Yesterday, I looked at the Sri Lankan team's series in Australia, and concluded that the host's batsmen may be more vulnerable than their bowlers. However, cricket is not that simple. It is a very balanced game, and adding strength in one place can create weakness elsewhere. So as well as looking to where Sri Lanka's weakness is, one must also try to detect what England must protect.

I started this little series by projecting the County Championship statistics of Owais Shah and Ravi Bopara into the Test arena. However, most of the likely Test team selections for the England team already have enough statistics at this level to make some practical projections without having to deal with the more highly variable County statistics. I've gone back as far as 2004-5 and assembled some projections. One gives different weights to each set of statistics, and throws in a factor for regressing to the mean. (Players have a tendency to drift towards the mean level of performance over time. Every time you see a run rate calculation in a one-day game on the TV, you're seeing this principle exerting itself. Commentators rarely seem to understand this.)

First of all, let's look at the England batting order. These are the players I reckon have guaranteed places, with their projected averages.

		Vaughan		32.74
		Cook		35.08
		Bell		33.49
		Pietersen	45.02
		Collingwood	29.56
		Prior		29.13
		Sidebottom	14.83
		Panesar		13.54
		Hoggard		 6.97

Panesar's projection is far more optimistic than I'd give him, an effect of regression to the mean. I'd be inclined to halve it at best. However, rather than undermine the principle of performance analysis, I'll stick with what the statistics show. This batting order adds up to about 240.

Things are not so easily summarized for the bowling, so I'm not going to produce a table. Instead, I'll tell you that assuming the Sri Lankan batsmen can stay in for about a hundred overs, based on the strike rates of Panesar, Hoggard and Sidebottom, plus an average England bowler for the 2004-7 period, a team can expect to score about 280 runs.

Now, if we cast our mind's eye back to what I projected for Bopara and Shah (or scroll down the page), you can see the nature of the problem. England score 240 runs. Assume we add a typical tailender, who will average around 6, England are still facing a deficit of about 34 runs. Bopara looks likely to be more expensive than an average England bowler, so selecting him and his projected batting average of 26 or so is a recipe for defeat. Shah's 32 carries England to within 2 runs of leveling the scores, and Shah plus someone like Anderson looks more likely to win the Test for England than Bopara.