19th December 2016
Hi ,
Is it possible to know, with any accuracy, what a horses chances of winning is?
The answer is yes and... no.
Okay, that probably didn't help clear it up very much. But I hope what follows will!
I'm going to say at the beginning that this can take some getting your head around. In fact, there are even profitable bettors who don’t understand this concept. Which means... that you don’t need to understand it to be profitable. But if you can understand it then I promise it's going to push you leaps and bounds ahead in your betting.
It's widely recognised that the odds just before the race begins are a very accurate representation of a horse’s chance of winning the race. We can test this theory by looking back at past races and seeing how often the 33% horses won and how often the 12% horses won etc…. and it does hold up that these odds provide an accurate estimate of a horses chances of winning. e.g. the horses whose odds indicate they should win around 33% of the time, actually do win around 33% of the time.
Actually, that's not quite accurate. And when we're talking about this we must be accurate. What I should have said is...
‘The odds just before a race begins are a very accurate representation of the horse’s chance of winning the race when looked at on average over the long-term.’
What does this mean?
Well, this is quite controversial but there's only ever one correct odds line in a horse race. This is simply...
The winner has 100% chance of winning the race and every other horse has a 0% chance of winning the race.
Think about that for a few moments. There is absolutely no other possible outcome and so the only completely accurate odds line would be one that represented this.
When we create an odds line what we're doing is trying to find the average chance of a large group of horses winning over a large number of races. By this I mean that we take all horses who we have assigned a 30% chance of winning a race over the course of the year and ideally we want these horses to win an average of 30% of the time.
This is crucial, and just to make sure it's clear I'm going to explain it in another way as well.
If we want to create an accurate odds line for just one race, and we're never going to look at another race, then we have to give one horse 100% and all the others 0%.
Obviously the horse we give 100% has to be the winner for our odds line to be accurate.
But, if we decide that we're going to be betting on one race every day and more importantly, over the course of the year we want our odds lines to be accurate, then we can create an odds line that represents what you recognise as a normal odds line!
This is because over the course of the year we may give 50 horse’s odds of 5.00 or a 20% chance of winning the race. If from the 50 horses 10 of them win then we know that for this group of horses we have been giving them an accurate estimate of their chance of winning the race. We have created an accurate odds estimate for a group of runners, we haven't created an accurate estimate for each of those runners individually.
It's quite a hard concept to explain and I hope that this makes sense.
Now that we understand this, what we need to make a profit is an accurate odds line which uses information in a different way to the public odds.
We've already confirmed that the public odds, just before the race, represent an accurate estimate of a the horse’s we assign these odds to chances of winning in the long-term.
Now, just to prove that there are multiple "accurate" odds lines but not all are profitable I want to show you an example.
The simplest way to create an odds line is using a 10 horse race. We give all the runners in the race a 10% chance of winning. If you do this for every ten horse race, you'll find that historically the odds line created is as accurate as the public one. In other words, the horses will win an average of 10% of the time.
Can you make a profit from it? Of course not, no more than you can from the public odds.
In order to make a profit you need to find information, or a way of using the information, available that is different to everyone else.
Let me take you through another example that demonstrates how you can make a profit with an accurate odds line that is different the general publics.
If you have a group of horses that you bet on through the course of the year, and they win 30% of the time, then when the public puts one of the horses at a lower chance of winning than 30% you know that you have a value bet.
Does this mean that my probability is better than the public one? No it doesn’t.
We know that the public odds are accurate in the long-term, we also know that ours are accurate in the long term. But the difference is that over the long term not all of our 30% horses will be 30% in the public line, some will be higher and some will be lower.
If we take all the horses with a 30% chance winning from our odds line and compare them to the 30% chance horses from the public odds, the runners will be different!
Sometimes we will get value on our selections when the public odds think they should be higher, other times we won't. But ultimately we know that the horses we put in the 30% group win 30% of the time.
This is a topic that is seldom discussed but, as I mentioned before, if you can get your head around it then you are going to be leaps and bounds ahead of other bettors.
Best Wishes,
Michael Wilding
