This is the final part of the series on how to Build A Killer Betting System and, before we get into it, let’s have a quick recap on the steps you’ve taken so far.
In part one you looked at the different types of factors that you need to measure to understand a horses’s ability in a race. In part two you broke down the race conditions of a race to determine the type of race that you’re going to be building your betting system in. Then last week, in part three, you took your ratings and converted them into Impact Values so that every rating was being measured on the same level.
That means that today you’re going to discover how to use this information to build your final betting system!
I’m going to share with you a process that very few people use when they build betting systems. But, if you want to build profitable ones, then this is the only way to approach it.
To be able to maximise this approach, I first need to show you why the normal approach is not very effective.
When you build a normal betting system you’ll start with an idea and then look to see how it performs. If it doesn’t make a profit, which it almost certainly won’t on the first attempt, you add a new filter (e.g. number of runners) and then check to see how this smaller sample from your data performs. If it improves the performance you keep this filter, if it doesn’t then you remove it and repeat the process.
The process looks something like this:
- Start with an idea
- Test idea to see if profitable
- Add a filter and see if it improves profit
- If improvement keep filter, if not remove and try another
- Go back to step 3
I’m sure that you’ve seen, and probably done, this many times before. Everyone does it.
To understand why this isn’t a good way to build a betting system, you need to look at exactly what is happening during this process.
Step 1 makes perfect sense, as does step 2. It may look like step 3 to 5 also make sense, but what’s actually happening here is you’re trying to find a set of conditions that produce a profit.
In other words… you’re back fitting heavily!
It’s all to do with mindset. Your mindset is:
I need to find the race conditions that produce the most profit
From today onwards, I want you to change your thinking to:
I need to find which horses run best in these races
It’s a subtle shift in thinking that’s going to make a huge difference to your bottom line. You already have the race conditions your investigating, you don’t need to change these to try and find a profit. What you need to do is find the “types of runners” that excel in these conditions.
There are horses that excel in all race conditions, your job is to find them. And, to do this, you start by looking for the ones that DON’T perform well under the race conditions you’ve chosen.
The purpose of build a betting system is to find selections very quickly each day, unlike form reading which can take an hour or more per race. So the question is, why does form reading take so long?
Because there are a lot of runners in each race.
Even if you spend only five minutes on each runner then a ten runner race is going to take you 50 minutes!
But a large proportion of those runners are not going to be competitive in the race, because they aren’t suited to the conditions. So what’s the point in even considering them?
None.
Which is why your first step in building a system needs to be…
To find the runners that don’t perform well under the race conditions you’ve chosen.
Sure, by removing these horses we’ll miss the odd winner. But I can guarantee you that these runners won’t make a profit, and are certainly not worth spending your valuable time on.
We’re going to call these runners the Eliminations. They’re the horses that we know aren’t going to perform well, and you can do this using your Impact Values. Look for ratings with Impact Values of 0.50 or less. This means that the horses are winning half as often, or less, than they would be expected to win.
These are the horses that you want to remove.
Once you’ve done this, you’re going to be left with horses that have some level of ability in these race conditions and, from these, we want to separate the ones that have the strongest preference for racing under todays conditions.
Notice how we haven’t started to look at whether they make a profit yet. That’s because at the moment this isn’t what we’re interested in. What we’re interested in is removing horses that perform poorly, and finding the horses that perform well under the current race conditions.
So, go out and mark all the horses that have an Impact Value of 1.10 or higher. These are the runners that are winning at least 10% more than they would be expected to win. In other words they’re strong contenders.
Now you’ve got three groups of horses. They’re…
- The horses that are expected to run poorly and are eliminated
- The horses that are average and don’t fit into eliminations or contenders
- The horses that are expected to run well and are contenders
While it’s possible to use the average group, it’s an advanced method and not one that you want to start with. What you want to do is to use the contender group, after all these are the horses that are expected to run the best.
And now that you’ve found them, it’s time to work out how best to bet them.
It won’t always be profitable to bet the strongest contender, this horse may be recognised as being the best in the field by the general public and so the odds won’t offer any value. This doesn’t mean they aren’t strong runners though, which is why they’re in your contender group.
However, the second or third best contender may not be recognised as being as good by the general public, and so may be profitable to back or place each-way bets on.
If the top contender is a horse that has such a strong chance you don’t feel you can bet in the race without betting on them, then you need to look at place betting or using them as a saver bet where you make your money back if they win.
You now have in your hands a roadmap to finding the best horses in the race, and you know that. Your investigative work now needs to focus around finding the best way to bet them to make maximum profit.
You’re no longer trying to find the race conditions and runners that make the most profit all at the same time. This results in massively back-fitting unless you’re very careful.
What you’ve done is separate out the two main components of system building. First of all you’re looking for the horses that are going to be the strongest in the race, and only then are you looking at how you can bet them to make a profit.
Although I’m not a big fan of hard and rigid ‘systems’ I have used similar approaches to develop ‘methods’ for quickly identifying possible value bets. I prefer to have an ‘angle’ into a race before I consider the race as a possible betting medium in the first place. This is in addition to the race conditions that you detailed in the blog. There are many ‘angles’ that can be identified as giving you an advantage before you even start to build your system and most of the ones that I use have come about from personal observation and then a detailed analysis to identify which types of race favours that particular angle. Race conditions are then also considered as a filter to fine tune the method.
If you take the ‘angle’ approach you will end up with several bespoke systems rather than a one size fits all approach which some days will mean no bets at all, but helps to keep your discipline in check. Think of it like a cabinet maker thinks of his tool box in that it’s all about selecting the right tool for the right job. You wouldn’t expect a system designed for sprinters on the sand to cut the ice for a 3 mile slog in the mud. Plus…. it’s nice to get the odd day where I can have a day away from the desk. Always paper trade the ‘angle’ first to validate your reasoning. Ideally, you need to see a minimum of a small profit (but preferably a large one), over at least one season from following it blindly. Then add your conditions by reasoned logic rather than looking at the results. If it improves the performance over the same period then you are on to something. Keep monitoring it for possible improvement, but only by reasoned and tested methodology.
Enjoyed the blog Michael and learned about ‘impact value’ ratings too. Every day’s a school day in this game!
Tim.
Hi Tim, thank you very much for your comment. It sounds like you have a great method of building your own systems using the angle and then going on to use the process I’ve outlined. I will look forward to hearing how your portfolio performs 🙂
I bought a book called ‘The Punter’s Revenge’ written by a guy called Tony Drapkin about 30 years ago and have been hooked ever since. In the book he applies conventional statistical analysis methods to horse racing data and attempts to make a profit. I used a lot of his techniques when I first started with marginal success but found that by using it to test angles into races it could be used to its best potential.
I did consider the idea of setting up a web site but busy enough as it is and often wonder how folks such as yourself get the time to do it, and still get time to put your own bets on….. never mind developing and testing new ideas.
I love the Punters Revenge, I am in occasional contact with Tony as well 🙂 Still an excellent read and Tony’s models are still working very well to.
I spoke to him on the phone just the one time about 15 years ago. Can’t remember why I called him now but I remember him to be a really pleasant and informative guy. I can recommend the book to anyone starting out but I imagine it’s been out of print for sometime now.
Like I said earlier….. I found the broad brush approach of trying to formulate probability tables to be very time consuming and a little too marginal in the long term with profits typically between 12 – 20% although this was in the days before Betfair and BOG offers and all the stuff we take for granted now a days.
Since then I’ve improved my action and now my best work (quality jumps handicaps) has returned 98% ROI to BFSP). I’ve been working this angle for about 3 years now. At the other end are Listed flat events where I’m currently around 42% ROI but generally speaking, I do better over jumps than I do on the flat. Average strike rate was around 15% last time I looked. So….. not only do you not get bets every day, you certainly don’t get winners every day either, but I prefer to play the long game myself as it can take a long time to recover if you take a knock playing the short end of the market.
The book is out of print but chapters are available online somewhere, I’ll try to find where. Those are superb ROI’s. I personally prefer to work to a much lower ROI such as 5% but have a very high turnover with bets in most races.
The odds that I get about the quality handicap (jumps) runners are very often into double figures with winners at 20.0+ not uncommon. The angle here is really about looking for a horse in a competitive race that has a reasonable chance, AND that has not been exposed to the market, either by the media or the racing post forecast. If the likes of Timeform or Spotlight rate it’s chances OR it is prominent in the betting forecast and you are getting a big price, (=> than the number of runners in the field), the chances are it will lose. In races such as these it pays to focus your attention on the rest of the field. You would be surprised just how many times good horses can go unnoticed by the racing press in good quality jumps handicaps.