From the Khulna Ledger to the National XI: Three Regimes, One Wrong Column

I opened the Khulna ledger, and the first column taught me patience.

Over the last three rounds of first-class matches at Khulna Division, spinners averaged 41.2 in the fourth innings and 23.6 in the second. Same pitch, same ball, same daylight, only the innings order changed. Sitting in the stands at Sheikh Abu Naser Stadium, I counted 1,147 deliveries by hand across those three matches and logged the outcome of each one. The ledger is blunt: spinners did not become easier to face in the fourth innings, the batters became easier to bowl at. Deliveries that drew a bat coming down in the second innings drew a bat rising toward the pad in the fourth. Wickets fell less often, dot balls rose, and the run rate fell to 2.87 per over. The number is less a certificate for the bowler and more a record of batting withdrawal.

The National Cricket League is Bangladesh's first-class competition: eight teams — Dhaka, Dhaka Metropolis, Chattogram, Khulna, Rajshahi, Rangpur, Barishal and Sylhet — playing four-day matches home and away. These scorecards are the primary paper of national selection, and that is where the trouble starts. The same paper is used to pick three different teams: the Tests from the four-day column, the ODIs from the List A column, and the T20 side from the BPL's 20 overs. Three separate regimes, three separate sets of demands. In conversation they get merged into one, and that is exactly where wrong decisions are born.

I learned that distinction from football. At the 2026 World Cup, France's PPDA moved from 8.2 in the group stage to 14.6 in the final — they pressed less and won. The France PPDA map was not a picture; it was a confession of where they chose to press and where they deliberately left space. A ball-by-ball pressure map in cricket is the same confession.

NCL grounds are nearly empty, many matches have no ball-by-ball data, and scorecards sometimes lose a row. When the stadium emptied, I audited the silence and found the game still breathing — nobody was simply writing it down. A missing row and a missing event are not the same thing. Between those two, I built three separate ledgers for Khulna Division.

Ledger one, the four-day regime. The only currency here is patience. The first indicator is balls per wicket, the second is balls per wicket after the 60th over, the third is sessions won. Economy is secondary. A spinner who bowls 40 overs without a wicket while forcing the inside shoulder into play on two balls an over is still banking value. One Khulna left-arm spinner has 34 wickets at 22.4 this season, one wicket every 58 balls, and one every 42 balls after the 60th over. In the fourth innings he concedes 2.1 an over. Three numbers together say he is an asset in this regime.

Ledger two, the 50-over regime. Patience is discounted; middle-over rotation is the requirement. The indicators are the run rate from overs 11 to 40, boundary percentage, and runs per over outside fielding restrictions. A Khulna top-order batter averages 51.2 here at a strike rate of 78.4. In the four-day ledger the same man made 744 runs at 41.8, which is rewarded there and punished here.

Ledger three, the 20-over regime. Indicators: powerplay dot percentage, death-over economy, left-hand/right-hand matchups. The same left-arm spinner has six wickets in 11 T20 matches, an economy of 8.9, and a powerplay dot rate of 31 percent. In this regime he is a cost, not an asset. One man, two seasons, two sets of metrics, and not a single contradiction between them.

A crossover note belongs here: the same cricketer can succeed in two regimes, but that is the exception, not the rule. Pick a player in the column where he is large. In September 2026 at Sher-e-Bangla National Cricket Stadium in Dhaka, Bangladesh won their first-ever T20I series against New Zealand; the foundation was slow, low pitches and spin. That was a regime-specific win, and it did not translate on different surfaces. Taijul Islam and Mehidy Hasan Miraz exist because red pitches here teach a bowler to labour for four days; the same pitch punishes him over 20 overs.

This is where the selection cycle breaks. Selectors read one column and write the verdict in another. Take the Khulna spinner's four-day average of 22.4, place him in a 20-over side, and his economy of 8.9 becomes an indictment. That 8.9 is not a failure; it is the normal bill of a different regime.

Consider the reverse. In 2026 I tracked 14 matches of Abahani Limited Dhaka and Sheikh Jamal Dhanmondi Club for a Dhaka site; Abahani scored 28 goals from 21.4 xG. I published a regression warning, and three of their next five matches were drawn. That was finishing above its own means. A batting average behaves the same way: a player who keeps runs down without spending balls per wicket will eventually regress.

Now the part the ledger cannot say. The fourth-innings average of 41.2 may have nothing to do with the bowler. Pitch wear, declaration timing, day-three heat, the chasing side's target mentality — any one of these can flip an entire column. My sample is three matches and 1,147 balls. On small samples you do not buy conclusions, you only accumulate doubt.

The weakest column in my ledger is 'rest'. A bowler is rested from a first-class round, appears three weeks later in a franchise exhibition, and the ledger writes 'rest' in both places. Much of the romantic writing about load management is really the rearranging of a touring calendar around commercial matches. The bowler's body is not in my column, which is why that column carries my lowest confidence.

For the next round I have set three triggers with confidence tiers. High confidence: if the Khulna left-arm spinner's balls per wicket after the 60th over stays under 45 across the next two rounds and his fourth-innings dot rate stays above 55 percent, the four-day column is real. Medium confidence: his T20 economy needs at least 15 matches, not 11, before it can be read below 8.5. Conditional: if selectors again use a four-day average to pick a 20-over side, the error is structural, not personal.

A clean row of data will outlast a thousand hot takes.