Betting with Ratings

In our last article we introduced the concept of ratings and how they are typically constructed. We also touched on how ratings are used as part of the form analysis process to help forecast the performance of each horse in an upcoming race. That forecast is often expressed as an expected rating for “today’s race.”

The different ways you can study the form using ratings to produce a forecast for today’s race could be the topic of an entire book itself and is beyond the scope of this article. Whatever approach you take though, once you have settled on an expected rating in today’s race for each horse, the key question is

“How do I take that rating and use it to make good betting decisions?”


Creating Your Own Prices

The traditional purpose of coming up with an expected rating for each horse in today’s race is to help create your own assessed price, which can be compared to the market to help make your betting decisions.

The technique I am providing here for turning your expected rating into a price can also be used without ratings. All you need to do is put your horses in ranked order and determine the number of lengths your assessment has each horse from your top selection if the race were to go exactly as you forecast.

If you are using a rating, then the correct values will depend on your rating scale and the number of points per length. Below is an example based on a rating scale of 2 points per length.


HorseForecast RatingLengths from Top
A950
B940.5
C931
E921.5
F912
G893
H883.5
I874
J845.5

Once you have something like the above table (either by using ratings or not), the next step is to look up the lengths each horse is from the top rated horse and apply a score according to the table below.

Lengths from Top RatedScore
01
0.10.95
0.250.85
0.50.75
0.750.63
10.51
1.50.37
20.24
2.50.18
30.12
3.50.085
40.06
50.03
60.015
7+0.007

If a horse is somewhere in between the numbers mentioned above, you can pick a score that is somewhere in between the two values.

Using the example above, below is the same table complete with the relevant score for each horse.

HorseForecast RatingLengths from TopScore
A9501
B940.50.75
C9310.51
E921.50.37
F9120.24
G8930.12
H883.50.09
I8740.06
J845.50.03
Total3.17

Notice how we have totalled the scores at the bottom. This is important for the next step.


Turning Your Scores into a Price

All you need to do here is take the score of each horse, divide it by the total (3.17) and then multiply by 100 to get the percentage change of winning.

For example, Horse A has a score of 1.00 divided by the total of 3.17 = 0.31545 multiplied by 100 = 31.55% winning chance.

To turn this percentage into a price you simply do 100 divided by the percentage number. So Horse A would be 100 / 31.55 = $3.17 assessed price.

You can also get the same result more efficiently via the formula:

1 / (score / total) = 1 / (1.00 / 3.17) = $3.17 assessed price

The following table shows the calculation for every horse in our example race.

HorseForecast RatingLengths from TopScoreChancePrice
A950131.60%$3.17
B940.50.7523.70%$4.22
C9310.5116.10%$6.21
E921.50.3711.70%$8.55
F9120.247.60%$13.19
G8930.123.80%$26.38
H883.50.092.70%$37.24
I8740.061.90%$52.75
J845.50.030.90%$105.50
Total3.17100.00%

Betting With Your Own Prices

Now that you have your own prices you may feel like you have unlocked the key to betting profits and it’s just a matter of time before you start backing good value winners and / or laying poor value losers.

A word of warning though…

One of the more dangerous beliefs a punter can hold is one which says:

“If I just price each race and then back the horses that are greater than my price or lay those much less than my price, I’ll make a profit.”

Just because you mark a horse at a certain price, that doesn’t make it the correct assessment of the horses winning chance, so betting entirely on that basis is very dangerous.

Race assessment and creating prices is a process of decision making in the face of incomplete information and general uncertainty. It involves making a number of assumptions along the way which may in fact be completely right, completely wrong or somewhere in between. If your assumptions are incorrect on one or more horses in a race then not only is your price on those horses wrong, but the price on every other horse is less correct as well.

With that in mind, please do not become too focused on the idea that a horse above your assessed price “must” be value and a horse below your assessed price “must” be poor value. Making good betting decisions requires a little bit more than that, which we’ll go into later.



What Matters Are Your Insights

Always remember that the quality of prices you can produce for a race is a function of how good your insight into that race is, not the technique used for creating the prices themselves.

In other words, how good your ability is to accurately rank chances and assign the relative margin between is what will determine how good your prices are. No sophisticated algorithm or other method to create your own prices can make up for shortfalls in your assessment of each horse.

The method I’ve outlined here is nothing magical, but does provide good structure to help you express your insight into a race and create a price output. You can even modify the score values in the table if you want to create more or less aggressive markets. Ultimately the price you end up with should feel right and make sense to you.


Adding Context to Your Price

Even if you have terrific insight into a race and spend a long time developing a price for each runner, it’s still dangerous to make betting decisions purely on that number alone. All prices have assumptions and differing levels of uncertainty attached to them.

For example, in Race A you mark a horse $3.00 and in doing so you’ve assumed the horse will make some improvement on its last start and handle a step up in distance. In race B you have a horse marked $3.00, but you’ve come up with that price by assuming the horse won’t make any improvement on last start and that you have allowed for some uncertainty about it handling the extra distance. If both horses are $3.30 in the market, are they both equally attractive betting prospects? If you are just using prices then both are technical overlays, so you would have to assume they are equally appealing.

However consider that in Race A you’ve made some positive assumptions about the horse improving and being effective up in distance and came up with $3.00.  Given that positive view, is $3.30 a compelling price to take? What if one of your assumptions are wrong? If you were slightly less positive in those assumptions, then you could have marked the horse $3.30 to $3.40 and then there is no real betting appeal at all.

In Race B you’ve made no positive assumptions and even allowed for some uncertainty about a step up in distance and have the horse at $3.00, with $3.30 available. That’s an entirely different betting scenario. If the horse does improve and / or handle the step up in distance then it could actually be better than a $3.00 chance.

What this highlights is how important it is to have some context around the prices you develop in each race… to understand the assumptions you’ve made to come up with that price.  When I price runners I add context by considering the evidence and assumptions applied to each horse, assigning it one of the following tags:

  • Short = I couldn’t price the horse any shorter than I have it now
  • Long = I couldn’t price the horse any longer than I have it now
  • Flat = There’s minimal uncertainty with the horse or we’ve taken a middle of the road view and could have the horse marginally shorter or longer if different assumptions were made.
  • Query = There’s a high degree of guesswork built into the price of the horse and the market is going to be a far better guide.

So in the above example of Race A. I might say the horse is $3.00 short. Considering the assumptions made I couldn’t possibly mark it any shorter than $3.00. A market price of $3.30 might be a technical overlay but in the context of how that price was developed, it’s hardly a compelling betting opportunity.

If on the other hand the horse was $2.50 in the market, then it is likely to be an attractive lay opportunity, because even after considering all of the uncertainty and assumptions, I can’t price the horse any shorther than $3.00

In the example of Race B, I might say the horse is $3.00 long. Considering the assumptions I’ve made I couldn’t possibly mark it any longer than $3.00. A market price of $3.30 in this instance has some betting appeal.

If I mark the horse at $3.00 flat, then that’s a reasonably balanced position and depending on how the remainder of the runners look, there might be a small edge there to play.

The last tag of query is for those cases where uncertainty around a horse is so high that you are purely guessing. Let’s say it’s an import runner that comes with good form but is first up and short of its best distance. Your price assessment is likely to be nothing more than a guess. If you make an initial assessment of $10 and later it is $3.00 in the betitng market then it’s foolish to take that price and all of your other prices as an accurate guide when you have clearly made a mistake on one of the other runners that is a main chance in the race.

With this context available you could adjust your betting market to have the query runner now $3.20 to $3.40 (allowing for the fact that on average all horses are marginally shorter in the market than their actual winning chance), then adjust your other prices and see how the betting scenario looks.


Using Prices to Make Betting Decisions

As I mentioned earlier, there is far more to a good betting decision than a numerical difference between assessed and market prices.

Do not fall into the trap of backing horses simply because they are technical overlays. For example, don’t take $3.00 about a horse just because you have it marked $2.70… if you know that the $2.70 has some uncertainty built in and you could have easily marked the horse $3.20 by making one or two different assumptions. That’s a decision you will regret if the horse loses and will most likely lose you money in the long-term.

Equally don’t lay a horse at $2.50 just because you have it marked $3.00 without understanding the assumptions you made in creating that price. If there is uncertainty about the horse and taking a different perspective could see you view $2.50 as somewhere close to the real winning chance, then don’t lay it.

The goal should be to use your prices as a tool to make betting decisions that make sense and feel right to you. Consider your price, the assumptions you’ve made and the market price on offer… then if a particular horse seems like good value then that’s the right scenario to bet. If the horse seems like particularly poor value then it’s the right scenario to considering laying the horse.

You should feel comfortable with your decisions regardless of the individual result and if your judgement is good enough, you will make money.

If you have the discipline to use your prices in this way, then they will become a key asset and ultimatley help you to increase your profits from both the back and lay side of the Exchange.


Related Articles

Introduction to Ratings

Ratings are numerical measures that reflect the quality of a horse’s past performances or other specific characteristics.

Read More

Barrier Positions

If you read or listen to any public commentary about an upcoming race, it will almost certainly involve numerous ...

Read More

Class Assessment

In thoroughbred racing, the term “class” is used in two different ways…

Read More