Lesson 1 - Designing your application

[Datapunter]

Lesson 1 ( Word document )

Punters Lounge JAVA programming course.

Lesson 1 - Designing your application

First thing to learn about programming is the preparation. This means defining what you want to make. That is done in 2 parts:

1.1 The functional design.

A functional design is a description of what it is you want your program or application to do.
Like this:

The application must retrieve the available odds on premiership football matches from different bookmakers. It must then find the best odds, calculate the overround and present that info so it can be reviewed at a glance.

1.2 The technical design.

The technical design is where you collect all the technical information you need to build your application. In our case we need to find out what bookmakers can be used. We need the formula to calculate overround. We need to design how we want to view the info.

Bookmakers:
We will be looking at bookmakers that run their site using HTML. These apply to:
www.victorchandler.com, www.pinnaclesports.com, www.paddypower.com, www.gamebookers.com, www.eurobet.com, www.coral.co.uk, www.capitalsports.com .

It is not possible on:
www.betfair.com , www.sportingoptions.co.uk , their site is not build with HTML only.

Note: as I'm writing this I've only had a quick look at those sites only to determine if HTML is used.

Overround:

To calculate overround: divide 1 by the odds and add the outcomes. Anything above 1 is the amount you lose betting all players. So a 2 player match with odds of: 1.35 and 3.5 has an overround of 2%. (betting on both will lose you 2% of the amount staked)

( 1 / 1,3 ) + ( 1 / 3,5 ) = 1.02

Or for a football match with the favourite at 2.00, the underdog at 3.3 and the draw at 3.6 its:

( 1 / 2.00 ) + ( 1 / 3.3 ) + ( 1 / 3.6 ) = 1.08 being 8%

Obviously finding an overround of less that 100% gives an opportunity of making a sure profit betting all outcomes.

Viewing the info:

What we want is a simple list of the odds at the bookmakers, then one line stating the best odds available calculating the overround each time. Like this:

Arsenal - Liverpool - Draw
1.75 - 6.0 - 3.35 = 3.7% at Bet365
1.85 - 5.0 - 3.30 = 4.3% at Ladbrokes
1.65 - 5.5 - 3.40 = 8.1% at WillHill
Best odds
1.85 Ladbrokes 6.0 Bet365 3.4 WillHill = 0.1% overround



SCHOLAR class: now I don't need to tell you the importance of spending sufficient time on your design, you should know by experience how much time it saves you along the way.


ROOKIE class: you may get the impression that it's a bit too much, lets just get on with it. Well in this case you are right, however as you start writing bigger programs you will need to spend more time on the design. Believe me, it will save you a whole lot of time along the way.
I rest my case.



Assignment 1.1

If you are content by following my lead as described above that's fine. But I invite you to think about applications of your own. Write some functional designs. What would you like a program for? And don't be afraid to use your imagination.

[Datapunter]

Any question on how to go about designing an application stick them in here.

When you are ready to start work on an application of your own start a new personal thread. Something like:
DP's project: Horse Racing Rating System

[kompressaur]

None at all Data

How can we tell which sites use HTML?

And if they don't, can we still use them in some way?

---

Good question,
it's not if they use HTML, it's if they use HTML in such a way that the information we are interested in is directly accessible.

Nearly all sites use HTML, some only HTML but most a combination of HTML, Javascript, background databases and possibly site-specific software. Depending on the complexity of the site accessing information can be easy or very difficult. For this course we will look at sites where the information can be accessed directly inside HTML, possibly after going from page to page a couple of times, but no more difficult than that.

How to find out ? Lets use this very page to try.
( i'm using Windows 2000 at work, may be different on other versions ! )

First i left-click in the area i'm interested in to make sure the current focus of the browser is on that area.
Then, keeping the mouse in that area, i right-click and select "Source" or "View source".
This opens a view, in my case using Notepad, of the source code of the page.
( you could save this info in a file if you wish )
On one of the top lines there should be a reference to the name HTML.
Then i do a search of the info i want like "Good question".
( using the search function of Notepad )

If it can be found i know the info is directly accessible inside HTML, and can be extracted fairly easily.

If it cannot be found then some other mechanism than basic HTML is used. In this case it is required to use other techniques to get to the info. Those will unfortunately fall outside the scope of this course.

We will go through the details of all of this in lesson 8 or 9.

So to answer your question, can we get at the info if it is not directly accessible in HTML ? Yes, the info can most likely still be accessed but there is no general method, it will have to be examined case by case.

Datapunter,

Being a Rookie i am tempted just to go with the standard app, but if not this is what i would like to do.

Taking the Sporting Life website they post Horse racing results up daily with in running comments for each runner. Each race has a seperate web address and i have checked them out and they appear to be standard HTML. Do you think for a rookie it would be too ambitious to create a program that extracted these results daily into a database that was searchable via horse name?

---

had a quick look, will take a bit of time to write but in itself its pretty straight foreward, would even dare to say like taking candy from a baby

So that's a No not too ambitious

Couldn't Javascript be parsed just as HTML is?

[Datapunter]

Short answer : Yes

But there are 2 levels to Javascript. Javascript is a programming language that runs inside a website. In most cases you can see the Javascript code on the webpage and yes you could parse that just like HTML. After all its just a series of characters forming words and lines.

The second level however is what the Javascript parts do once a website is accessed by a user. This you obvioulsly cannot parse because it is different for each user and each session.

You could parse the Javascript code, and i would say decode not parse, what it actually does. After all Javascript is bound to a set of fixed rules and can therefore be de-coded. In theory that is, i have not done this yet.

It's my impression, i could be wrong, that this would involve an enormous amount of work. The question then is : is it still worth it ? I would tend to look for alternatives and only go into de-coding Javascript if there is no other option.

I think there may be another dimension to your question. Maybe if you have an example i can make it more clear and specific.

[Datapunter]

empty post, done to re-order all the sticky topics.