Has anyone has experience with any of this software?

[csperberg]

I have a list of some software that I was just wondering if anyone can chime in on and its usability for horse racing. Whether you used it or not, heard through the grape vine or what have you just would like some opinions.

I am providing a link to each sites home page incase anyone is not familar

1.Maple 10 - http://www.maplesoft.com/products/maple/

2.SPSS 15 - http://www.spss.com/spss/index.htm?source=homepage&hpzone=ad_box

3.NeuroSolutions 5 Developers Edition - http://www.neurosolutions.com/products/ns/features.html

4.SAS V9 - http://support.sas.com/software/index.htm

5.DatabaseSpy 2007 - http://www.altova.com/products/databasespy/database_tool.html


I was just wondering what ones would have the best practical use for horse racing. Does the cost factor of some of these make them a little too pricey, or is that extra premium you pay worth the extra investment for what you can do as compared to a less costly piece of software.

Some of these also have Excel add-ins, like Maple 10 and NeuroSolutions 5.

Once again just looking for opinions thanks.

[NEhandicapper]

I have used Maple a lot, not sure how it can be used for horses unless you need the power of an advanced mathematics program. I used Maple for helping in modeling and presenting structural and hydro systems when I had to fill in the gaps from the CAD programs.



If you have a large amount of data and would like to evaluate this data using advanced statistical methods, then I can see using Maple. You can accomplish many statistical methods with MS Excel, but for some real muscle, you could extract this data and have fun with it in Maple.



But, if you have an ambition to invent the “magic” equation that could be applied for handicapping, then Maple can help with the grunt work. I always thought that a good partial differential equation would solve everything for horseracing, just need to discover the right boundaries :bang: .

[K9Pup]

I

4.SAS V9 - http://support.sas.com/software/index.htm


I was just wondering what ones would have the best practical use for horse racing. Does the cost factor of some of these make them a little too pricey, or is that extra premium you pay worth the extra investment for what you can do as compared to a less costly piece of software.

Some of these also have Excel add-ins, like Maple 10 and NeuroSolutions 5.

Once again just looking for opinions thanks.

I used the mainframe version of SAS back in my working days. I know the PC version is a fairly robust piece of software. The learning curve could be a problem.

[Red Knave]

2.SPSS 15 - http://www.spss.com/spss/index.htm?source=homepage&hpzone=ad_boxA few guys on the board have mentioned this so you may want to do a search for SPSS in the archives.
3.NeuroSolutions 5 Developers Edition - http://www.neurosolutions.com/products/ns/features.htmlI tested a much older version of this for a while. It may be me but, since I have no grounding in AI, neural nets, genetic algorithms etc., I really didn't know what I was doing or how to interpret the results I got Iif I got any :) ). Again, there are some guys on this board who know this stuff inside and out. Do some searches on AI, Neural etc. and see what you find.
4.SAS V9 - http://support.sas.com/software/index.htmLike K9, I worked with a mainframe version of this 10 years or so ago. My opinion is that, if you just want to set up a database that you can query, use one of the many open source ones like Firebird, MySQL, PostGreSQL and others. A lot of guys on this board use Abcess, I mean Access ;) .
5.DatabaseSpy 2007 - http://www.altova.com/products/databasespy/database_tool.html (http://www.altova.com/products/databasespy/database_tool.html%20[/QUOTE)The above also holds true for this product. Unless you already have a database created and set up you could use other, less costly, tools to do the querying and reporting.

Just my .02.

[csperberg]

much appreciated opinions from everyone.

I am sure that everyone of these has pros and cons to them and their useage.

Was just more curious on if any were worth the time to learn and the cost as well. Especially knowing that there are always other cost effective like products available as well as open source. What is it that these higher priced software offer that set them apart from their less costly competitors if any?

The DatabaseSpy program I guess is really in a differnt field than the analytical and mathematical software since it really is just a database tool more than anything. Though I thought the flexibility it offered under one package was sort of appealing. Also checked out their XMLSpy, that was a nice XML editor easy to use for anybody.

[K9Pup]

Was just more curious on if any were worth the time to learn and the cost as well. Especially knowing that there are always other cost effective like products available as well as open source. What is it that these higher priced software offer that set them apart from their less costly competitors if any?



A lot really depends on where you are coming from and where you want to go. Someone with a lot of programming knowledge will have a easier time learning these products than someone starting from scratch.

You can do simple statistical functions with just about any of the mentioned DB programs. If you want to get into more complex statistical analysis then you might need SPSS or SAS or one of the others. I guess the big question is there something you want to do today that you can't.

[garyoz]

I have experience with SAS and SPSS. Not sure that complex statistical analysis will get you anywhere in this sport. I've tried. It can be time consuming and doubt that you will have a eureka moment. Correlational analysis is sufficient.

You might want to read some of the threads on multiple regression analysis on this board. One thing I've always thought could have potential is factor analysis, a type of multivariate analysis where variables are combined--sort of power numbers for different but related factors.

There are many issues about the application of formal statistical analysis to racing--ultimately the input tends to be one of pacelines--and their selection is a whole other issue. Ken Massa at HTR has consistently done the most rigorous analysis of handicapping factors in his free newsletters (IMHO). HSH (Dave Schwarz) has a program that appears to do the best database analysis and model building. There are other programs that others favor. You might be better off building on what these programs already do.

[ryesteve]

I have a list of some software that I was just wondering if anyone can chime in on and its usability for horse racing. Whether you used it or not, heard through the grape vine or what have you just would like some opinions.
I've programmed in SAS for the past 20+ years. There's nothing you wouldn't be able to do with it data-wise. It's also very fast, even when dealing with large datasets. The downside is that it's a programming language with a steep learning curve, not a point-and-click query front end that you'd be able to immediately dive into.

I have experience with SAS and SPSS. Not sure that complex statistical analysis will get you anywhere in this sport.

You might be better off building on what these programs already do.
I don't disagree with this advice. But I've found that SAS is immensely useful as an adjunct to the software I use (JCapper), mining and analyzing its db and output in ways that the program itself can't.

[prank]

I have a list of some software that I was just wondering if anyone can chime in on and its usability for horse racing. Whether you used it or not, heard through the grape vine or what have you just would like some opinions.

I am providing a link to each sites home page incase anyone is not familar

1.Maple 10 - http://www.maplesoft.com/products/maple/

2.SPSS 15 - http://www.spss.com/spss/index.htm?source=homepage&hpzone=ad_box

3.NeuroSolutions 5 Developers Edition - http://www.neurosolutions.com/products/ns/features.html

4.SAS V9 - http://support.sas.com/software/index.htm

5.DatabaseSpy 2007 - http://www.altova.com/products/databasespy/database_tool.html


If I may add my $0.02 to the data modeling tools discussion, here are the tools I've used, am using, or am considering using.

As background, I'm a PhD student in statistics. I'm not about to handicap horses -- I've got to get done with my PhD (and my wife would kill me, unless I was actually winning, but winning means spending the time to get a model, etc., which means not working on my research...), though maybe I'd do something in my "spare" time...

For the last 3 years, my primary tools have been R and Matlab. R is free & open source; it's generally considered a functional clone of S-plus. It has a large community, and it can handle the data used for horse racing (NB: I'm being very general here: you may need to reformat files in order to use them with R or any of the other tools; I just mean to address the concern about the usability for handicapping). Moreover, there are a lot of R libraries (add-ons) that you can use for advanced visualization and modeling. If you know how to program BASIC, then R wouldn't be difficult. If you know C, Python or other such languages, then R will be quite easy.

Matlab is my drug of choice for very large data sets. For instance, I'm wrapping up a project with a million variables and millions of observations. I've got to use it because R is unsuitable for this scale of data.

I'm very seriously contemplating Python for my next project, because it gives me more control over data types, it's object oriented, I can do threading and distributed computing, and there's neat stuff that's been developed for scientific computing, such as Numpy.

Now, that's probably way more than you're looking to bite off.

If you're looking for fairly standard statistical analyses, I'd recommend SPSS. It's the program of choice for the social science fields, and though some may say it makes it too easy to get into trouble, it is better to bang your head against the wall to understand statistics than it is to bang your head against the wall to understand programming. The former may lead to insights, the latter leads to, at best, one roadblock being substituted for another. (Some of us have no choice, if the tools really aren't available, though...such is research.)

If you've got a mathematical bent, I'd say that Matlab is the only way to go. There's an open source clone, called Octave, which I've not tried. It may be acceptable, and I'd recommend a glance if you're willing to explore it.



I was just wondering what ones would have the best practical use for horse racing. Does the cost factor of some of these make them a little too pricey, or is that extra premium you pay worth the extra investment for what you can do as compared to a less costly piece of software.

Some of these also have Excel add-ins, like Maple 10 and NeuroSolutions 5.

Once again just looking for opinions thanks.

I'd say that you needn't really worry about Excel, as all of these can handle visualization of data. The nice thing about a spreadsheet program is that it makes data entry easier. Matlab and R, as far as I know, really suck at making it feasible to explore and modify specific values. In rare occasions, I get data from some source, send it to a CSV (comma separated value) file, tweak it in Excel, and then save as a CSV for use in R or Matlab. (TSV, tab separated values is another fine format.)

Maple and SAS should be fine for developing complicated analyses, but I wouldn't be so quick to recommend them, only because the former isn't as established for statistical analyses and the latter is, to put it politely, idiosyncratic (though it *is* extremely popular).

Cheers,

Prank