An idea of how ‘spread out’ the values are in any data set.
Photo: AFP
How many runs? Calculate the average.
Now, let’s go through the same steps for our inconsistent batsman. In this case, the differences between his average of 50 runs and his six test scores of 10, 90, 95, 5, 100, 0, would be: 40, 40, 45, 45, 50, 50. Already we can see that the bats
man often scores very differently from his average — much more or much less. If we were to square these differences, we’d get: 1600, 1600, 2025, 2025, 2500, 2500. Taking the average of the squares of these differences, we would get 2041.7, or approximately 2042.
Performance and average
Compare this number to 10 (which is what we got for the reliable batsman) — isn’t 2042 much larger than 10? It is larger because the variation number of runs (or spread in the scores) made by the inconsistent batsman is large. Often, the number of runs he scores is much larger or much smaller than his average — and his performance varies greatly from one test to another.
This is, of course, only one way to get an idea of the variation in a set of numbers — there are other equations which also provide numbers that give us an idea of how “spread out” the values are in any data set. Some of the other ways to calculate variance actually provide more accurate measures. But more on that at a later time, perhaps…
Printer friendly
page Send this article to Friends by
E-Mail
