A Ranking Is Not a Scorecard: Marco Jansen's Two Crowns and the Date the Story Left Out

The most important figure in this week's ICC rankings update is the one that is nowhere in it: the date. Sixteen data points, five hard facts, two background notes, one opinion — and zero timestamps. Marco Jansen now tops both the Test bowling and Test allrounder charts; Abhishek Sharma is back at No. 1 among T20I batters. Both headlines are true. Both are also incomplete, because the moment the ratings were cut is unstated, and an ICC rating is a moving window, not a fixed photograph.

I learned this in 2026 on a Delhi data desk, building a ball-by-ball ledger. A scorecard without an over number is not a scorecard; it is a summary. Cross-checking more than 14,000 deliveries against broadcast records taught me that two descriptions of the same innings — the television's and the scorer's — diverge within half an hour, and whichever was filed first becomes the record. A ranking is not a scorecard; it is the page of that ledger on which nobody wrote the divisor.

Start with the facts as filed. Ten wickets in two Tests, an innings return of 5/33, and 85 from No. 8 carried Jansen to the top of both Test lists. The third Test of the South Africa–Australia series is still in progress, so this picture can change inside four days. Abhishek's recent sequence reads 29, 27 and 113, the last of those off 38 balls — a strike rate of 297.4. Ishan Kishan's injury supplies the surrounding context. Temba Bavuma (179), Harry Brook, Scott Boland, Simon Harmer, Shaheen Shah Afridi and Azmatullah Omarzai also moved. India's T20I series against New Zealand and the ODI tri-series in Pakistan and the UAE appear in the same roundup.

The question starts here, and it is blunt: are Jansen's two crowns two achievements? The intuitive answer is yes. The ledger's answer is no.

An ICC rating is not an aggregate of points. It is a weighted average: not every innings carries the same recency, the scale factor shifts with the strength of the opposition, and each format keeps its own table. Six hundred points this week are not the six hundred points of last week. A reader sees a table; on a data desk the same object looks like an equation with the window length, the innings count and the opposition coefficient all hidden inside it.

That applies twice over to the allrounder list, because the structure there is different. The widely documented construction multiplies a batting rating by a bowling rating and divides by 1,000, with a qualifying threshold attached. The update note publishes none of it — no coefficients, no threshold, no window. That is precisely where a ranking becomes news and the method disappears.

The multiplication is the story. Suppose Jansen's batting rating sits near 500 and these two Tests lifted his bowling rating by 40 points. That adds 500 × 40 ÷ 1,000 = 20 points to the allrounder rating. Read it backwards: the higher his batting rating, the harder any bowling surge registers on the allrounder list. The numbers here are illustrative, not the ICC's actual coefficients, but the direction is not in doubt. Jansen did not win two crowns in two weeks; one week's work passed through two windows, and both windows counted it in full. Two crowns read as two achievements. In arithmetic they are two receipts for one performance.

Five for 33 is the brightest sentence in the story and the most incomplete. Ten wickets across two Tests must be spread over four innings; the roundup shows us one. Where the other six fell — which innings, which pitch, which part of the order — is unrecorded, though it sits in the scorecard for anyone who looks. Saying the bowling rating is 'deserved' without that breakdown is the same carelessness I never tolerated with any number. My ledger carried three columns beside every wicket: ball type, the batter's score-state, and where the field stood in that over. A wicket that fell three overs after a field change and a wicket that fell to the bowler's plan both read as '1' in the scorecard.

And 85 from No. 8? This is where the 2026 ledger comes back onto the table. I coded every innings by phase — powerplay, middle, death — and learned that year that 42 off 61 means three different things depending on whether it happened in the 12th over or the 42nd, and what the required rate was at the time. The ICC batting model does not weigh a batter's position that way. 85 at No. 3 and 85 at No. 8 get the same scale. That does not mean Jansen was over-rewarded; it means the crown does not testify to the rarity of the innings it appears to celebrate. In my account, a No. 8's 85 carries a different weight from a No. 3's 85, because expected runs were lower and risk was higher. The ranking cannot see the difference. That is its boundary, not its error.

Now Abhishek. 113 off 38 balls. A strike rate of 297.4, close to three runs a ball. The arithmetic is simple: scoring 113 requires at least 19 balls to have gone to the boundary (113 ÷ 6 = 18.83). More than half the innings spent clearing the rope is exactly the zone of T20 where the sample is small and variance is large. Cricket calls that innings 'form'. A dataset calls it the right-hand tail.

A rating is a weighted average, so that innings arrives carrying 29 and 27 with it. This is where the headline and the mathematics take separate roads. If the 113 is the newest row in the window, it deposits its heaviest weight this week and then dilutes with every innings that follows. So the No. 1 spot has not 'come back' as a verdict on form; it is the product of a sharp improvement across three innings, two of which added little. A fast innings did not hand over the chair; it closed the gap underneath it.

Ishan Kishan's injury is the least examined cause in this story, and the cause is not an action but a mechanism. A player who does not play does not lose rating points — his rating freezes while the ratings around him move. The injury froze Ishan; Abhishek moved his own. Putting both in one headline is easy, but one is a cricket event and the other is a calendar event. 'Returning' in a ranking table can mean a return to form, or it can mean waiting for someone else's absence.

Bavuma's 179, Brook, Boland, Harmer, Shaheen, Omarzai — that is not a ranking list, it is a fixture list. The largest single driver of ranking movement is not talent; it is scheduling. Players whose teams are playing climb; players in a rest cycle sit still and look neither better nor worse. Reading the apparent stability of Mitchell Starc, Matt Henry, Ollie Robinson, Jasprit Bumrah and Ravindra Jadeja as stability of form is the same mistake. That stability often reports only how many Test deliveries their window contained. When I mapped the 2026 Russia World Cup's expected-goals logic onto cricket as expected runs and wicket probability, the most useful lesson was that expected and actual are different objects, and printing them in one column is the worst available error. A rating is closer to a probability than to a score, and it is being read as a score.

The third problem is structural. One report carries a Test bowling table, a T20I batting table and an ODI tri-series in Pakistan and the UAE. In the reader's mind those three become one scale. Putting 10 Test wickets and a 113 in a T20I on the same board is like writing a 400-metre time and a marathon time on the same wall. The methods differ, and that difference is nowhere stated.

One more omission misleads. The allrounder definition belongs to the ICC — how many innings batted, how many bowled, what threshold. None of that appears in the update. The reader consequently reads a table as a selection committee's opinion when it is a mathematical entry gate.

And one blank that no rankings roundup fills: the women's tables. Sixteen data points were spent on three men's lists and not one on women's cricket. That is not politics; it is an editorial habit, and habits harden into institutional memory.

Now the part where I argue with my own thesis. The easiest conclusion is that the rankings are meaningless. That conclusion is wrong, and I wrote the null hypothesis before the analysis: if the rankings genuinely record performance, every movement in this update is justified. Tested against the evidence, the null survives. Jansen did produce the numbers that belong at the top of both lists, Abhishek did play the biggest innings, Bavuma did make 179. The ranking is not lying. It is answering one question while the reader is asking another.

The ranking asks: what was this player's weighted average over the recent period? The reader asks: is this player the best in the world right now? The first is bookkeeping, the second is forecasting. The moment a newspaper prints the answer to the first as the answer to the second, a statistical claim becomes public belief. This is where the data desk loses ground to the commentary box — the box answers instantly, while the desk asks for the divisor first.

And that line about the top two Test bowlers being liable to change quickly is being read as excitement. I read it differently. If the gap between No. 1 and No. 2 can be closed by one good innings, that is not evidence of volatility — it is evidence of the divisor. A model whose rating depends that heavily on the last few innings was built to be sensitive. That is design, not defect. But then it should be read as 'the fastest climber of the last month', not 'the best bowler in the world'. Precision is being traded for thrill.

And 'takes back the crown'? The source does not carry Abhishek's full ranking history; all it establishes is that Ishan overtook him in July. The intervening months are absent. A phrase that needs three lines of data to verify was published without them. It is a small omission, the kind that accumulates into institutional memory.

I think of 2026, when the IPL moved to the UAE without crowds and my first draft came close to declaring home advantage dead. Sixty matches, more than 7,200 balls, a conditions-adjusted model — and the finding was that home advantage had not died, it had hidden in the dew, and the toss effect clustered around moisture rather than crowd pressure. Mumbai Indians won the title and Jasprit Bumrah's 27 wickets gave the model a clean test case. I learned that year that a counter-intuitive thesis is easy to build and hard to defend in front of data. This piece does not exist to prove the rankings wrong. It exists because the way we read them should be stricter.

So what do the next rounds test?

First, the third Test between South Africa and Australia. If Jansen's bowling rating dips after a flat, wicketless innings, we learn how durable the summit is. Second, Abhishek's next three innings — the 29, 27, 113 pattern says the weight of his rating rests on one knock; two more innings above a 200 strike rate will settle the chair, two cheap dismissals will shake it. Third, the ODI tri-series in Pakistan and the UAE, where names will climb a table that will not appear in the next Test list, and readers will repeat the format confusion this roundup invites.

One question I leave open. Should the ICC publish three things with every update — the date, the window length, and the points gap between the top two? I know it will not happen, because a ranking's entertainment value lives in its vagueness. But in this trade your conclusion is only as trustworthy as your timestamp. This week we have two No. 1s, fourteen names, one injury — and a missing date.