ICC Rankings Update: Marco Jansen Tops Both Test Bowling and Allrounder Tables, Abhishek Sharma Back at No. 1 in T20I

The ranking file on my desk that morning had three sheets. The first listed Test bowlers. The second, Test allrounders. The third, T20I batters. On the top row of the first two sat the same name: Marco Jansen. Ten wickets in two Tests, 5/33 in one innings, and 85 runs batting at No. 8. Those 85 runs were counted in the allrounder table; that 5/33 in the bowler table. One series, one man, two different arithmetic columns, two summits. On the third sheet, Abhishek Sharma had returned to No. 1 among T20I batters after 113 off 38 balls, even as Ishan Kishan's injury vacated a line in the same table.

I have been reading scorecards since 2026, and one habit has hardened with age: when a No. 1 headline appears, I first check how many matches sit inside the sample. The answer here is short — two. Nobody had asked me what the numbers looked like before, so I kept them.

Verdict first, evidence after: Jansen's double peak is real, but it is as much a calendar fact as a cricket fact — and the deeper flaw in this rankings brief is that three separate ledgers, Test bowling, Test allrounding and T20I batting, are wrapped into one story as though their maths spoke to each other. I will publish the specific condition that would prove this wrong further down.

Ledger, page one: how the number is built

ICC rankings are not an annual awards list. They are a weighted average — recent matches carry more weight, older matches decay. That means the table reflects two things at once: how well you played, and how often you were allowed to play. The second half never makes the headline.

The allrounder table shows that second half most plainly. A player must bat and bowl in the same matches to accumulate. A specialist who never bats, or never bowls, cannot climb. So this is not so much a talent table as an opportunity table, and opportunity is set by team structure and calendar, not by the player.

It is also worth being explicit about the window. The South Africa–Australia Test series was live, its third Test in progress at the time of the update. Alongside it ran the India–New Zealand T20I series and an ODI tri-series in Pakistan and the UAE. Three competitions, three formats, one rankings bulletin. That is both the brief's strength and its weakness.

One item the original bulletin did not carry, and which I am recording openly: the publication date and the ranking period. A ranking is a living document; without a timestamp it stops being a snapshot and becomes a rumour.

Ledger, page two: Jansen's two rows

The straightforward part first. Ten wickets in two Tests, 5/33 in one innings — that is form, and it is the form of a fast bowler with height, a left-arm angle and the new ball. It is the cleanest road to the top of the Test bowling table.

The tangled part is the 85 at No. 8. It was a good innings, but its weight in the table came less from the runs than from the size of the sample. A two-match window is small enough that a single innings becomes a large share of it, and a weighted average forgets old patience quickly. The effect is sharpest in the allrounder table, where every player's sample is thin by design, because no allrounder contributes with bat and ball in every match.

Somewhere around the four-thousandth row, a spreadsheet stops being arithmetic and becomes a season — but reaching that row requires twelve or fourteen Tests. South Africa's Test calendar has been among the thinnest in the world in recent years. That is where my scepticism sits: when your country plays few Tests, every good match speaks louder in the weighted average. That is not Jansen's fault. It is the table's design.

The reverse is equally true. A bowler who plays twelve Tests a year has his bad weeks absorbed just as quickly, and his rating settles. It is for this reason that the same bulletin noted the top two Test bowling spots could change soon. That is not speculation; it is arithmetic, and the arithmetic says a two-match peak is a peak, not a throne.

Ledger, page three: Abhishek Sharma and the most volatile table

T20I batting is the fastest-moving of the three tables, for an obvious reason: the smallest sample and the fastest scoring rate. Abhishek Sharma's series read 29, 27, and then 113 off 38 balls. The third innings almost erased the first two. That is the economy of the format — one innings can rewrite a week's arithmetic.

One word needs correcting. The headline said he took the crown back. The record shows Ishan Kishan had moved ahead of him in July, and Kishan is now injured. So the return had two contributors: Abhishek's 113 and Kishan's vacated slot. A comeback in the rankings is usually the product of two windows — one you opened, one that closed for somebody else. The 113 did the work, true; without the injury, the swing might have waited for the next update.

So I read the No. 1 T20I line as a temperature reading rather than a certificate of form. It says who is hot now, not who will be best over six months.

Ledger, page four: three formats, one story

Here the structural weakness shows. A single report moves between Test bowling, Test allrounding, T20I batting and ODI tri-series. The experienced reader keeps the formats apart. The trend reader does not. When four tables flow through one paragraph, the reader assumes all four obey the same maths — while the weighting, the overs bowled and the sample size differ in each.

This is why I prefer to read such briefs as three small stories. The Test table tells you whose patience is holding. The T20I table tells you whose bat is warm this month. Reading them together creates the illusion of a single ladder up which everyone is climbing. There is no such ladder.

Ledger, page five: the names the headline left out

There was movement beyond Jansen and Abhishek. Temba Bavuma's 179 is a serious row — an innings of that size, carried alongside the captaincy, climbs straight up the Test batting table. Harry Brook moved up, Scott Boland and Simon Harmer advanced in the bowling list, Shaheen Shah Afridi's return was logged, and Azmatullah Omarzai appeared among the upward movers.

The list is not a story by itself, but it does say one thing: the top floor is crowded. Between names like Mitchell Starc, Matt Henry, Ollie Robinson, Jasprit Bumrah and Ravindra Jadeja, a few points separate them and the order shifts week to week. When the top ten are that close, a No. 1 headline should be read as a position, not a result.

Ledger, page six: the row that was never in the table

Nowhere in the bulletin was there a single line of women's rankings. An empty row does not mean an empty dataset; it means a different question is being asked. The women's record has filled up fast in recent years — the franchise structure that began in 2026 and the major tournaments that followed built a category that previously had no ledger. When a rankings roundup skips that category, it is not dropping information; it is dropping an audience.

I do not read that as a failure of the players. It is an indexing habit, and habits change slowly. But the custodian's complaint stands: a row that never gets printed does not exist for the next generation.

The Hindi and Urdu sports bulletins I read daily often frame this update differently — more weight on Jansen's 85 than on his bowling, because an innings travels faster with the general reader. That is not wrong. It is a different ledger. Only reading both completes the picture.

The contrarian angle: a ranking is a bank statement, not a mirror

Now the part that demands care. The brief's implicit assumption is performance, then ranking movement, then headline. The chain is attractive and it skips a step. A ranking does not tell you who the best bowler is; it tells you whose weighted average is currently favoured. The gap between those two sentences comes from three places — matches played, quality of opposition, and series context.

In Jansen's case, the bulletin does not say on what pitch, against which batting order, in which innings the 5/33 came. Nor does it say what situation the 85 was made in — a match-saving innings or a match-winning one. Without that context, 85 and 5/33 are just figures, and the table counts them the way a ledger counts turnover.

The other direction is equally true: there is no alternative to rankings. A player who performs repeatedly stays near the top over time; no artificial spike survives. My objection is not to the long-run ranking but to the way it is read week to week.

And here is the condition that would prove me wrong, stated in the open: if Jansen stays in the top rows of both tables through the coming subcontinent tour and the twelve-plus Tests that follow, my argument that this is a calendar effect weakens badly. Then it would be a genuine rise, not a two-match flare. Conversely, if he takes a few fewer wickets in a home series and slips three or four places, the sample-built nature of the peak is confirmed.

The third point stings most: the missing timestamp. No publication date, no ranking period, no source scorecard for the match figures. This is a common carelessness in rankings reporting and, to my mind, the most expensive one — because the entire value of such a brief sits in its time-stamp. Without a date it is not news. It is rumour.

Forward signal

Three things I will watch over the next few weeks, read from scorecards and not headlines. First, the rest of the South Africa–Australia series: does Jansen's line hold when the pitch is slow and the ball old. Second, India's next T20I assignment: does Abhishek Sharma's No. 1 survive Ishan Kishan's return, or does it prove to be the two-window comeback. Third, and most important to me, the next ICC update: does it carry one line of women's table, or does that column stay blank again.

A ranking is a ledger somebody else fills in each week, and those of us who read it treat it as final truth. My argument is only that it should be read as a ledger — row, context, date, and then opinion. Will the next update deepen Jansen's double peak, or will it confirm that a two-Test sample was only ever a two-Test sample?