This is usually something done in the workplace to find out how long a job should take and then they would have a good idea if a worker was average, slow or fast. Average in racing is quoted as the Median time - the average time it should take the average horse to cover a certain distance - the distance in my world of racing only concerns five and six furlong races, the type of races where there is usually a fast pace for the whole distance and a more reliable yardstick than longer distance races where they could be run to varying degrees of early pace.
I cannot see how you can ever find an average horse, or more specifically the official handicap rating of an average horse. I have studied the Median times at all flat courses for sprint races and I found it impossible to judge a horse using this approach. I was looking for the key to open up the secrets of speed variations between courses and I tried many ideas.
One of the first problems is the actual distance of a five-furlong race at each course, they are all classed as five furlong races but not all of them are exactly that distance. Bath is five-furlongs and eleven yards, Brighton is fifty nine yards over which is the biggest difference, Leicester is just two yards over the described distance. So naturally I did the simple maths of dividing the times from each course by the exact yardage and then multiplying it by 1100 which is the proper distance in yards. In some cases there is very little difference but any course where it is actually more than the described distance has to be brought back into line. As I have said before, over this distance a horse can cover between 20.52 yards and 18 yards per-second and at the slower figure it would take a horse just over three seconds to cover the extra 59 yards. I'm not going to bore you with times and stats for every course but you can see the times given as a general five-furlongs are not accurate and obviously these discrepancies have to be sorted to have the full picture of how times can be used to your advantage.
It is this discrepancy that gives my figures more accuracy even then it is not the end of my ideas of getting even more of an edge. I have already written about how can you measure a 70-rated horse's performance against a 110 rated horse over the same distance and going. The way I have overcome this problem is to use a sliding scale from the highest rated horses right down to a lowly rating of 65 - anything below this is not worth looking at because it equates to selling class or worse.
Another variance that has to be sorted is the fact that if a low grade course like Catterick, which is classed as a grade four course, the times at this track would be lower if the better class horses had run there on a regular basis. The grade one courses such as Newmarket and Ascot have top quality horses running there regularly so the times are obviously more accurate than a lower grade course where lower quality have run regularly but they are probably not good enough to make the times as accurate as a class one course so another variance that I have taken into account.
I am not saying my times are absolutely accurate all I am claiming is that my figures are more accurate than the ones gained using the Median time as a guideline. What I will say is that I am confident that the horse that comes out top of my figures after making the usual adjustments for weight is the horse most capable of running the fastest. I do then add a few other filters, which I refer to as stats, these can be very useful when having to decide between two horses that are equal on my figures.