Won the Series, Lost Percentage Points: The WTC Sum the 2-1 Scoreline Does Not Show

Cape Town, Saturday. A margin of 176 runs. For anyone reading the scoreline, the file is closed: South Africa beat Australia 2-1, took the series, and moved to No.1 in the ICC Test ratings. I wasn't watching the scoreline that evening. I was watching the percentage column of the World Test Championship, and that column was moving against the team that had just won. The side that won the series fell from 75.00 to 71.43. The side that lost it fell from 80.00 to 69.23 and dropped to third. And the side that did not bowl a single ball in this series sat at 72.22 and became No.1. All three figures are correct, and all three come out of the same arithmetic — I ran them backwards, broke them into wins and losses, and rebuilt the table nobody printed. Two screens, one scorecard, and a column no one re-reads. The story that ran was “Australia lose their No.1 ranking.” The story in the table was that winning a series does not raise your percentage. In India the Test-ranking argument usually gets stuck on why two tables contradict each other. They measure different things. The ICC Test rating is a rolling, multi-year weighted rating — every result lands on a stack of older results and the numbers do not start at zero. Australia sit at 122, South Africa at 124. Two points. The WTC table is the opposite: current cycle, twelve points a match, percentage = points earned ÷ points available. One table measures standing, the other measures efficiency inside a cycle. Reading them together is the key to this whole episode, and it is the one thing the original report never explains. Under standard WTC convention a win carries full points, so every percentage decomposes into a clean win-loss ratio. Australia were on 80.00 before the series — eight wins in ten, 96 points from 120. After Cape Town they have thirteen matches, nine wins, four losses: 108 from 156, or 69.2308 per cent. South Africa were on 75.00 from four matches with three wins; now they have seven matches, five wins and two losses, 60 from 84, or 71.4286. New Zealand sit at 72.22 — thirteen wins in eighteen, or 104 from 144 built on eight wins, two draws and two losses in twelve. Both routes land on the same figure, and that ambiguity is itself a signal of how many draws are buried underneath. Those three numbers force a pre-series table: Australia 80.00 first, South Africa 75.00 second, New Zealand 72.22 third. That table appears nowhere in the report, even though every claim in it rests on this table — Australia dropping two places, South Africa second, New Zealand top. None of the three can be checked without it. A reader with only the newspaper has no way of knowing where South Africa stood before the series, which is precisely why the reader cannot tell whether winning the series helped them or hurt them. The shape of the series is also written backwards. All the text gives you is that Bavuma's 179 “brought the Proteas back into the rubber” — meaning South Africa lost the first Test, levelled at Gqeberha, and settled it at Cape Town. The result of the first Test, its venue, and the win-loss sequence are all missing, and that sequence is the entire story: a side 0-1 down coming back to take the series and the top ranking. The reader has to reconstruct it. The reader who doesn't walks away believing South Africa were never behind. Here is the part nobody wrote. South Africa won 2-1 and their WTC percentage fell. That is not a contradiction, it is the definition. The denominator is matches played, and every loss adds a full match to it. Three wins in four matches is 36 from 48, or 75.00. Winning the series added 24 points to the numerator — two wins — but added three full matches to the denominator, because a loss is also a full match. 36 plus 24 is 60; 48 plus 36 is 84; and 60 over 84 is 71.43. Winning a series is not a guarantee of a rising percentage. It depends on how many you lose while winning, and how small your old denominator was. I ran the percentage backwards and it gave me a series in which the winner's arithmetic ended up worse than the loser's. Which is how New Zealand got to No.1 without bowling. Their 72.22 sat banked while South Africa's 75.00 slid to 71.43, because South Africa lost one match inside a series they won. It is the same arithmetic that gets relitigated every IPL season around net run rate, except in Test cricket it shows up cleaner because the denominator is small. In a seven-match table, one Test is worth roughly fourteen percentage points. Fourteen. One Test, one series mood, sometimes an entire cycle. This is where I have to doubt my own table, and where the original report is entirely silent. South Africa's 71.43 rests on seven Tests. Australia's 69.23 rests on thirteen. New Zealand's 72.22 sits somewhere between twelve and eighteen, depending on how the draws were counted. A percentage normalises the denominator; it does not normalise sampling variance. A seven-match percentage is far more volatile than a thirteen-match one — the next result moves South Africa to roughly 72.00 or to 62.50. I am not arguing South Africa don't deserve No.1. I am arguing that the claim is built on a seven-match sample and should be read with that condition attached. What would change my mind: South Africa's next home cycle, where the match count grows and the ratio settles. On the ICC ratings the picture is thinner still. South Africa 124, Australia 122. Two points. In a rolling rating, two points can turn inside an innings, not a series; one heavy defeat takes three or four off. The report presented the handover as settled, when the arithmetic says it is the most reversible kind of change available. New Zealand, India and England were named in the top five without their points — selective data that props up the “power changed hands” framing rather than the closeness of the contest. And the confusion that will have misled the most readers: New Zealand are third in the ICC ratings and first in the WTC. Not a contradiction, two answers to two questions. The rating asks what you did, against whom, where, over years. The WTC asks how many points you banked in this cycle. New Zealand have been efficient in this cycle; the rating reflects a longer and harder history. Printing both tables together is fine. Printing them without the distinction convinces readers that cricket's statistical machinery is itself confused, when the confusion lives in the explanation, not the tables. Bavuma's 179 at Gqeberha is the hinge the series turned on, and the only place in the report where a human voice appears. His line — that this was his best win — is quoted, but the bracket attached to it belongs to the editor, not the speaker: he was talking about a series, and the text converts it to “best Test win.” A small thing, except that adding a word inside a quote takes a meaning away from the reader. The larger problem is the silence from Australia's side — no captain, no coach, no selector. A team loses the top ranking and not one sentence from their camp is recorded. That is not a balance problem, it is a reporting gap. The 176-run margin carries no detail either. No totals, no bowling figures, no spell that turned it. A decider with a 176-run gap usually has two or three spells that built it: the third seamer's first-innings burst, the first hour after lunch, a dropped catch that reset the shape of an innings. The report looks for none of them. That is the biggest miss for me, because 176 runs is an outcome, not a cause. The tape nobody re-watches is where the table's real mistake hides. If I were writing one line on Australia's defeat, I would first decide whether the error was systemic or individual. Losing one series after nine wins in thirteen is not one fielder's fault. It usually sits in selection architecture: fast-bowling workload, no extra spinner on a turning surface, a batting order reshuffled mid-series. The report investigates none of this because it never spoke to Australia. When you don't ask the losing side, your story stops being analysis and becomes publicity for the winner. The report also stops being objective at the close. It says South Africa's win “changed nothing for India.” Half-true is dangerous here. India's own points and percentage are unchanged — correct. But India's direct rival for a final place is Australia, and Australia have just dropped two WTC places, from 80.00 to 69.23. That is an unambiguous gain in India's relative position. “Nothing changed” only holds if you limit the table to India's own column and ignore the other six teams. In a tight race, relative position is the thing. Weaker still is the next claim: that the only way India can be absolutely sure is by winning all seven remaining Tests. Winning all seven is sufficient, not necessary. A WTC draw is worth four points, a win twelve. India have two Tests against New Zealand and five against Australia, 84 points available. Five wins and two draws gives 68; four wins and three draws gives 60. Neither scenario is out of qualification, provided the teams above drop points — which Australia did this week by losing two matches. Saying “win all seven” is easy, but it is the language of intent, not arithmetic. A scenario table would have taken four lines. It wasn't built. There is a bigger question the report never raises: is the method itself wrong? If the series winner's percentage can fall, the table is not rewarding what it claims to reward. Alternatives exist — more points for a win, a lighter loss penalty, a denominator built on series rather than Tests. Each has a cost, and I won't advocate one here, because recommending a method change off one cycle's numbers is the same error I am charging the writer with. But this much is fair: when the table's output runs against a reader's plain reading of events, the reader is not the one who is wrong. Headline and body also work against each other. The headline foregrounds the loser — “Australia lose their No.1” — while the first paragraph opens with the winner. Deficit framing pulls readers, that is an old trick, but here it mislabels the event. This week's news is that South Africa reached No.1, not that Australia slipped. And paragraphs two and three restate paragraph one three times without spending the freed space on match detail. The room reserved for the series went into repetition. One more thing I am holding as unverified: the report says last year's WTC final was contested between these same two sides. That sentence carries no source, and it is the sentence that fixes the whole timeline — which cycle is running, which series was “last year,” which season India's seven Tests fall in. When the timeline hangs on an unsourced line, every date in the piece becomes suspect. And something that sounds technical but isn't: the date on the table. WTC and ICC ratings both change after every match. When a report doesn't say which day the numbers were pulled, the piece cannot be checked, cited, or used as a comparison six months later. That is not a style failure, it is an archive failure. A caveat I have to attach to my own reading. The cleanliness of these percentages raises suspicion. 69.2308, 71.4286, 72.22 — so exact that not one draw appears anywhere. Thirteen matches, nine wins, four losses, no draws. Possible, not typical. So one of two things is true: the numbers are real and very carefully written, or they were built to satisfy the checks. Internal consistency raises my confidence in the arithmetic, not in the events — and that is the difference between an audit and a report. Until a dated ICC table is in my hand, I am marking these figures as pending verification. India's two Tests against New Zealand come in the next few weeks; five against Australia come early next year. When you open the table, watch the denominator, not the points. A small denominator can swing fourteen points on one match; a large one dies slowly. The next time someone says a team became No.1 by winning a series, ask which sum they used — and which day's table they are reading.