Software Installation Preferences - Horse Racing Forum - PaceAdvantage.Com

DeltaLover · 11-08-2010, 12:49 PM

When it comes to distribution of an application we have various methods that each one has it's own advantages and disadvantages... Can you please vote on your personal taste?

DeltaLover · 11-09-2010, 01:20 PM

Interesting...

Although I expected the vast majority to prefer a pure Web 2.0 application without anything to be installed locally the poll reveals that a stand alone WITHOUT connection to the internet is the most desirable set up!

I really do not understand why this is the case considering the pros and cons of both worlds.

A web based application does not have any installation or update overhead, is cross platform and although it might lack some advanced features of a its full blown desktop counterparts still can deliver a pretty good user experience..

As far as the connectivity to the internet I think that its benefits are pretty easy to realize in comparison to a stand alone (real time updates, intraday modifications etc)...

I would like to hear some reasoning about your preferences so I can understand further your reasoning.....

Ghostrider5 · 11-09-2010, 04:41 PM

With a web based application if something happens to the website or the software vendor. You don't have any software to use. For example i bought a handicapping software program called propace a couple of years ago. But the software vendor died in 2009. I needed a reinstall because my computer crashed. The software required the software vendor to give me unlock key to reinstall the software. Got lucky and was able to get the programmer of the software to get me a new program with unlock key. That's why i rather own the software program outright. So i don't have to worry about situations like that again.

ghostrider5

DeltaLover · 11-09-2010, 05:47 PM

How about a software that actually is free but what you actually buy is the data that will be used by it?

My understanding is that the program you are referring to has some kind of import mechanism to import data from some kind of a of source like a data base complying to a specific schema or a specific type of file format like for example the Bris MCP or DRF.....

This type of an approach requires a relative complete data base and it complicates development effort trying to achieve compatibility with all the existing data vendors.

I am referring to an application that will propagate all data centric processing to the server and use the client's machine as a pretty thin layer of interaction...

Ghostrider5 · 11-09-2010, 06:09 PM

That sounds good! Like the free software idea. But need more details of what kind of software? What are data service provider options? How much control would i have over the standalone capablities?

ghostrider5

DeltaLover · 11-09-2010, 06:31 PM

Let's say you have a software that allows you to describe any handicapping factor and back-test it for its ROI and IV.

I mean something like this for example:

For all maiden claiming races ran in Belmont, Aqueduct and Saratoga evaluate all the horses that were coming from a 20-30 layoff, were stepping down from a MSW and switching from a sprint to route while the 1st and 2nd betting choice had more than 5 starts while were favored in at least one and there were at maximum two more horses had ever started before in a MSW.

You describe this factor in a DSL (domain specific language), submit it to the server and get back the corresponding ROI and IV.

Of course in top of this (simplified) factor you could be able to write your custom track variant estimate and apply it for the whole card or you can just use the standard one that is providing you with pace and final figures...

Although there is some learning curve associated with it following the DSL approach, it can certainly be simplified by a front end driven by a UI.

Is this something that might be interesting?

Ghostrider5 · 11-09-2010, 11:12 PM

Yes! Will check it out when you are ready to release the software.

ghostrider5

dukeofperl · 11-13-2010, 10:34 PM

Someday web apps will rule but we're not there yet.

If I'm developing a handicapping system, especially a heavyweight app that is mission critical there are 2 things I will surely leave out ... the web and relational databases. Both are simply too slow and one (the web) is unreliable compared to a standalone app running with flat file databases.

my 2c

Dave Schwartz · 11-16-2010, 02:47 PM

Delta Lover,

As a developer, all of my friends see the world as you do. When I told them a couple of years ago that my industry didn't want a web app, they told me "everyone wants a web app."

Simply not true. And especially not true of horse players who may want to take their laptop to the track or racebook where very few would have a connection.

There is also the fear that the company will fail and their software... the one they finally got to work right for them... would be obsolete and useless.

There is a programming concern... There is (potentially) too much data for you to query via the web, unless you are building your own mini-cloud. Imagine it is Saturday morning and you've got 200 users hitting your machine(s) each minute. You better not be doing anything too robust because if you are the queue will just get longer with each passing minute.

Finally, if you are successful, the EQB guys will come talk with you about data licensing. Be prepared.

Regards,
Dave Schwartz

It means we have to operate in the desktop world longer.

Dave

thorobasePA · 11-17-2010, 01:42 PM

Quote:

Originally Posted by Dave Schwartz

Delta Lover,

As a developer, all of my friends see the world as you do. When I told them a couple of years ago that my industry didn't want a web app, they told me "everyone wants a web app."

Simply not true. And especially not true of horse players who may want to take their laptop to the track or racebook where very few would have a connection.

There is also the fear that the company will fail and their software... the one they finally got to work right for them... would be obsolete and useless.

There is a programming concern... There is (potentially) too much data for you to query via the web, unless you are building your own mini-cloud. Imagine it is Saturday morning and you've got 200 users hitting your machine(s) each minute. You better not be doing anything too robust because if you are the queue will just get longer with each passing minute.

Finally, if you are successful, the EQB guys will come talk with you about data licensing. Be prepared.

Regards,
Dave Schwartz

It means we have to operate in the desktop world longer.

Dave

It is even just a Windows world for this market too?

I've explored some business models for this kind of thing (hosted) - Dave pretty much spells it out that while the tech might be sexy, it's not what this (albeit very small) market is looking for. They might spend 95% of their time at home with a good connection, but that 5% when they want to use it offline can be a dealbreaker. Local storage + sync approaches could work, but you are dealing with such huge data stores that this isn't easy at all.

There might be some interest around the "investigate-any-possible-angle" hosted approach, especially as it can be managed with scalable resources like EC2, when added to a marketplace to buy/sell computerized angles, but serious horse players still tend to be lone wolves.

Rich Rosa had a hosted system for a while - I don't think it worked out for him.

harntrox · 11-24-2010, 12:40 AM

Is DRF Formulator the leading web handicapping app?
How many people have went back to old Windows Formulator 4.x after waiting 10 minutes for a PDF for the WEB version ?

If DRF has experienced lackluster response, whats the point?

Dave Schwartz · 11-24-2010, 11:58 AM

Quote:

It is even just a Windows world for this market too?

To write handicapping software for the Mac is equivalent to writing software for the Radio Shack TRS-80 in comparison to Windows.

Okay, I am exaggerating. But writing software for a 7% market share is just not a good idea.

Quote:

Microsoft Windows has about an 88% share, Apple Mac OS 7% and Linux 1%

Source: Wikipedia

Dave

CBedo · 11-24-2010, 01:37 PM

Quote:

Originally Posted by Dave Schwartz

To write handicapping software for the Mac is equivalent to writing software for the Radio Shack TRS-80 in comparison to Windows.

Okay, I am exaggerating. But writing software for a 7% market share is just not a good idea.

Source: Wikipedia

Dave

I'd be shocked if in the horse racing market, mac users were anywhere close to 7%!

I use a mac (two actually), but I still have to run windows in a vm to run some horse racing software, and even though I develop ON the mac, in this environment, you still have to develop FOR windows.

DeltaLover · 11-24-2010, 06:03 PM

Quote:

Originally Posted by Dave Schwartz

Delta Lover,

As a developer, all of my friends see the world as you do. When I told them a couple of years ago that my industry didn't want a web app, they told me "everyone wants a web app."

Simply not true. And especially not true of horse players who may want to take their laptop to the track or racebook where very few would have a connection.

There is also the fear that the company will fail and their software... the one they finally got to work right for them... would be obsolete and useless.

There is a programming concern... There is (potentially) too much data for you to query via the web, unless you are building your own mini-cloud. Imagine it is Saturday morning and you've got 200 users hitting your machine(s) each minute. You better not be doing anything too robust because if you are the queue will just get longer with each passing minute.

Finally, if you are successful, the EQB guys will come talk with you about data licensing. Be prepared.

Regards,
Dave Schwartz

It means we have to operate in the desktop world longer.

Dave

Dave,

although i've spent years of desktop development using several frameworks (WinForms, MFC, Swing, C based Win API and many more) I have to admint that Web 2.0 based UI clients are the present and the future... Almost all new development is done based in some sort of of a Web platform such as Ajax, Silverlight, Flash, HTML5, jQuery...

As far as using the laptop at the track, this is exactly what I've been doing for several years using a wireless connection to the interent through an air-card which only costs me $60/month...

The problem I see with a stand alone application has to do with data.

In my case I am using a pretty extensive historical data base to generate my betting decisions so for my software to be of any value it will require from the user to have a large db (at least 30,000 races).

Overall, I think your points are valid and based in the poll's results, a stand alone application seems to be the way to go...

Dave Schwartz · 11-24-2010, 06:17 PM

Quote:

As far as using the laptop at the track, this is exactly what I've been doing for several years using a wireless connection to the interent through an air-card which only costs me $60/month...

I see it as you do. To me $60 per month is just another cost of doing business. Just like the RTN subscription I have that costs me $104 per month (net).

However, I assure you that there are a significant number of horse players who seriously do not like the idea of having to spend $60 per month to take their laptop to the track.

And even those who do not go to the track still want to go to Las Vegas or some other destination once or twice per year. For them the commitment is not looked at as $60 per month but rather $700 per year.

But don't take my word for it. Do another poll.

Dave