Nine Wickets, Still Behind: The First-Class Numbers That Are Not Test Numbers

For the past few years I have kept a ball-by-ball log of tour matches — not scorecards, an over-by-over event file in which every delivery's line, length and the batter's footwork are tagged separately. At stumps on day three of this match, the loudest signal in that file was not a shot. It was two numbers: Mohammed Siraj 4/118 and Mukesh Kumar 5/87. Nine wickets between them, and India A still behind on first innings. Before that sentence travels any further, one condition has to be accepted: every figure above is a first-class record, not a Test record. This is an unofficial Test — a four-innings, multi-day tour match, a warm-up for the two-Test series reported to begin on November 19, 2026.

The structure of the fixture has to stay in view. Results here are not guaranteed, the frame is four-innings first-class cricket, and the opposition is running its own experiments. At stumps on day three, India A were batting in their second innings; the game is live, and the first-innings lead is the most expensive asset on the table, sitting in New Zealand A's pocket. I hand-coded 380 League One matches before I trusted the model, and the habit that survived is a simple one: before I use a number, I write down its format, its date range and its source. Empty stadiums taught me to measure what crowds conceal; in a tour match there is no crowd at all, which leaves pitch age and ball age as the only honest variables. Fill those three boxes here and the first warning arrives in the name of the format, not of Siraj or Mukesh: reading a first-class wicket as a Test wicket is pricing two different currencies at the same rate.

Strip the arithmetic back. Siraj's four wickets cost 118 runs, or 29.5 apiece. Mukesh's five cost 87, or 17.4 apiece. The gap is 12.1 runs per wicket — more than 70 per cent. The nine-wicket headline is a sum, and a sum buys two bowlers' work at a single price when the work was never priced the same. A selector reading only the addition is reading an uneven investment as an even return.

The picture sharpens when you look at the stand. New Zealand A's innings rests on a 208-run fourth-wicket partnership between Boyle and Foxcroft. The wicket column is not a progress bar; it is a survival curve. A flat 208-run line suddenly went vertical, and that vertical stretch is what later appears on the card as nine wickets. Which raises the question that matters more than the total: when did the wickets actually fall — inside the new-ball window, or once the innings had already slowed on its own?

The tail deserves separate treatment. Clarke 41, Sheat 40, Lennox 32 — 113 in all. That is the largest block of the innings after the 208-run stand, and it arrived precisely in the phase when India A's job was to close the innings. Tail runs in first-class cricket carry a different price: older ball, lower tempo, a field set for spin. I do not have the ball-by-ball split for those three innings in front of me, so the claim stays bounded. What can be said is this: the 113 from the tail is a large slice of the first-innings lead, and it came after the fourth wicket fell — which means India A's best bowling phase ended in a leak.

One more measure. The nine wickets cost 205 runs in total, 22.8 apiece. Siraj sits 6.7 runs above that team average per wicket, Mukesh 5.4 below it. The small spread is the point: nine wickets are not nine events of equal value — five came out of the fault line in the middle of the innings, four came from around the tail.

An old memory attaches itself here. In the 2026-20 season my survival model gave Charlton Athletic a 71 per cent relegation probability unless they raised their defensive line. The recommendation was declined; they went down 22nd on 48 points. The spreadsheet knew the relegation before the stadium did. This match reads the same way: the scoreline says first-innings lead, the per-wicket split says something else entirely. For the Danish FA at Russia 2026 I delivered 41 pre-match briefs, each capped at 400 words and one chart. The cap teaches the discipline: no more than three numbers in a paragraph, claim first, evidence second, caveat last. A 400-word brief can hide a thousand hours of silence — a scorecard hides more.

Now the part where I argue against my own numbers. A five-for is the easiest certificate in the world to hand out, and the least earned. In a tour match the opposition's intent is different: they treat the innings as practice, trial new shots, and occasionally bat deliberately slowly to test a bowler. Pitch age, ball condition, field settings and the absence of DRS in an unofficial fixture mean a first-class wicket count is not directly comparable with a Test wicket count. Correlation is not causation. The real examination is whether a bowler who took four with the older ball can produce the same with the new one — and that examination is still pending. Credit where the model holds: after the 208-run stand was broken, the bowling that kept the run rate in check was the clearest plus on the day, and both bowlers shared it.

So I pre-register the decision threshold rather than inventing one later. If, in the second innings, the per-wicket gap between Siraj and Mukesh narrows from 12 runs to six or seven, and the tail is held under 40, I still will not write 'ready'. I will write 'two independent signals pointing the same way'. The question now is whether the selectors for the series beginning on November 19, 2026 will read the sum, or the split.