of different people making different picks, like I pointed out, you could come up with a set of rules that people would adhere to (a best approach to using the tool for example).

YOU just cannot get away from the human element and PERSONAL RESPONSIBILTY for your OWN selections. Get over the black box type of thinking asap for your own benefit!

Get over the black box type of thinking! It is the type of black box thinking that has led to computers, software -- handicapping programs. As for black box thinking, artificial intelligence and forecasting has shown that it can out compete so called "experts" in various fields.

My results of using AI to predict horses winning races, has done pretty darn well so far. Once I get it to a certain stage, I'll prove it to anyone that black box type approaches do work.

--DD

[PaceAdvantage]

Actually, PA, I'm not sure this would prove anything either. Just because Dave or Ken or whomever can or can't make money off their own program, doesn't mean that you or I couldn't make money off it, or that we wouldn't lose our shirts trying even if they didn't.

It's the fairest way I can think of to have different software products duke it out. Who better to run the show than the creator himself?

But, to appease folks that think like you do, I would allow that if the creator does not wish to run the software and make the picks, then he can designate whomever he thinks is his most CAPABLE customer to make the selections in place of the software creator/author.

[betchatoo]

Though, in the end, it still might not prove what is the best software, a fun contest might involve having a team approach (say 5 users) from each software. These teams would work together on selections. Let each software creator choose his (or her) team, put together a set of rules that makes sense and come out firing

[ryesteve]

Your argument is that many of the handicapping software out there is as such, that give it to a hundred different people and you'll get a 100 different picks. Then what's the use of having the software in the first place?
That's kinda like saying, "100 people who read the Wall Street Journal end up investing in 100 different stocks. I don't understand why they bother printing that thing in the first place"

[Pace Cap'n]

Guarantee you they wouldn't come up with more than 12, or 14 tops.

[BillW]

That's kinda like saying, "100 people who read the Wall Street Journal end up investing in 100 different stocks. I don't understand why they bother printing that thing in the first place"

Kinda sounds like the DRF doesn't it!

Evaluating commercial software using an objective procedure when no responsible vendor would recommend using their software in that manner seems irrational to me. It would be extremely difficult to remove human judgement from the handicapping process and even more difficult to evaluate the improved performance of a software package when that element was added back in. The results of an exercise such as this would certainly have little to do with the relative quality of the software.

May the luckiest program win :rolleyes: .

Bill

[midnight]

It'd be a no-win proposition for the software authors.

Steve Wolsom posted 500 plays for his software and showed a profit. I doubt the world beat a path to his door.

Even a month-long contest wouldn't prove anything. The spot play I've posted in Selections the last three days has a three-year win rate of about 16% and an ROI of 1.10. It's had bad months of 8% and 0.55 ROI and good months of 24% and 1.80 ROI, a lot of fluctuation.

If the software has a bad run, it's branded as junk, and anybody with an agenda can point to the contest thereafter as "proof". If the software has a good run, it's dismissed as luck. There's nothing to gain, in my opinion.

[BillW]

I agree. It would simply bring out the wingnuts making irrational criticisms and personal attacks aimed at the vendors.

[46zilzal]

I moderate a group that discusses a software approach and with the SAME DATA, SAME READOUTS, we get many different interpretations...there is no black box because the wagering side has little if anything to do with the 'capping side.

many folks (self included) learn that at specifc courses, one can outright dismiss many of the readouts just based on personal records. A tool: works for different people different ways.

[Lasix1]

I moderate a group that discusses a software approach and with the SAME DATA, SAME READOUTS, we get many different interpretations...there is no black box because the wagering side has little if anything to do with the 'capping side.
Truer words were never spoken, 46zilzal. Professional handicappers who I know, say that the relationship between betting and handicapping is as much as a 70/30 ratio favoring betting strategy.

That's why PA's suggestion about a shoot-out between competent players using different software might prove a lot about the players, but virtually nothing about the software.

It might not even prove all that much about the players either, since, at least in the short run, good players are like classy horses---they take turns beating one another.

[PaceAdvantage]

I guess this is one reason why we didn't have a contest.

The other reason is most of the software vendors are chicken....LOL

[Jeff P]

The other reason is most of the software vendors are chicken....LOL I thought the real reason the last thread like this one fell apart was that those thinking of participating couldn't seem to agree on an acceptable format. A contest like this could actually be a lot of fun for all involved. Let's throw some suggested formats out there and then let's pick one. I'll step up to the plate and play.

-jp

.

[Bruddah]

I am a computer novice compared to the intelligencia of this board. However, Jeffs' software has given me the ability to use my own trainer angles and develop a data base, I had only dreamed of before. He patiently walked me through the learning curve, which in retrospect now seems so simplistic.

Personally, in this contest, I see no gain, for any of the developers of these programs. There can be only one winner, and I am sure there is more than one excellent program out there. For me, it is JCapper and Jeff P.. Both are in a class by themselves.

It is no suprise to me Jeff has taken up the challenge first, with really nothing to be gained. He is very dedicated to our sport and finding the "Rosetta Stone" to electronic handicapping. He has complete confidence in the program he has developed, as do I.

The above message is from one satisfied customer. :)

[joeyspicks]

Why dont you simply run a contest with catagories for each piece of software......and just see WHO is good at using that software??

I mean there would be a group of people using HTR, JCapper, Netcapper, HSH,
ETC, ETC

Within each group people would post (each person can join one group). I dont know if this can be done or what the mechanics of it would be......but it would be interesting to see the wide range of results .

[andicap]

Take the 5 players for each software approach but instead of ONE selection (how the hell would they ever agree on anything?), use ALL their picks in a simple, Win-place format. No exotics -- to keep the scoring simple -- and add up their points in the end.
If not 5 players -- might be tough to find 5 who find to take part -- then try 3.
I doubt the software vendors themselves would be interested in such a contest because if they didn't fare well it would be terrible publicity. Can't blame them.

I think such a contest could be fun IF it was made clear it wasn't a test of the particular software.
After all, if a handicapping pro finishes out of the money in the World Series of Handicapping, do we think he's a terrible handicapper?
If a top poker pro finishes far down the list in the World Series of Poker do we think he's a bust as a player?

Of course not. Short-term luck plays a huge role in both handicapping the horses and playing poker. It's those who have the odds on their side, the mental agility and toughness and overall instincts who win in the long run.
Same thing with software.

As for having the owners of the software take part, well, Ken Massa is a fine handicapper, but I'm sure he would admit there are others who do better with HTR than himself. One reason is he has to spend a lot of time coding instead of handicapping/research, etc.

[Tom]

I have an idea....post the printouts of various programs for selected races and let everyone pick thier own horses based on all the programs. Enen then, I am sure some programs will not lend themselves readily to this, but it would allow people to review actual races - before they are run - and see how they would handicap them.

[ryesteve]

I have an idea....post the printouts of various programs for selected races and let everyone pick thier own horses based on all the programs. Enen then, I am sure some programs will not lend themselves readily to this, but it would allow people to review actual races - before they are run - and see how they would handicap them.
That wouldn't work too well for a db handicapping program like JCapper, and from what I understand, HSH.

[kingfin66]

I have an idea....post the printouts of various programs for selected races and let everyone pick thier own horses based on all the programs. Enen then, I am sure some programs will not lend themselves readily to this, but it would allow people to review actual races - before they are run - and see how they would handicap them.

I've actually tried to post software output on PA before, but have received a message that the file size is too large. Of course, being that it's me and all, there could be a user error involved in the process.

[BillW]

I seem to remember 32K is the file size limit - check with PA

[Jeff P]

Originally Posted by Tom
I have an idea....post the printouts of various programs for selected races and let everyone pick thier own horses based on all the programs. Enen then, I am sure some programs will not lend themselves readily to this, but it would allow people to review actual races - before they are run - and see how they would handicap them. I really don't forsee any problems on my end if this turns out to be the agreed upon format. I could either post simple cut 'n pastes of report content into the contest thread or upload the reports themselves to my own site and then simply post links to the reports in the contest thread.

-jp

.

[ryesteve]

I really don't forsee any problems on my end if this turns out to be the agreed upon format. I could either post simple cut 'n pastes of report content into the contest thread or upload the reports themselves to my own site and then simply post links to the reports in the contest thread.

-jp

.
So what happens when the reports show a couple bombs marked with one of your personal UDMs, and people ask you to tell how it's defined? ;)

[Jeff P]

So what happens when the reports show a couple bombs marked with one of your personal UDMs, and people ask you to tell how it's defined? If posting a report turns out to be the agreed upon format and IF I manage to hit a bomb or two (I can dream can't I?) then I wouldn't really mind explaining what I liked about a horse if asked.

-jp

.

[PaceAdvantage]

I'm sure each vendor knows of a player or players who are winning with their software. Why would they be afraid to let that player represent them in a contest?

The contest could go on for as long as everyone agrees. One month, two months, whatever. It should go on long enough to negate much of the "short term luck" people talk about on this thread.

BTW, if such a contest were to take place, software that did not come with an official endorsement of participation from the vendor/author would not be allowed to compete.

[betchatoo]

Here is my suggestion for the contest. Let's try and get 5 player teams with a consensus pick (each team to make their own rules about how a consensus is reached). This makes sense for a couple of reasons but chief among them is, if this is going to be an every day contest for an extended period, not everyone may be available every day to make their selections. This would screw up the scoring. With 5 players you would have at least one player available each day to make the team's selections.

Rules (as I envision them)
1. Contest to last 6 weeks

2. Play 5 tracks each day. 3 major (say the top NY, Fla, and Ky track) and 2 minor (i.e. Mnr and Pen). This would give people at least an idea how the software works on the type of tracks they like to play (tracks to be adjusted on days when all 5 do not run)

3. Play 3 races from each track (a total of 15 races per day), individual teams to select which races they wish to play. This will give the ability to use software in those races they feel offer most value

4. $2 win/place bet on each horse (if this works out well exotics can always be included in a future contest). No caps. If a team can pick a $100+ horse, good for them!

5. Team with highest amount of money earned is declared winner