Average is still the better number for a Test opener and I will not apologise

Tempo statistics have taken over the discussion of batting, and the oldest measure survives every argument in the longest format.

The fashionable position is that average is a relic, a number that rewards occupying a crease and punishes the player who scores quickly and gets out. In white-ball cricket that is broadly right. Applied to a Test opener it is close to nonsense, and the reason is what an opener is actually being asked to produce.

His job is time as much as runs. Facing the new ball means absorbing the period when the bowlers are freshest, the ball hardest and the movement greatest, and every over he survives is an over the middle order does not have to face in those conditions. That value is not visible in his own scoring column at all. It shows up in somebody else's hundred two hours later, which is why average, which rewards not getting out, tracks the job better than any tempo measure does.

There is a real cost and the modern argument is right about it. A batter who bats slowly enough can average forty and cost his side the chance of a result, particularly on a surface that is deteriorating and where a fourth-innings target has to be small. Draws made by grinding are a legitimate outcome but they are not a strategy, and a side of accumulators cannot force a win away from home.

So the honest answer is that both numbers matter, and I would still weight average higher, because the failure mode it protects against is worse. A side that scores slowly usually draws. A side whose openers keep failing loses inside four days, repeatedly, and rebuilds a top order every eighteen months.

If you want a single improvement, it is not to replace average but to split it. Publish an opener's average in the first twenty overs of an innings separately from the rest, because that is the passage he is picked for, and a man averaging sixty who arrives properly only once the ball has softened is doing a different job from the one on the team sheet.

What has happened is that a set of ideas which are correct in the shortest format have been imported wholesale into the longest, mostly because the same analysts and the same broadcasters cover both.

They are different sports played with the same equipment. The measurement should differ too.