Starc's Four in Six: Australia's Cape Town Comeback, but 329 Is Not Yet a Rescue

Mitchell Starc began the over with the second new ball, and inside six deliveries four names had joined the scoreboard. In a rented room in Bengaluru my ledger lay open on the table — the same ledger in which I record every evening that television compresses into a thirty-second clip and the scorecard forgets. It was the final session of the first day in Cape Town. Rickelton went off the fourth ball of the over, Jansen off the sixth, Coetzee off the first of the next over, Rabada off the third. Ball numbers 4, 6, 7, 9 — four wickets inside a six-ball window. This was not a T20 death over, where such figures are manufactured weekly. It was the second new ball of a Test match, and that difference is the key to the whole story.

The series stood level at 1-1 that morning. Australia had lost the second Test at Gqeberha by an innings and 31 runs, and the third match was the decider — not merely a series, but an entire tour hanging on one innings. Pat Cummins won the toss and chose to bowl, on a surface the original report itself calls batting-friendly.

Newlands needs no introduction: pace and carry in the morning, a breeze across the ground, seam movement under cloud, and spin growing more relevant from the third day. Choosing to bowl there means accepting that the best batting conditions are being handed back, and that the last innings will be played on a surface Lyon and Connolly will find friendlier than Cummins' quicks did on the first morning.

By lunch South Africa were 140/0. No wicket, no control. It was Australia's worst session of the match, and at that point Cummins' decision was being read as plainly wrong. The report carries no weather information at all — no cloud, no wind, no humidity. Omitting that on day one at Newlands is like leaving the over number blank in a scorebook.

A Test day is paid for in sessions, not in wickets. After lunch South Africa lost their first wicket at 158 — Marques Ackerman out, the opening stand broken, the score 158/1. What followed was not a long partnership but a cascade: six wickets were gone by 232.

One thing is worth recording here, because it explains the architecture of the innings. A side that was 158/1 and reached 232/6 lost five or six wickets for 74 runs. That happens when the pitch is not actually poor but the middle order is soft — Markram, Bavuma, Bedingham, Mulder, four names falling in the same shape. Had the surface genuinely seamed, 140/0 would never have existed; had the batting had depth, 232/6 would never have existed. Both truths hold at once: this was a bimodal innings — 158 runs for the first wicket, drought, a 95-run floor, and then four wickets in six balls. That is a good pitch plus a fragile middle order, not a hostile surface.

Then came the 95-run seventh-wicket stand. It carried South Africa past 300, and it is the least examined passage of the innings. If 329 reads as a triumph for Australia's bowlers, remember it was built from 232/6, not stopped there. When the lower order adds 95, the bowling story shifts from comeback to reprieve.

The reprieve closed in the final session. The second new ball became available after 80 overs, and Starc turned it into a weapon immediately. Left-arm pace, angled into the right-hander's stumps, a hard shiny ball — that is a Test-specific mechanism with no T20 equivalent. What seventy-four overs with the old ball could not produce, six overs with the new one did. Rickelton, Jansen, Coetzee, Rabada: all four fell to shine and angle, none to a magic yorker. The burst was not a miracle of the surface; it was a consequence of the ball's age.

Here the ledger makes me stop. I re-added the report's numbers by hand, because for three decades I have checked arithmetic before headlines. At 232/6 six wickets had fallen; Rickelton went as the seventh, Jansen the eighth, Coetzee the ninth, Rabada the tenth, and the total of 329 reconciles. But then only five wickets fell between 158/1 and 232/6, not six — which makes 'six wickets for 74 runs' wrong, or requires the Rickelton-Jansen stand to have been the eighth wicket. Only one of those can be true. A wicket is also simply missing from the list: Ackerman, Markram, Bavuma, Bedingham, Mulder — five names, and the sixth is unnamed. A report whose wickets do not add up is not data; it is a draft.

The headline will not go here. Calling 329 all out a brilliant Australian comeback is easy, but on a pitch described as batting-friendly, 140/0 to 329 is a South African failure rather than an Australian feat. A total around 330 in the first innings is below expectation on that surface, and it was reached only because the seventh wicket added 95.

Second, 'Australia are in control' is half a sentence. By stumps Australia had not scored a run or faced an over. Control in a Test is the currency banked as a first-innings lead; everything else is projection. Cummins' toss decision deserves the same discipline — it looked wrong at lunch and right at stumps. That is outcome validating process, not process earning it. A high-variance call was rescued by a final-session collapse, no more.

Third, and most important to me. The report carries no date, no byline, no agency. Even the spellings wobble — Marques Ackerman appears more than one way, Nic Maddinson and Beau Webster arrive in shifted forms. And the document's own subject label does not match an Australia-South Africa fixture. I mark the corner of such a sheet in red. It does not mean the cricket is wrong; it means the numbers are unverified. The logic survives. The figures do not yet.

The first hour tomorrow is the most expensive hour of this Test. Australia must open against the same new ball that took four wickets, and must bat past 329 to bank a lead — because the surface left for Lyon and Connolly on days four and five will not forgive a chase. The question is not whether Starc took four wickets in six balls. It is how many runs Australia turns those four wickets into.