Statistics The Pattern In Numbers

Statistics is what happens when you have many numbers and want one number that summarises them. Cricket commentators do this every day. They say Babar averages 50 in ODIs. They say Pakistan's median first innings score is 280. They say the most common score in T20 is around 160. Average, median, mode. Three different summary numbers for the same data.

Mean, also called average, is the sum of all values divided by the count. If a batter scores 30, 40, 50, 60, and 70 in five matches, the mean is 30 plus 40 plus 50 plus 60 plus 70 equals 250, divided by 5, equals 50. The mean tells you the typical level if you smooth out the highs and lows.

Median is the middle value when the numbers are sorted. For scores 30, 40, 50, 60, 70 the median is 50. The median is more honest than the mean when one value is very extreme. If a sixth score is a duck of 0, the mean drops to about 41.7 but the median is still around 45. The median resists outliers.

Sample bias is the silent killer of statistics. If a survey of which biscuit is most popular only asks people inside one Imtiaz store in DHA Lahore, the result will not represent Pakistan. People in Larkana, Mardan, or Quetta might prefer a totally different brand. A small or unfair sample produces a confident wrong answer. Always ask, who is in this sample, and who is missing.

Today's habit. Pick any number that comes daily, like the temperature in your city, the highest score on any one Pakistani channel cricket replay, or the price of one tomato at your kiryana. Record it for ten days. Compute mean, median, and mode. See what the three numbers say together.