There comes a time when things start to come together and it now appears that my efforts have been worthwhile.

For six years I have been recording results of selections from the UniTAB web site using a particular range of the Ratings and form.

For most of that time I concentrated on TAB numbers 1 and 2 in handicap races and all TAB numbers for set weights, set weights with penalties and weight for age races. The primary reason for this was that I was collecting the data for the selections manually by entering the data from the web site into a spreadsheet.

Then last year I was able to make use of a web scraping program so I expanded the selections I was using to include all TAB numbers. I used the same form lines as before but added an extra rating number.

This of course added some extra race types (e.g.: maiden races and the occasional hurdle race) and I also included New Zealand races.

One of the problems was that I still must enter the race type and the number 1 in handicap races to my win and place dividends of each race. I must also convert each TAB particular codes which I devised some years ago.

So to make use of the power of the web scraping program I set a goal to collect all the data for 2009 while continuing to collect it daily. Then I started to analyse the results. When I had finished analysing the results I decided to collect the data for another month (December 2008) and determine if the rules which I had written worked for that month. The rules worked for that month so I was very confident that future predictions would be profitable.

What I have now is results for all race types (country maidens’ right through to group 1 weight for age races) for horses with the form lines below and UniTAB Ratings 96 to 100 inclusive.

All of the following information is old hat for many readers and users of my ‘systems’ but I include it for first time readers.

These are the form lines which I use and the range of Ratings numbers from the UniTAB web site:
- All last start winners
- Any horse second up from a spell (as shown in the UniTAB web site)
- Any horse which won its start before last and is not second up from a spell
- Any horse which ran second at its two most recent runs and is not second up from a spell

So the form lines can be any of the following:

From 1 December 2008 up until 26 March 2010 there have been 26,127 selections.

The overall loss is 4,295 units (16.4%). When I apply the rules I have developed there is a very good profit, easily exceeding 10%.

The innovative aspect of the next stage of this project involves the determination of whether a selection meets any of the rules. Because there are many potential bets each day the work to manually apply the rules to the selections is very time consuming and prone to error. The other problem is that many of the rules are track condition dependent, so to predict a winner using the track condition early in the day only to find that the track has changed some time during the afternoon is frustrating.

So I am getting a software program written which will automatically apply the rules for the day’s selections. That is not very innovative but I think the solution to my ‘track condition change problem’ is.

The answer is easy if you think about it – have the program check all the potential rules for each track condition – Fast, Good, Dead, Slow or Heavy.

So the output of the program will tell me if the selections for each race can be bet on the track condition which applies at the time.

The output will look something like this:



The programmer is my stepson Alexis. When he said that we should patent this work, my reply was that nobody would patent something like a horse racing system. However I got curious and started looking. If you type ‘horse racing system patent’ into Google you find US Patent 7645191.

So my curiosity meant that I read the patent and there is a step of the process which says something like ‘If the prize money of the race which the horse is racing in today is more than 30% greater than the race it competed in last time then do  something, otherwise do something else’

(Note: I wish GTX could do that).

So that got me to thinking again about prize money change, which I have written about before.

The next day was Wednesday 16 March 2010 and the major race anywhere that day was the Newcastle Newmarket Handicap. I took a look at the field the night before and instantly found the horse, Walking or Dancing that I believed would win the race. It was the only horse which was dropping in prize money; it had won at the track and at the same distance and was to be ridden by Glen Boss, one of this country’s top jockeys.

It was quoted in the early fixed price market at $8.

I couldn’t believe it. This horse should be much shorter so I sent an email to a colleague and his reply was that he would not bet in really difficult races like that and with so many starters (original field size 16).

Well you know what happened; the horse bounced straight to the front, stacked them up and then kicked away in the straight. It was getting tired but won reasonably comfortably.



Some of these horses are highly rated in the Winform Ratings and others are not. I look for these simply because they win races and sometimes at good prices. These situations are what is called ‘doing the form’, rather than relying on mechanical systems. The first part of this article is about a mechanical system and sometimes after further examining the field and prices (doing the form) I find it is not likely that the selections from these systems will win but in the long term they make money. So sometimes I will look at a field where there is a ‘mechanical system’ selection and see if there is an ‘obvious’ danger or more likely winner.

Here are some more examples of what to look for. Of course they don’t always work but the idea is that even though the horses might have what looks on paper to be poor form, there is generally a reason for it and even in spite of the poor form they win.

I have highlighted in RED the horses for consideration. The only reason to consider them is that they are dropping significantly in prize money from their previous race.