Return on Investment or ROI%

This is a very useful stat to know as it shows in percentage terms how much profit or loss has been made. Using the building society analogy again this would be the equivalent of the interest rate.

If a race has a favourite's ROI of 20% then you would have made a 20% profit (or get back £1.20 for every pound staked). If the ROI% is -5% then you would have made a 5% loss (or get back 95p for every pound staked).

The formula for calculating ROI is:

ROI% = Total Profit / Total Staked * 100% (at level stakes)

e.g. if a series of 135 bets (all at a level £1 stake) returned a profit of £8.45 then

ROI% = £8.45 / £135 * 100% = 6.3%

The ROI is useful for identifying unique stats about a runner. This could be a sire's first time out stat, a trainers record in claiming stakes races, or a jockeys record when he teams up for a particular trainer.

Again, the stat can be used to show good and bad conditions. Positive ROI's are good whilst negative ROI's are bad.

Just as with the level stakes profit stat the ROI stat can be flawed if there are a few big priced winners. If a jockey has a rare 100/1 winner his ROI could look impressive but overall he may not be that profitable.

This flaw can also work in reverse. If a trainer has many outsiders which rarely win then the ROI could be quite negative. If you filter out the big priced runners you may find that the trainer is actually profitable to follow.

John Dunlop has a weak ROI with his 2yo horses. Backing all his 2yos in the past 10 years on turf has returned a loss of 38%. But if you dig deeper, and check out the prices of all his 2yos, you will find that he rarely wins with his outsiders whilst his fancied 2yos do not return as much of a loss.

The table shows that Dunlop's 2yo return a loss of 17% when they are in the top third of the betting, whilst his unfancied runners return a loss of 80%. The stats show that Dunlop had 445 2yos which went off at prices of 20/1 or more. Only 4 of them won and it is this group which make up the bulk of the loss.

This shows that you have to be careful when using the ROI as a guide to profitability. Whilst an ROI stat for a favourite is pretty reliable you should always check to see if there are lot of unfancied outsiders are skewing a stat.

Impact Value

An impact value is a statistical index which shows if a particular group win more than their fair share of races. The technique was first used in the 70's by Fred Davis and has been one of the most widely used stats in the USA but rarely used in the UK.

You won't find impact value stats in daily newspapers, or mentioned by tv racing pundits - they consider stats such as this to be too complicated for their readers and viewers.

The impact value is not a complicated stat at all. In fact, it is easy to calculate and is one of the most powerful stats to know. All you have to look out for is figures which are greater or lower than the magic 1.00 figure. Any figure above 1.00 shows that the group is winning more than their fair share of races whilst lower than 1.00 and the group is winning less than their fair share of races.

A good use of impact values is for determining which group are best in certain race conditions. e.g. In 2yo Maiden Stakes which gender are best, or is a first time out 2yo better than one which has race previously.

In this example colts have an impact value of 1.19. This means that they win more than their fair share of races whilst fillies and geldings win less than their fair share of races. Another way to look at this is that a colt is twice as likely to win a 2yo Maiden Stakes race than a gelding (1.19 / 0.59 = 2.0).

The impact value is a powerful stat because it takes the number of starters into consideration. It is much more informative than the strike rate or number of wins.

To calculate the impact value you need to know the number of winners and runners for a group, and the total number of winners and runners for the type of race being analysed.

The following example calculates the impact values for different weight ranges in 2yo handicaps (nursery races) on the turf.

The formula for calculating the impact value is:

I.V. = (group percentage of winners) / (group percentage of runners)

For the highest weight group (9st 4lbs+) the formula looks like:

(313 / 1139) / (2307 / 13212) = 1.57

The middle weight group impact value is:

(509 / 1139) / (5601 / 13212) = 1.05

Finally the bottom weight group impact value is:

(317 / 1139) / (5304 / 13212) = 0.69

Just as with the ROI% stat the impact value can be flawed if there is a price bias. In some cases it is worth checking the impact value on fancied runners, or those priced in single figures, just to check that a lot of outsiders are not skewing the figures.

Article created 01-Nov-08. Stats may have changed since. Data analysed from Nov-98 to Nov-08