Auditing Selection from the Khulna Ledger: Bangladesh Cricket's Three Regimes and One Wrong Sum

A November morning at Sheikh Abu Naser Stadium. A first-class match in the National Cricket League, first session of the second day. The stands can be counted; I counted twenty-seven people, twelve of them local club coaches. A right-handed middle-order batter from Khulna Division was on 29 off 87 balls, a strike rate of 33.3. A young reporter beside me whispered that the man was doing nothing at all.

The column I was watching was not the runs. It was balls per dismissal, the number of deliveries a man survives in an innings. His previous four innings read 121, 64, 143 and 97, and in none of those four did his strike rate cross 45. That morning he stopped on 41 off 146. The scoreboard wrote failure. My ledger wrote something else: the partnership he built with Khulna's number four, across those 146 balls, was what gave the side a first-innings lead.

I opened the Khulna ledger, and the first column taught me patience. When the stadium emptied, I audited the silence and found the game still breathing.

Bangladeshi cricket runs four separate regimes at once, and our selection talk keeps reading one in the language of another. Test cricket is one regime, where balls per dismissal, overs per innings and fifth-day reserves carry value. One-day cricket is a second, balancing scoring rate against wickets in hand across fifty overs. T20 is a third, built on boundary balls. The fourth is franchise cricket, which keeps its own clock, its own contracts and its own travel schedule.

In the domestic structure the only long-format ledger is the National Cricket League, where Dhaka, Khulna, Rajshahi, Chattogram, Sylhet, Barishal, Rangpur and Dhaka Metropolis play as separate sides. The season usually squeezes into October to December, and December to February belongs to franchise T20. So a batter who just spent three months writing a ledger of patience is asked, over the next three months, to take a risk every fourth ball with the same hands.

The real problem is not that one format is better or worse. It is that we are making one regime's decisions with another regime's numbers.

The first question is always about sample. A batter who averages 39.2 at a strike rate of 48.7 across 24 first-class innings, and averages 24.6 at a strike rate of 138.4 across 15 franchise innings in the same year, sounds like two different men. In reality it is one man honouring two different contracts. In 2026 I tracked Abahani Limited Dhaka and Sheikh Jamal Dhanmondi Club across fourteen matches for a Dhaka-based football site; Abahani scored 28 goals from 21.4 xG. I published a regression warning and they drew three of their next five. That habit has stayed with me, and it is now my first rule in cricket: ask which regime the number belongs to before you decide who the man is.

Recognising the regime is not enough on its own. Inside a single regime you have to count against the clock. Test bowling load and T20 bowling load are not the same object, even though the domestic calendar files them into the same week. A first-class seamer bowling 22 overs in an innings is ordinary; in T20 his maximum spell is four overs. Training, recovery, travel and physio paperwork all differ. Yet in the selection room the two numbers sit side by side, because both fall under the same name in the same database.

Take one concrete case. Bangladesh won the ODI series against New Zealand 3-0 in Dhaka in September 2026, and in that series at Mirpur Shakib Al Hasan took his 700th international wicket, becoming only the third bowler to reach that mark. It is a large number, but what does it prove? It proves that two decades of work across three formats, accumulated in one place, builds a pillar. It does not prove that the same bowler can carry an equivalent load in a Test the following month. That September his longest ODI spell was ten overs, and the domestic and franchise calendars pulled at him almost immediately.

Another case, from 2026. In August and September 2026 Bangladesh beat Pakistan 2-0 in a two-Test series in Rawalpindi, their first Test series victory on Pakistani soil. In the first match Bangladesh made 565 across five rain-affected days; in the second they won by six wickets. Across those two matches Bangladesh's seamers used seam and reverse seam on lengths that a four-over T20 spell never teaches. Ask the question and you find the spells were built on long overs counted in first-class cricket, not on franchise economy rates.

Consider Nahid Rana. When he arrived in the national side in 2026 the conversation centred on his pace, and pace is an extraordinary sensation. But his value in Test cricket is set by repetition, spell over spell, and that is not manufactured without 20 to 25 overs of domestic first-class work. Those who only watch his highlight reel are watching a different cricketer.

Some of the talk around workload management is necessary. A larger part of it is staged. Rest is often announced in exactly the week a multi-nation series or a commercial tour begins, and the scientific basis for that rest never appears in a public ledger. If rest really is the output of a physical calculation, then spells, deliveries per week, training volume and recovery days should be published as four columns. As long as those columns stay hidden, the phrase will remain the name of a closed column rather than the name of a medical record.

When I built the France PPDA model in 2026 I learned one thing: that map was not a picture, it was a confession of where they pressed and where they let go. Cricket's balls-per-dismissal column is the same kind of thing. It is a confession, not a verdict.

The contrarian angle is here. The easy story is that domestic first-class cricket solves all our problems. But correlation and cause have to be separated. It is true that men who bowl and bat many first-class deliveries do better in Tests. The cause may not be the volume of balls; it may be age, physical maturity, or simply when the opportunity arrived. The number of first-class matches played each season in Bangladesh is not enough to build a full load. Domestic pitches tend to favour spin, leaving little room for seamers to learn. A larger number can therefore accumulate the wrong kind of experience.

There is another trap: domestic economy rates. An economy of 2.9 in first-class cricket is a good figure, and it does not transfer to T20. Batting intent differs, fields differ, the age of the ball differs. A bowler who took three catches at fourth slip in a first-class match will not find a fourth slip in a franchise game.

The audit gets easier the moment we admit that four ledgers must stay open at once, with crossover notes running between them. A clean row of data will outlast a thousand hot takes, but it will not be one row. It has to be four rows read together.

From years of sitting in grounds watching the game, I would say the least discussed signal does not come from fitness or from a referee's call. It comes from the missing row. A domestic scorecard never carries a bowler's spell-by-spell load, so we do not know who was carrying what in which over. When the stadium empties, that absence is not a political statement. It is an institutional one.

Watch three signals over the next round. One, whether the spells-per-innings figure for Khulna and Rajshahi's first-class seamers has risen or fallen over the last three seasons. Two, the balls per dismissal in the first two innings of anyone who walks straight from the franchise league into the Test side. Three, whether the domestic calendar increases the number of days per match rather than the number of matches. The third signal will carry the most information.

The number nobody is writing down yet will sit at the centre of the next selection argument.