System, Let’s Make A Winning One! – The Final

It is the third week of our system build and things have been going well. If you haven’t had a chance to read what we did in week 1 and week 2 then please have a look and send in your thoughts.

Last week we ran a poll to see whether you would prefer backing favourites or non-favourites. 68% of you preferred to bet on non-favourites and 32% chose favourites. This told me that while the preference was for non-favourites there was also some consideration needed for the favourites in races. The benefits of using favourites is that the strike rate is increased but for the following method I ultimately chose to go with the preference of our readers and look at non-favourites as this also helped to increase the profit and ROI which were the top results in our poll for week 1.

That is enough chatter for now, it is time to carry on with making our system. This week I decided to use a piece of software available to everyone called Proform Professional (they have offered a discount of 25% to Race Advisor readers who use the code PRORAT25) to create the system. I did this because I received an email asking if it was possible to create profitable systems without your own database. Of course it is I replied and I thought I should backup my reply by using it here.

If you remember from last week we are concentrating on class 6 and 7 races on all-weather tracks. Although I am going to have to cut out a lot of the process (we don’t want this to turn into a thesis) there are going to be quite a few tables in this article but I think it is important to show you as much of the process that I went through to get to this stage.

Straight away I decided to remove all favourites and second favourites as well as any joint 1^st and 2^nd favourites. The first extra piece of information I wanted to look at was to see how the results looked if I only looked at horses who were beaten in their last race by 0.5 to 1.5 lengths. I know you are thinking that this means they weren’t winners! By not looking for horses who won but horses that ran a good race, we are looking for runners that may not be so heavily bet in their next race (in other words value horses) but have still showed merit in their performance.

This gave us the results below:

This didn’t produce the best results to SP even though we made a profit to the Betfair SP Estimate (a very useful feature). A logical way to try and create a better profit is to only consider horses whose last race was on the same race type as the race we are looking at. If we narrow down to only horses whose last race was on the all-weather we get the results below.

Well a pretty good start, we are already profitable in two out of four years and that is to SP. We are profitable to the Betfair SP Estimate in every year. On top of that we have a nice number of runners each year and didn’t lose too many by reducing just to all-weather horses in the last race. The strike rate is a bit low and of course we haven’t tested it in the year 2009 to see if it holds up.

However we can do better than that, this was just our first attempt! It is all very good that we are looking at horses within a certain distance of the winner but what happens if it was a tight 10 runner race and eight of the runners were within our distance. In this situation we don’t want to be considering them all, we only want the ones who had the slight edge over the others. With this in mind I added a rule to only consider horses that had finished in fourth position or better in their last race as well.

There are now only losses in the year 2007 to SP and that is just one unit and the strike rate has improved very slightly. Things are looking good so far but we want to be profitable in all years to SP, if we can then beat SP slightly by taking Betfair SP odds or early odds at the bookmakers with our selections we shall be making a good profit. By searching through the results I found out that our selections performed much worse in races with two year olds.

I like to have a valid reason for putting any rule into a system; if it doesn’t make sense to me then I don’t use it. Over the years I have discovered that the rules that don’t make sense tend not to work in the future. It would make sense that races with two year olds are going to be slightly less predictable because we have horses with little form who may be improving rapidly. These horses are going to be hard to judge and so by sticking to races with a minimum age of three we can increase our profitability.

There we have it, we are now profitable in every year to SP. Okay I know that 2005 and 2007 weren’t exactly hugely profitable but at least they were in profit. If you were betting to Betfair SP then you would have made a very good profit every single year. The strike rate hasn’t increased much and so we are betting on higher priced horses (average odds around 11.00 in fact) but they are making us a profit. The next step is to look at our trial year and see if the system holds up for 2009.

You can see above that the system holds up in 2009 with 2.75 units profit to SP and 60.29 to Betfair SP Estimate. 2010 is not profitable to SP yet but we are less than half way through the expected number of bets in the year and is only very slightly down so we can expect that to turn around. Of course to Betfair SP Estimate it is just profitable.

Above is a profitable system using just seven simple rules and information that can be found in any paper or website. It was created using the commercially available Proform. Of course it took more trial and error than I showed here but I hope I have shown the general method.

Will it continue to be profitable? A question I can’t answer but there is no reason it shouldn’t continue to be. Can it be made better? Of course, but I have taken you this far and you can now continue to improve it through your own analysis or daily handicapping to finalise a system to be added to your portfolio. That way we may preserve the odds on offer for these selections.