Sam Konstas's 149: three innings inside one, and the selection question nobody is asking

24 October, the WACA. One-Day Cup, New South Wales bat first. Sam Konstas makes 149 off 131 balls, dismissed in the 47th over. Fifteen fours, four sixes, strike rate 113.7.

That is the line that became the headline the next morning. It is also the line that, on its own, proves nothing.

Cut the innings into three pieces. First fifty: 59 balls, strike rate 84.7, reached with a six. Second fifty: 47 balls, strike rate 106.4. Final 49: 25 balls, strike rate 196.0. One scorecard, one name, three different batters. The Konstas who batted through the first hour at 84 went at 196 through the last half hour. Which of those three is the selection panel watching?

I coded 34.1 overs by hand in the Eden Gardens scorer's box on 13 March 2026 and learned that night that a scoreline published without a pressure state is typesetting, not reporting. The rule holds thirty years later. Pressure is a logged variable; outcomes are just noise.

Here is the number nobody printed. Konstas made 149 of New South Wales's 371 — 40.2 per cent. The other ten batters made 222. If NSW batted the full fifty overs — which 371 and a 47th-over dismissal strongly imply — they had 300 balls, of which he faced 131, or 43.7 per cent. The remaining 169 balls produced 222 runs. Strike rate 131.4.

Konstas's strike rate in that innings sat below his own team's. That single line breaks the "he dragged the innings alone" story. But I do not publish point estimates. The unknown here is extras. The claim survives extras up to about thirty and collapses beyond that. I do not have the extras figure, so I publish the bracket, not the verdict.

What a scorecard can give you

Counted: 131 balls, 149 runs, 19 boundary balls (15 fours, four sixes), three phases (59/47/25 balls), three stands (81, 85, 82), and NSW's 371.

Not counted: ball-by-ball. So no control percentage, no false-shot count, no spell map, no par score at the WACA for this competition, no strength rating of Western Australia's attack. And one provenance flag — the source attaches "off 89 balls" ambiguously to Josh Phillipe's 51 or to the 81-run stand. I read it as Phillipe's own innings (strike rate 57.3). If it belongs to the stand, a third of my argument falls over immediately, and I would rather publish the assumption than the inference.

The partnership algebra

NSW lost Riley Kingsell in the second over, so Konstas walked in at No. 3 and batted roughly forty-five overs, second over to the 47th. In that time, three stands: 81 with Phillipe, 85 with captain Lachlan Shaw, 82 with Jack Edwards. That is 248 runs in which Konstas was one end.

Two readings follow. One: he anchored three partners, lasted 45 overs and finished at 196. That is a mature two-speed one-day innings. Two: Phillipe went at 57.3, Shaw made 44, Edwards 45, and the rest of the side's collective rate sits near 131 — meaning when Konstas finally accelerated, the team was already going faster than him. Strike rate against the bowling actually faced was the only column that mattered at that auction, and it is the only column that matters here.

The weight of nineteen balls

Eighty-four of the 149 came off 19 balls (4.42 runs per boundary ball). The other 112 balls produced 65 runs at 58.0.

Had five of those nineteen boundary balls gone to hand — statistically unremarkable in an innings that lasts to the 47th over — the knock becomes 110-odd and there is no headline. Selection should not turn on nineteen balls of luck.

In 2026 I coded 412 one-day internationals India played between 2026 and 2026. The database had 412 matches; the byline had someone else's name. The real question was always which part of an outcome would not recur. Here, that part is nineteen balls.

The map of four fifties

The claim everywhere is "four consecutive 50-plus scores." Open the list: 67 against India A in Puducherry, 71 for NSW against Western Australia at the WACA, 151 not out at the same ground against the same side, and 149 at the same ground against the same side on 24 October. 438 runs, three dismissals, average 146.0.

That is not a form graph. It is a single-venue, single-attack cluster with one innings abroad bolted to the end. Three of four at the WACA. "Consecutive" is not a property of the batter; it is a property of the list somebody built. What was dropped from the sequence is never printed.

And 371 has no meaning without a par score for the venue and competition. Without a benchmark, reading a total is like reading a share price with no index.

"Second-highest total ever" for NSW deserves the same scrutiny. "Ever" is not an achievement of a batter; it is an editorial decision by whoever compiled the archive. The 1970s scorecards that were digitised late are the foundation of that record, and the scorer who logged them has no byline.

The Test ledger: 163 runs, ten innings

December 2026, Boxing Day, Melbourne. Sixty against Jasprit Bumrah on debut. I watched that from Kolkata and wrote in my diary that night: this is evidence of a floor, not a ceiling.

Then 103 runs in nine innings. 163 runs across ten innings. Average 16.3.

Those sixty runs against Bumrah have lived longer in public memory than the 103 that followed. Both belong to the same sample. The sample says 16.3, and it also says the sample is small. Fifteen months out of the side since the West Indies tour is the context, not the correction.

Access is not evidence

Australia's top order is shaking. Nic Maddinson is struggling on return from South Africa, Marnus Labuschagne's problems persist. David Warner has publicly backed Konstas at No. 3.

Warner's endorsement is an opinion, not a metric. A former player carries weight in the media ledger and zero in the selection ledger until the panel publishes its weights. In 2026 I handed a method to an investigative reporter, not names. It is a habit I have not broken.

The counter-ledger

I do not have Labuschagne's numbers for this window in front of me. I do not have Maddinson's South Africa series ball-by-ball. So I cannot say picking Konstas is right or wrong. I can only say that the comparison currently being made is being made on an incomplete list. A panel does not choose between players; it chooses between error types — a false positive on a hot streak against missing a 22-year-old in form.

What to watch

If he is picked for New Zealand, log three things: his batting position, his first thirty balls of the first innings, and whether the panel publishes Labuschagne's numbers with the same rigour. His domestic evidence is thinnest in exactly that new-ball phase — his first 59 balls at 84.7 were solid, not aggressive, and Test cricket asks for solid. The workload question is separate: Puducherry to Perth to Shield to one-day cup to a possible Test squad is a calendar no physio can fix.

Fifteen fours and four sixes go in the newspaper. The counts 59, 47 and 25 do not. Selection is always decided by the column nobody printed.

Method appendix. Sample: one innings (149/131), one cluster (four innings, 438 runs), one Test baseline (ten innings, 163 runs). Computed: phase splits, partnership sum 248, share of total 40.2%, share of balls 43.7% (assuming fifty overs). Exclusions: no ball-by-ball, so no control percentage, no spell map, no par score, no attack strength rating; extras unavailable, so the team-rate claim holds only to about thirty. What would falsify me: a 120-plus at 90-plus strike rate away from the WACA in his next innings, and I will publish the correction.