Four Balls That Never Reached a Column: Starc's No.1, Harmer's Ten-For, and the Uncounted Arithmetic of the ICC Bowling Table

I started with the balls nobody counted, and the ledger did the rest.

The new page of the ICC Test bowling rankings arrived with only three usable lines on it: Mitchell Starc reclaimed No.1 and pushed Marco Jansen to second; Simon Harmer took ten wickets in Cape Town; and the same Harmer climbed to seventh. Everything below those three lines — more than forty rows — was not cricket at all. It was the technical residue of cookie consent, a page-scraper's leftover. I say this at the top because when the source itself is half-missing, the first duty is to declare how much is absent. My entire reading rests on those three lines plus general Test-cricket knowledge; every other figure I am holding with my own fingers on it.

What happened was not a delivery but a mechanism. The ICC ranking runs on a rolling, weighted system — recent matches carry greater weight, older performances decay gradually. The direct consequence is that "No.1" is not a certificate of a bowler's present-day capability but a sum of his recent form and the decay of rivals' older points. Sixty-four matches fit into twelve columns, but four voices kept escaping the table — here they are: Starc's knee, Jansen's decay, Harmer's single match, and the Cape Town seam.

For Starc, returning to Test No.1 is consistent with a left-arm quick at the peak where new-ball swing and old-ball reverse operate together. But the report gives no average, no economy, no recent spell — only the position. And for a fast bowler, the position is the most fragile document in the file, because beneath it sits the entire workload ledger: how many matches, how many rest days, how many flights, how many consecutive Tests. Starc's career carries an injury record, and a rankings update never mentions it. I read that line twice — first for the number, second for what the number left out.

Harmer's story is more interesting because it stands on a single match. Newlands in Cape Town has historically been a seam-friendly ground; an off-spinner taking ten wickets there is technically a large signal — either the surface dried and roughened for the second innings, or Harmer built a line-length-flight match-up the batters could not break. But the report does not say who the opponent was, which series it was, what happened in the first innings. A ten-for in one match is a loud sound, but a single sound. Reaching No.7 demands more than one match — perhaps a few other recent innings pointed the same way, but the source is silent. I write that only as inference, and calling an inference an inference is my first rule.

I remember that join from the spring when sport stopped. For five months alone I rebuilt 824 IPL matches, roughly 197,000 deliveries, tagged by pitch, dew, toss and venue — only so that when play returned, the noise would have somewhere to come back to. That exercise taught me something directly applicable to this rankings update: data desks won the tactical debate years ago, while the commentary box still stands where it always did. And the problem here is not a new metric, it is translation. Hearing "No.1", a viewer understands "the best bowler" — when the correct sentence is "the highest points in this window of this arithmetic." That is the gap the microphone never explains.

My second lesson came from the 2026 World Cup, where I built live xG and PPDA feeds across 64 matches and later audited my own product: of twelve metrics, only four reached the broadcast. A producer replaced PPDA with possession percentage. After that day I began writing every data piece in two layers — the metric first, then a plain sentence a co-commentator can read aloud. For this rankings update, that plain sentence is: Harmer's seventh is the upper edge of his Test career, but that edge is built on a one-match staircase; Starc's first is his form, not his age.

Now the turn where the number deceives. Correlation is not causation. The line drawn between Harmer's ten wickets and his seventh place is the ranking algorithm's line, not a line of sustained performance. If the same Harmer bowls on a flat subcontinental pitch next Test, the points will dissolve quickly — and that is the algorithm's natural behaviour, not an injustice. Likewise Jansen slipping to second is no proof of a form collapse; it may be the ordinary arithmetic of old points decaying, because the report records only Starc's and Harmer's gains and mentions no loss for Jansen. When a table shows only the climb and prints none of the slip, that table is incomplete. I trust a column only after I have walked its exceptions twice.

The coverage boundary must also be stated plainly, because data is not neutral — it is produced where money and cameras went. Harmer's Cape Town match is visible because it was a Test and it was broadcast; a ten-wicket spell in a domestic Ranji or County fixture does not climb a weighted table the same way. That inequality is not the ranking's fault; it is the shape of the cricket economy. Reading this page from Colombo, I keep the gap between two economies in mind — a domestic season in one country costs what a single fixture costs in the other, and that same gap decides whose balls get counted and whose do not.

One more thing absent from this table: injury. Return timelines for fast bowlers are now set by PR teams, and "week-to-week" often conceals that the injury is nowhere near healed. The real question about Starc at No.1 is not how many points he holds but how many overs will be demanded of his body in the next six months — how many Tests, how many rest days, how many flights. The ranking does not count that. I say this as inference, but it is inference drawn from the same workload framework I open separately for every fast bowler.

In all of this, risk is low and the risk of over-reading is high. A rankings update is no controversy, no rules crisis; it is the ordinary output of the ICC's administrative engine. The real risk is treating Harmer's one match as a trend, and reading "No.1" as "greatest of all time." South Africa's two bowlers — Jansen second, Harmer seventh — in the top ten is a genuine signal of a strong current Test attack, but pronouncing on team strength from three numbers would be the same error I keep refusing.

So I read this page as a window, not a wall. Three things to watch next quarter: whether Starc stays at the top across two or more ranking cycles; whether Harmer takes wickets away from Cape Town, on any surface, in the next Test; and whether Jansen returns to No.1. The quietest row in the sheet usually has the loudest season behind it. Starc's No.1 is a date; Harmer's seventh is a condition. The real cricket will be whatever walks off the Cape Town seam and proves itself on the next pitch — or does not.

And finally, what this table can never tell you: the name of the spinner who took twelve wickets in a domestic match this very week and still never climbed a weighted calculation, because there was no camera there. Sixty-four matches fit into twelve columns, but four voices still escaped the table.