You cannot average a strike rate across innings and expect a true number

Adding up per-innings rates and dividing by innings weights a nine-ball cameo the same as a hundred-ball anchor.

Add the rates, divide by the innings, and you have made a mistake.

It is a common one, it appears in graphics on national broadcasts, and it produces a number that overstates almost every lower order batter who ever lived. The correct calculation is total runs over total balls, weighted naturally by how many balls each innings contained. The incorrect one treats a nine ball twenty at two hundred and twenty as equal in weight to an eighty-five ball seventy at eighty-two.

Do that across a career and the cameos dominate. A player with a lot of short, brisk innings ends up with a headline rate he never achieved over any meaningful stretch of batting.

This is not an obscure statistical nicety. It is the difference between a ratio of sums and a mean of ratios, and the two answers diverge most exactly where the batter's innings lengths vary most, which is to say in the middle and lower order, which is to say where strike rate gets used hardest as a selection argument.

There is a second layer that even careful aggregation misses. Strike rate is not a constant a batter carries around. It is a function of balls faced, and it rises steeply as an innings develops, because a set batter on a known surface can take on shots the same player could not attempt on ball three. So the career figure is really an average over the distribution of innings lengths a player happened to get. Change his batting position, change the distribution, change the number, without changing the batter at all.

Promote a finisher to number four and his strike rate will fall. Everyone will say he has lost his edge. What actually happened is that he now spends more deliveries in the part of the innings where nobody scores quickly, and the metric has no way to report that.

The fix is to publish strike rate as a curve against balls faced rather than a scalar. Ten to twenty balls, twenty to forty, forty plus. It fits in three columns and it answers the question people are actually asking, which is what this player does at the point in the innings where you would use him.

One number was never going to carry that. We keep asking it to, then arguing about the answer as though the argument were about the player.