A No. 1 Built on Two Tests: What the ICC Rankings Measure, and What They Don't

Marco Jansen took ten wickets in two Test matches. One innings reading 5/33. And 85 runs batting at No. 8. In the latest ICC update, those three lines bought him two crowns in the same week — top-ranked Test bowler and top-ranked Test allrounder. The headline is clean. The arithmetic is clean. What is not clean is the sample: two matches.

Forty-five years of reading scorecards has taught me to look for three columns — what happened, against whom, and over how many matches. The third column is the first thing a rankings story drops. The No. 1 badge made no sense to me until I subtracted the schedule underneath it.

First, what an ICC ranking actually is, because most readers treat it as a league table and it is not one. The engine runs on a rolling window: every match earns points, recent matches carry more weight than old ones, and the strength of the opposition is folded in. Rating equals total points divided by matches played. That makes a rating a flow of recent results, not a stock of ability. Flows swing; stocks do not. The ICC publishes its method, but the part that matters most never gets printed — which matches are still inside the window, and which have quietly fallen out of it.

The picture in this update is wide. Abhishek Sharma returned to No. 1 among T20I batters on the back of 113 off 38 balls, a strike rate around 297, with Ishan Kishan's injury supplying the context. Temba Bavuma's 179, Harry Brook, Scott Boland, Simon Harmer, Shaheen Shah Afridi and Azmatullah Omarzai made their moves, while Mitchell Starc, Matt Henry, Ollie Robinson, Jasprit Bumrah and Ravindra Jadeja sat in the background having not played that week. Three formats on one page: Test, ODI, T20I — the South Africa–Australia Test series, the India–New Zealand T20Is, and an ODI tri-series in Pakistan and the UAE. As a roundup it works. As a reading experience it carries a risk, because three currencies get deposited into one account.

Start with the allrounder table, because that is where the largest misunderstanding sits. Most readers assume the ICC allrounder ranking is an independent measure — some third skill being graded. It is not. The allrounder rating is the batting rating and the bowling rating added together, not a standalone index. Which means that for a player already on top of one department, the second crown is half achievement and half bookkeeping. Jansen's two crowns are the product of one two-Test window, not two separate narratives. Had the 5/33 and the 85 arrived in different months, the two badges would likely have landed in different weeks and the story would have looked far less luminous.

Now the 5/33 itself. Which pitch it came on matters less than which opponent and which weather. South Africa against Australia is seam against seam, bounce against bounce. In that climate a fast bowler's wickets-per-ball reads better than his career average, and the ranking treats the better ratio as permanent. The ranking treats one evening as a full year; the ground only ever knew it as one evening. When I built a conditions-adjusted model in Dubai in 2026, that was the lesson it kept returning: every number needs its environment written beside it or it starts lying. Here, the environment is nowhere in the copy.

Second, the 85 at No. 8. That innings lands in the batting rating and hits the allrounder table twice over. Consider the sequence — a bowler who had already reached the top of the bowling table saw his batting rating rise after one of the better innings of his career, and the two columns were then added. His arrival at No. 1 among allrounders was close to inevitable. It is a genuine achievement and should be written as one. Reading it as "the best allrounder in the world" is something else: it is adding two columns and calling the sum a third truth. The addition is not wrong. Calling the addition a truth is.

Third, sample size, which belongs in the sentence and not the footnote. Two Tests. In bowling terms that is a few hundred deliveries, with the heaviest weight sitting on a single evening. My estimate is that at this sample size the ranking's confidence band is wider than several places inside the top five — a gap of three or four rating points that one good spell erases. The table this week is not a forecast of next month; it is a receipt for last month.

Fourth, Abhishek Sharma's 113 off 38. The T20I formula weights strike rate heavily, and that is correct: in T20I cricket, run rate is the actual currency. The nuance lives elsewhere. One innings of 113 moves a rating further than four scores above thirty in the same series, because the rating rewards speed rather than consistency. That is a design preference, not a flaw. But when someone says "the No. 1 batter," what he is actually saying is that the man recently played the fastest innings — not that he won the most matches. Both are good. They are not the same.

And Ishan Kishan? His name was at the top and it slipped because of an injury. Those are two separate events stitched into one story. Abhishek's No. 1 is the output of his runs; Ishan's slide is the output of his injury. There is no causal link between them, only one chair. In news prose the join looks elegant. In data prose it is two ledgers. When I was handling scorecards at total-cricket.com two decades ago, the rule was already old: putting two events in one sentence does not make them one cause.

Fifth, the names who did not play that week — Bumrah, Jadeja, Starc, Henry, Robinson. Here the structural problem shows itself. A player who is not on the field carries a frozen rating; a player who is playing has a rating that moves every match. Comparing the two is like adding an old share price to today's market and calling it profit. The table is technically accurate and practically misleading. The fix is not to abandon rankings; it is to print the date of each player's last match beside his name. Then the reader can see for himself which number is alive and which is parked.

Sixth, the format blend. Bavuma's 179 came in Test cricket, Brook and Boland earned their points in Tests, Harmer's rise came from England's spin-friendly run, and Shaheen and Omarzai from the ODI tri-series in Pakistan and the UAE. Three leaderboards in one report. Readers have to do the separation themselves because the roundup structure does not do it for them. My habit for years has been to print each format's table on a separate sheet, then watch which names stay on their sheet and which slide between two.

The peripheral names carry the real story. Shaheen and Omarzai's rise comes from an ODI tri-series most Indian viewers barely watch. Bavuma's 179 was made under the weight of captaincy, and Test batting rankings reward such innings slowly — which is right, because Tests should reward patience. Harmer and Boland are two names whose rankings are built less on talent than on conditions. Weighing all three on one scale in one table is not cricket's error, it is the ranking's design limit — and the copy never states that limit.

Seventh, the part cricket journalism mentions least. The ICC ranking is not a measurement; it is a product. The weekly publication calendar exists so that a headline exists every week. Boards use it in marketing, franchises write "No. 1 ranked" into the deal sheet before an auction, and for an institution that sells broadcast rights it is a moving advertisement. The table we treat as truth is an institution's calendar-driven product, built to generate a weekly news cycle at least as much as to measure the game. This is not conspiracy. It is business. And the reader deserves to know what he is holding when he holds a ranking.

Now the turn. Near the end of the report sits a line saying the top two Test bowling spots could change quickly. It reads like caution. It is actually the sharpest comment on the whole system. If the world's No. 1 bowler can change hands twice in a fortnight, the table is not telling you who is best — it is telling you who played most recently. A second objection sits underneath. Jansen's No. 1 rests on a two-Test sample; Boland's rise came from wicket volume in friendly conditions; Harmer's number from England's spin-friendly stretch. All three are products of their environments, and no column says so. In 2026, with 60 matches across three empty stadiums in front of me, I learned that conditions are not background — they are the lead character. The ranking never shows that character. And a number earns my trust only after I have watched it fail. I have watched many No. 1 badges arrive and leave, and the pattern holds: the week the badge arrives is the week it means least.

I rebuilt this table line by line, and the headline number quietly dissolved.

So the next time someone tells you a player is No. 1 in the world, ask one question — No. 1 over how many matches? For Jansen the answer is currently two, and the next Test of the South Africa–Australia series will decide whether this was a spike or a beginning. For Abhishek the answer is 38 balls; if he makes 40 not out on a slow pitch next series, his rating will fall while his value rises. For Shaheen and Omarzai the answer is an ODI tri-series almost nobody is watching — meaning the table moving the most is the one in the format watched the least. The next rankings update will tell you whether Jansen's No. 1 was a rating or an echo. I will wait for that week, because the post-mortem is the only room where the market tells the truth.