American writer Fred Davis is somebody that serious  punters should be aware of. In 1974 he wrote a book called "Percentages  and Probabilities" where he discussed a new way of analyzing horse  racing information via the statistical approach of “Impact Values”.  Since then, especially in America, Impact Values are used by private  handicappers and many serious punters in an attempt to gain an edge.

 

The use of Impact Values seems less common in the  UK – the whole statistical approach is still not “in vogue” as the die  hard traditions of form study reigns supreme. Things are changing  slowly, but a statistical approach is still frowned upon by many.

 

However, I believe it is a very important method in  terms of appreciating how strong a statistic actually is. Essentially  the operation is a fairly simple mathematical division sum, but it may  be easier to give an example:

 

Example – looking at draw bias at Chester.  In 10 runner plus 5f handicaps over a recent 7 year period, we find  that the 3 lowest drawn horses won 24 of the 47 races – in percentages  terms that equates to 51.1%. The highest 3 drawn horses won just once  equating to a success rate of 2.1%. Hence, if we want to work out the  Impact Value we divide 51.1% by 2.1% which gives us a figure of 24.3.  Hence, the Impact Value is 24.3, and what it means is that horses drawn  in the bottom 3 stalls are 24.3 times more likely to win than those  drawn in the top 3 stalls. Quite a strong statistical indicator that  one!

 

Not all Impact Values are this big / potent.  However, working out Impact Values gives you a clearer picture than  simply looking at say a wins to runs ratio (strike rate). Let us look  at one of the top flat trainer’s record with older horses in all age  handicaps – again over a recent 7 year period:

 

 

On first glance, what is your first impression?  Essentially, you should be thinking that this is a fairly good record –  a 1 win in 6 strike rate in handicaps suggests it is more a positive  record than negative one. However, how good is this performance in  reality? Well one thing we need to do is compare this trainer’s  individual record with the record of all trainers in the same race  type. Now the average strike rate of all trainers in all age handicaps  is around 7.5%. Hence to calculate the Impact Value we need to divide  16.8% by 7.5% - this gives us an Impact Value of 2.2. Hence, this  trainer outperforms the average trainer by over 2 to 1 in all age  handicaps. Hence, knowing this Impact Value means that this trainer  requires further investigation and research on his performances in all  age handicaps.

 

However, despite an Impact Value being extremely  useful, it is important that you not only consider the Impact Value,  but also the Return on Investment (ROI). ROI is calculated by dividing  profits or losses by the number of runs. Hence if 100 bets produced a  profit of £24.00, then the ROI is +24%; if 100 bets produced a loss of  £15.00, then the ROI would be -15%.

 

If the trainer in the above example had an ROI of  +30% then it indicates that not only are these trainer’s horses winning  more than they should, they are also winning at prices that are bigger  than perhaps they should be. A combination of positive Impact Values  and positive ROIs is the ideal scenario and those horses should offer  us, the punter, a good chance of making profits. A positive Impact  Value coupled with a negative ROI gives us a dilemma. Clearly the  horses are winning more than they should, but unfortunately the prices  on the winning horses have not big enough to produce profits.

 

Sounds easy then – find a positive Impact Value  coupled with a positive ROI and we have the way to the Holy Grail of  successful betting. Unfortunately again, just having those two figures  is not enough to have full confidence in our statistics. Let me explain  further. Below is a list of hypothetical trainers and their imaginary  performances in maiden races focusing on Impact Values and ROIs.

 

 

Looking at the above list, which trainer or  trainers would you choose to follow? Clearly Smith looks the best one  to follow – a success rate of over 5 times the norm and percentage  profits (ROI) of 200%. Jones looks very good also with a success rate  roughly 3 times the norm and percentage profits (ROI) of over 70%.  However, if we add some other key stats to the table then it becomes  very clear that we need to rethink:

 

 

Smith and Jones (not the comedy duo I hasten to  add!) have had such a small number of races that their Impact Values  and ROIs are both falsely inflated, whereas Johnson now looks the  trainer to be most interested in considering that his two “positives”  have been generated by a significant sample size. An Impact Value of  1.4 is not that high, but still indicates that he is 1.4 times more  likely to win maiden races than the average trainer. Also profits of  nearly 30% are very acceptable over a good number of runs / bets.

 

What people do next will vary – however, it may not  be prudent to simply back all Johnson’s runners in maidens without any  further thought. There is a fair chance that you might make a profit  thanks to the combination of the past positive stats, but even at this  point, further research and analysis should be a priority. You have  arrived at what looks like a profitable trainer/race type combination,  however there is more work to do. What happens next is entirely up to  the individual, but if it was me, I would do at least two things:  firstly I would examine the winning prices of the 32 winners. This  would help determine whether there are any further patterns – we have  already seen that stats can be misleading. Despite all this information  we have now about Johnson’s runners, if one of the winners had been  priced at 66/1, then all the profits generated from the 234 runners  would have essentially come from this one horse. That would immediately  create a re-think and make you wonder whether this area needed further  exploration. Assuming the prices of the horses fitted into a fairly  standard pattern, then the second thing I would do is break the results  down into years, or perhaps blocks of 25 races. This would be an  attempt to see if the results followed a fairly consistent pattern over  the time period. The results of this secondary research might indicate  that results have tailed off in the past couple of years – again this  would set the alarm bells ringing. Whatever the outcome of this further  analysis, be it statistical or more form based, at least you would be  giving yourself the best opportunity to profit from your research. A  combination of a positive Impact Value, a positive ROI, and consistent  results are a powerful combination. A little hard work perhaps, but the  chances are it would be worth it. Now if only I could create an Impact  Value for hard work and profit making – now that would be something!

Article provided by Punter Profits