Dedicated to John Cole's Virtual Cricket game

Tuesday, May 09, 2006

Standard Deviation and VC5

A recent article on cricinfo examined the influence of standard deviation in working out which batsman was best.

Basically, standard deviation is a statistical "thing" that, in the article, determines how consistent a batsman is.

Say you are comparing Batsman A and Batsman B. Both are great players and have a batting average of 50. But let's say Batsman A tends to score runs like this:

200,0,0,0,200,0,0,0,200...etc.

So while he averages 50, he scores a duck 3 out of every 4 innings but then scores a massive 200 1 in 4.

But let's say Batsman B scores like this:

50,50,50,50,50,50,50,50... etc.

The guy scores the same amount of runs as Batsman A, but is consistent throughout. He does not, however, score a century.

Batsman A has a very high standard deviation while Batsman B has none.

According to the cricinfo article, Jacques Kallis is the most consistent top batsmen around while someone like Brian Lara is prone to scoring big but having very lean periods in between. This is one of the beauties of cricket.


It would be great if VC5 had some sort of standard deviation for both batsmen and bowlers in its design.

No comments:

Statscounter


View My Stats (from 21 Feb 2008 onwards)