The Science of Impact Values

What are Impact Values? Impact values are statistically derived figures that allow you to assess whether a particular factor in horse racing sees horses win more often than expected, or less often than expected.

They were invented by an American called Fred Davis and published for the first time in 1974. Although they are used throughout America as significant figures, they are hardly ever used in the UK. Nick Mordin, a well known racing author, used to be the only person who talked about the significance of using impact values.

First of all let us see how we calculate these figures, before we go on to look at how we can use them to help us make a profit. The word statistical may have got you a little worried however, as you are about to see, the calculation of these figures is actually very simple. The only requirement is a database of past races as Impact Values can only be calculated using the information from previous races.

The formula to use is:

(number of winners with the factor / all the winners in the sample of races)

Divided by

(number of starters with the factor / all the starters in the sample of races)

As you can see, this is a very easy calculation. I will now work one through with you. We shall look at the Impact Value of horses starting from Draw 1 at Ascot race course. We have 105 winners from Draw 1 from 1513 winners in our sample of races. There were 989 horses that started from Draw 1 from 17000 starters in our sample of races. This is all the information we need.

To work out the Impact Value for Draw 1 at Ascot we now perform the calculation above which, when the figures are put in, looks like this:

(105 / 1513) = 0.0693

Divided by

(989 / 17000) = 0.0581

= 1.19

The Impact Value is therefore 1.19. This means that a horse that is drawn in Stall 1 at Ascot is 0.19 times more likely to win than a horse that is not drawn from Stall 1.

To calculate how much more likely this is as a percentage you do the following sum:

Impact Value (1.19) – 1 = 0.19

0.19 x 100 = 19%

Again clearly and simply, you can see that this means that our horse in Stall 1 is 19% more likely to win the race than a horse that is not drawn in Stall 1. I am hoping that you can already see how useful this figure can be.

Impact Value however, used in such a general way, it is not much use. If we were to narrow this down to include a specific going and distance then the figure would start to become much more useful. The Impact Value should also not be used on its own. This is because, although a horse may be 19% more likely to win a race than the other horses because it is starting from stall 1, it does not mean that everybody else does not also know this information and the odds are quite likely to have been adjusted because of it.

Therefore you should also take a note of the ROI. Betting the horses in this example would have lost you 15.75% of your investment. So then how can we make a profit using these figures? Well, we can do this in several ways.

I. We can consider the ROI and look for horses that have a positive ROI as well as a good Impact Value. The problem with doing this, is that it is going to be very hard to find these horses;

II. We could combine different positive impact values to create a rating. This is good in principle but you would need to make sure that you then trial your ratings to confirm that you have got both a positive ROI as well;

III. We can use Impact Values is to assess this race by race.

I believe this last option to be the most effective way of using Impact Value. What I mean by this is, you want to look at a specific set of circumstances, to try and establish what type of horse wins that race, or is most likely to win that race. For example, let us look at races at Ascot with 8 to 12 runners. Look at horses that were favourites last time out. To calculate the Impact Value we run it through our database and see that there were 409 starters with 44 winners. 67 of those had the additional factor of being favourites last time they were out, of which 7 won.

Our calculation for the Impact Value then becomes:
(7 / 44) = 0.159

Divided by

(67 / 409) = 0.1638

= 0.97
So we know that last time out favourites are not a statistically important factor in races at Ascot with 8 to 12 runners. Even this is quite general. I would suggest that you start with a track you like, and focus on the type of races that are run there regularly. Separate them into distances – plus or minus 5 furlongs, going types, handicaps, race types.

You can then look at the factors that you want to get the Impact Values for, for example: speed ratings, form ratings, prize money won, win percentage at course etc….. Calculate a selection of these to see which have the highest impact value and you will start to see which types of horses are most likely to win a particular race. This will take some time, but will be very beneficial to your betting.

Another way of using Impact Values is to tell you information about a horse. For example fitness; you could select an age and sex of horse, and run through different factors that may affect the fitness of the horse. When you have a number of them, you can use this information to tell you which horses of this age and sex in a particular race are actually likely to be in good race fitness.

There are many ways in which Impact Values can be used, and it is up to you to start exploring. However, they all require time and patience to make them a worthwhile factor adding value to your betting. The methods in this article only represent a few of the possible uses for Impact Values and are intended to give you a head start, with time you will come up with more of your own.