From the Khulna Press Box to the Mirpur Floodlights: The Invisible Numbers of Bangladesh Cricket

A morning from last season's National Cricket League is still stuck in my notebook. Sheikh Abu Naser Stadium, Khulna, first day, second session. I was logging ball by ball, and the scorebook showed me a number that never made a headline that day. Overs 32 to 50 — across those eighteen overs the batting side's boundary percentage was 11.4, almost exactly what it had been in the previous eighteen. Yet the run rate fell from 5.8 to 3.2. The batsmen were not getting out, they were not missing the ball, no wickets fell. Still the scoreboard went quietly cold.

That day a senior journalist sitting in the next chair in the press box told me, "The pitch has died, that is the real story." I closed my notebook and pulled out the fielding map. The pitch had not died. The field had moved.

The spreadsheet was my daily prayer; the data, my daily office. And the first condition of entering that office is this — the story comes after the model, never before it.

Context: the country where cricket is played, and the country where it is read

There is a permanent crisis in how we write about Bangladesh cricket. We watch the game very deeply, but we measure it very shallowly. In domestic cricket we have almost no ball-by-ball data store. Four-day National Cricket League matches, fifty-over Dhaka Premier League games, twenty-over Bangladesh Premier League games — three different environments, and yet we build one story for all three out of roughly the same basic statistics.

When I built my first shot-level model for the Bangladesh Premier League from Khulna in 2026, it was not a job requirement. It was stubbornness. Because I had watched, sitting in the Khulna press box, two journalists read the same match in completely opposite ways — one said "batting failure", the other said "suffocating bowling". Both were right, both were wrong, because neither had the dot-ball location, the field set, or the wicket-value calculation.

Three ground realities should sit at the centre of our study of Bangladesh cricket. First, the slow, low-bounce surface at the Sher-e-Bangla National Cricket Stadium in Mirpur — where after the fourth over the ball begins to fold into the spinners' hands. Second, the Khulna and Chattogram surfaces, where seam in the morning and spin in the afternoon make two different games on one ground. Third, our weather. The heat of April, the dew of October, the fog of December — three seasons that create three different games for the same team.

Nobody denies these three realities. But nobody measures them either. And the cost of not measuring is simple — we explain matches through player names rather than through environment. Yet in Bangladesh domestic cricket, environment is the biggest selector of all.

The powerplay: a two-faced number

Let us begin with the first six overs, because this is where our reading goes wrong. In world cricket the average powerplay run rate currently hovers between seven and eight. When I stitched together seven seasons of ball-by-ball data from Bangladesh domestic cricket, the number came out two-faced — the ball moves plenty in the first eight to ten deliveries, but once a batsman settles, the run rate suddenly swells.

In Bangladesh domestic cricket the real currency of the powerplay is not the opening pair's strike rate; it is the over in which the second wicket falls. In my log, teams that lose two wickets inside twelve overs finish their innings at an average of 238 (in a twenty-over game). Teams that reach twelve overs with one or no wicket down average 274. That thirty-six-run gap is not tied to the batsman's name; it is tied to how many overs the set batsman was allowed to bat.

Here is my controversial observation. In Bangladesh domestic cricket we select openers by personal strike rate, when the match is actually decided by their miss-proof strike rate. The same batsman scores 18 off 12 in the first ten balls — a 150 strike rate, excellent. But if he plays six dots in those ten balls, the team has wasted fifty per cent of its resource in ten deliveries. In my model, one dot ball on a Bangladeshi surface costs an average of 1.31 runs — meaning every six dots costs a side roughly eight runs, which usually never appears on the scoreboard, it is merely absent.

I built the model in the Khulna press box, then let the league speak. Seven seasons of accounting say the powerplay is less about boundaries and more about the quiet traffic of ones and twos between an opener's dot balls. Those who understand this game will not say after a match, "the batsman batted slowly"; they will say, "the batsman hit hard in the wrong place."

The middle overs: where the match is actually lost

Overs seven to fifteen — in my notebook I have written these eight overs down as "the zone of silence". Boundaries are not rare here, wickets are not rare either, but the run rate is easiest to control. Because this is where spinners bowl, this is where fielders come inside the circle, and this is where the Bangladeshi pitch shows its true character.

In Bangladesh domestic cricket, overs seven to fifteen are the window in which eighty per cent of fielding captaincy should be evaluated, and yet only ten per cent is. When I drew the middle-overs dot-ball map, the dots did not fall randomly — they came from specific positions. A silly point pushed inside square leg, a deep midwicket set slightly above long-on, and a low-catch specialist at slip — these three fielding positions together create roughly four to five extra dot balls per over on a Bangladeshi surface.

Here my model raises an uncomfortable question. We judge Bangladeshi domestic spinners by wicket count. But in my seven seasons of data, many of the best economy spinners take fewer than two wickets a match — yet in the overs they bowl, the opposition run rate is one and a half to two runs lower. Their value hides not in the first column of the scorebook, but in the third (overs) and in the fielding map.

And one thing my notebook shouts again and again — the gap between a right-hander's strike rate against left-arm spin and a left-hander's strike rate against right-arm spin is far larger on Bangladeshi pitches than the world average. Because here the ball does not turn much; it frustrates more — it comes slowly, turns a little late, and the batsman's hands go early. This "spin of frustration" is our own resource, and we do not measure it.

I trust the model, but I audit the story it tells. This middle-overs silence is no coincidence. It is the result of a deliberate design, or of the absence of one. A side that sends a fielder to the boundary in the middle overs concedes four to six extra runs in the next ten overs. The number looks small; across four matches it accumulates and turns a whole points table.

The death overs: the true price of a wicket

Overs sixteen to twenty. In Bangladesh domestic leagues the run rate climbs to nine and eleven here, and every match's emotion collects in this pocket. We measure the death bowler by economy, the death batsman by strike rate. Both are incomplete measures, because in these five overs the biggest currency is the time-exchange of a wicket.

The best death bowler is not the one who concedes least; the best is the one who knows how to buy the right to concede with a wicket at the right over. In my model I calculated that a wicket falling in the 17th over saves a side an average of 7.4 runs, while a wicket in the 20th over saves only 3.1. Because in the 20th over the ball is nearly done; after the wicket the new batsman gets only one or two deliveries. Yet in the Bangladeshi press box a 20th-over yorker-wicket draws the loudest praise. The praise is not wrong, but it is misplaced.

A practical conclusion follows. The "specialist death bowlers" that Bangladesh Premier League sides buy are often evaluated in the final over. But the bowler who takes two wickets in the 16th and 17th overs and makes the opposition score ten fewer runs shows his value in the next over, not his own. Data first, then the name.

Pitch, dew and light: Bangladesh's three selectors

Now back to the three realities we acknowledge but do not measure.

Mirpur. Here the first session of the day is not the hardest — the hardest is the second and third, when the top layer of the pitch dries and begins to take spin, while the bounce drops even lower. My log shows that batsmen batting after position four in Mirpur average roughly fourteen strike-rate points below the top-order batsmen. This is not a variation in batsman skill; it is a punishment of position.

Khulna. At Sheikh Abu Naser Stadium, morning seam and afternoon spin — two different games on the same day. The consequence of this duality is that the toss-winning side almost always chooses to bat first, but once the morning moisture is gone, that becomes a trap. In my count, the advantage a side batting first gains in the first ten overs in Khulna is outweighed by the three or four wickets it loses in the middle of the session.

Chattogram. The pitch here is comparatively batting-friendly, but sea breeze and humidity keep the ball damp. On this ground a death-over yorker is least effective, and a slower ball most. We do not write these ground-specific differences into match previews; before the toss we only say "the pitch is good".

Three grounds, three different games. Yet our domestic cricket analysis is roughly one. This is our biggest failure — we measure cricket, not environment.

Fielding: the runs that never reach the scoreboard

One line I have heard again and again in the Khulna press box — "we are not good at fielding." The line is true, but nobody has ever proved it with numbers. I tried.

A fielding metric does not really mean catches held, but how many catching chances are created — that is, whether a fielder is cutting off the ball's path in advance. In my seven seasons of logging, a session-tracked match in Bangladesh domestic cricket produces on average three to four "half-chances" that never register as catches, because the fielder never reached them. Behind these half-chances lies the fielder's initial position — his one or two steps before the ball is released.

And one more thing — the pressure of the throw. Run-scoring is slow on Bangladeshi pitches, so run-outs are few. But if a boundary fielder's throw does not arrive on top of the stumps, a single cannot be stopped, and the batsman takes another run the next ball. These "invisible two runs" accumulate across an innings and make the difference in a match. They never appear in fielding statistics.

Contrarian: the pitch did not die, the story did

Now back to that morning in Khulna, overs 32 to 50. The senior journalist said the pitch had died. I say the pitch did not change one bit — the field changed.

In the first eighteen overs, deep cover and deep point were near the boundary, so batsmen could hit over cover. In the next eighteen, two fielders moved four steps in from the rope, and one came from slip to the leg side. The boundary percentage stayed about the same, because batsmen were still playing shots — but the shots no longer fell into empty space. The run rate fell because the easy road to runs was closed.

Here we must accept an uncomfortable truth. A fall in run rate in cricket is not always a batsman's failure; often it is the success of fielding geometry — and we have not learned to tell the two apart. Match reports say "the batting collapsed", when the fielding map shows the collapse was not caused — it was arranged.

There is another danger, one I have seen in my own work. The more data we get, the more we want to turn everything into numbers. But Bangladesh domestic cricket contains variables no model captures — the pressure of an age-group side, family expectation, the mentality of playing through injury, or the guilt of playing at home. Build a model without these variables and the model is not wrong, it is incomplete. And explaining a match with an incomplete model means we quietly do an injustice to the cricketer.

The press box taught me humility: noise is data too. Without that humility, data becomes a telescope, not an eye.

Takeaway: a signal for the next round

In the next phase of Bangladesh domestic cricket I will be looking for one specific signal. Not a batsman's name, not bowlers' average speed. It is this — how many fielders inside the circle move within four steps in the middle overs. My model says this one variable will decide the gap between the middle three teams in the points table next season.

The question is this — do we have the courage to look outside the scoreboard, or will we pass another season writing "the pitch has died"?

A closing word

On the road to becoming big, Bangladesh cricket's biggest obstacle is not talent. It is attention. We hunt talent, but we do not measure the environment in which talent grows. A Khulna morning, a Mirpur second session, the humidity of Chattogram — these three places are the real factory of our cricket. I stand there and gather data, because you cannot price a product without understanding the factory. And in cricket the product is runs, the currency is overs, and the profit and loss is written in the invisible lines of the fielding map.