A batter's first ten balls should be scored as a separate innings

One strike rate for a whole career flattens two entirely different tasks into a single misleading number.

Split it. That is the whole argument, and it takes about four lines of code to do, yet almost no published batting number does it. A batter faces his first ten deliveries in near darkness, judging pace off the pitch and how much the ball is holding, and he faces his sixtieth with all of that information already in hand. Those are different jobs. Averaging across them produces a figure that describes neither.

The practical consequence turns up in selection. Two players score at the same overall rate. One is slow for fifteen balls and then extremely fast, the other starts quickly and plateaus. In a chase where wickets are in hand these are close to interchangeable. In a collapse, where a new batter arrives in the sixteenth over with four needed an over and no time to look around, the first is worthless and the second wins you the game. Yet the scouting sheet says they are the same player.

There is a bowling-side use for it too, and it is the one I find more interesting. If you know that a batter's rate against pace is respectable but that his first fifteen balls against spin are close to static, you do not bowl your spinner at him in the eleventh over. You bowl him the moment the batter arrives. The window is small, it closes fast, and the whole thing depends on measuring the early period on its own rather than folding it into a season total.

Now the objection. Small samples. A batter might face only a handful of new innings against left-arm spin in a year, and any number built on that is noisy. Fair, and it is why I would not want a public league table of it. But the same objection applies to almost everything we already publish with confidence, and nobody seems troubled by that when the number is flattering.

What annoys me is the pretence of precision. A career strike rate carried to one decimal place, presented as a settled fact about a player, is aggregating conditions, formats, roles, phases and eras that have nothing to do with each other. It has the appearance of measurement and the content of a rumour.

If you want one number instead of two, use the tenth ball. Runs scored from the eleventh delivery onward, per hundred balls. It is cruder and it is far more honest.