The Dot-Ball Ledger: Auditing Bangladesh's T20 Middle Overs Before the 2026 World Cup

Zahur Ahmed Chowdhury Stadium, Chattogram. A December evening in 2026. Bangladesh are chasing 164. At 13.2 overs the score reads 84 for 3 — seven wickets in hand, 80 runs needed, 40 balls left.

The next 22 deliveries produce 17 runs. Five dot balls in a row, split across two overs. Of those 22 balls, six go straight to fielders, four are merely pushed to rotate strike, and three are left alone.

The scoreboard records that passage in four lines: 13.2 — 84/3; 17.4 — 101/5.

My ledger records it in seven columns: ball number, bowler, line-length zone, shot type, contact quality, field-pressure index, and expected runs. The expected runs for those 22 balls were 28.4. The actual return was 17. The shortfall was 11.4 — in one match, in one phase, in one window of pressure.

That 11.4 is the subject of this piece. Because before the 2026 T20 World Cup, Bangladesh's largest numerical loss is not in the powerplay, and not at the death. It happens between overs 7 and 15 — silently, every match, nearly every over.

CONTEXT: WHY A LEDGER, WHY NOW

Between January 2026 and December 2026, the Bangladesh men's team played 47 T20 internationals. I hand-coded every ball of those 47 matches — 11,084 legal deliveries. For each ball, six variables were logged: bowler type (pace or spin), line-length zone, batter's hand, shot type (drive, pull, sweep, cut, defend, leave), contact quality (middle, edge, miss), and whether there was defensive pressure in the field.

The expected-runs model is deliberately simple. The base rate for each ball is drawn from three layers: the global average outcome for that line-length and shot combination, the historical scoring pattern at that venue, and the average outcome for a batter at that stage of an innings with that many wickets in hand. The three layers are averaged, because using only one layer destroys venue control, and without venue control a 140 at Mirpur and a 170 at Sylhet sit in the same column.

I keep clean columns so the messy truth has somewhere to land.

Two caveats must be stated plainly. First, this is my own coding, not an official database. Every figure therefore carries an uncertainty interval. Forty-seven matches is a medium sample, not a large one. Second, this ledger makes no claim to explain results. It simply keeps ball-by-ball accounts so that someone else can later reconcile them. The ledger does not replace the match; it remembers what the match forgot.

When I built my first expected-goals ledger in Chattogram in 2026 — 22 Bangladesh Premier League matches, every shot charted by hand — one habit took root: accounts before narrative. Stories are always entertaining. Accounts are never entertaining, and that is precisely why they are needed.

CORE ANALYSIS

1. THE POWERPLAY: THE PROBLEM EVERYONE SEES

Bangladesh's powerplay run rate was 7.42 in 2026 and 7.86 in 2026. That is improvement. But compare: over the same period India's powerplay run rate was 9.31, Australia's 8.94, Afghanistan's 8.52. Bangladesh are improving more slowly than the field.

Powerplay dot-ball rate is 51.2 percent. That looks alarming until placed in context: the global T20I powerplay dot-ball average sits between 46 and 49 percent. Bangladesh are two to five percentage points behind. This is the first myth worth breaking — the powerplay is not Bangladesh's primary problem.

Boundary rate in the powerplay is 16.8 percent, close to the global average. The issue is forced strike rotation. In the powerplay, Bangladesh batters change strike 1.9 times per six balls; India's figure is 2.4, England's 2.6. A small gap, but accumulated across 47 matches it produces roughly 900 extra dot balls.

1. THE MIDDLE OVERS: THE PROBLEM NOBODY SEES

Now the central number. Between overs 7 and 15, Bangladesh's run rate is 7.31. In this window no side wants to fall behind, because this is where set batters bat and boundaries come easiest — fielders are in, pacers return for their second spell.

Comparisons: India 9.12, Australia 8.87, Afghanistan 8.34, South Africa 8.69, Pakistan 8.01. Bangladesh 7.31.

Against India the gap is 1.81 runs per over. Across nine overs that is 16.3 runs. Across 47 matches it is roughly 766 runs.

That 1.81 is the centre of my accounting. It does not lose any single match. It is spread across 47 matches, so nobody notices. When a team scores 16 fewer runs in one match we say "the runs did not come today." When a team loses 766 runs across 47 matches we give it no name at all, because it is not an event — it is a condition.

1. THE ECONOMY OF THE DOT BALL

Bangladesh's middle-overs dot-ball rate is 39.8 percent. India's is 31.4, Australia's 33.2. That is roughly five dot balls per over. Across nine overs, 45 dot balls — nearly 45 balls producing zero runs, every single match.

Here a subtlety matters. Not all dot balls are equal. I divided them into three classes:

Class one — earned dots. A good ball the batter could not do anything with. This accounts for 38 percent of Bangladesh's middle-overs dots.

Class two — passive dots. The ball was not good, but the batter took no risk — neither played for a single nor attempted a boundary shot. Forty-four percent.

Class three — error dots. The ball was in the batter's shot zone but the shot selection was wrong. Eighteen percent.

The real story is in class two. Forty-four percent passive dots means that in the middle overs, Bangladesh batters decline to make a decision on roughly one of every two dot balls. Not deciding is itself a decision, and in T20 cricket it is the most expensive one available.

1. SPIN VERSUS PACE: THE WRONG QUESTION

A common belief in Bangladesh is that the side cannot play spin. The accounts say otherwise.

In the middle overs against spin, Bangladesh's run rate is 7.04 and strike rate 108.2. Against pace, 7.61 and 114.8. So there is a spin problem — but pace is not a solution either.

The real difference is in dot balls. Against spin the middle-overs dot rate is 42.6 percent; against pace 36.9. But against spin, earned dots account for only 31 percent — meaning that against spin, Bangladesh batters are more often undone by their own decisions than by good bowling.

The problem is not that spin cannot be played. The problem is that against spin in the middle overs, batters are not prepared to rotate strike, because they treat the single as a luxury rather than a risk.

1. THE BATTER LEDGER

Now the individual columns. Among those who faced at least 300 middle-overs balls across the 47 matches:

Litton Das: powerplay strike rate 132.4, middle overs 108.6. Excellent in the first six overs; from over seven his rate falls by 23.8 points, and his boundary rate drops from 16.4 to 8.9 percent.

Najmul Hossain Shanto: overall strike rate 118.9. Middle-overs dot rate 46.2 percent — six points above the team average. His earned-dot share is 41 percent, meaning the issue is tactical rather than purely environmental.

Towhid Hridoy: middle-overs strike rate 141.2, boundary rate 16.4 percent, dot rate 34.1 percent. One number stands out — he hits a boundary every 22.4 balls in the middle overs, against a team average of 34.7. He is the exception that proves the role: a sustained aggressor at numbers three to five.

Jaker Ali: middle-overs strike rate 128.7, death-overs 148.7. But his average middle-overs innings lasts only 14.3 balls — he cannot survive that phase, only exploit the last one.

Mehidy Hasan Miraz: middle-overs strike rate 96.4. With the ball his economy is 6.84, among the team's best. This creates a genuine strategic question: was sending a 96-strike-rate batter in that phase a batting decision, or a bowling-balance decision in disguise?

1. THE BOWLING LEDGER

There is uncomfortable news here that few want to state. Much of Bangladesh's middle-overs problem is not actually a bowling problem.

Bangladesh's collective middle-overs economy is 7.29 — better than the global average of 7.81. Rishad Hossain takes a wicket every 18.4 balls in that window, which is excellent. Mehidy Hasan Miraz's middle-overs economy is 6.84.

The bowlers keep the team in the game through the middle overs. The batting then fails to hold that line. In 19 of the 47 matches Bangladesh conceded more middle-overs runs than the opposition — meaning they fell behind with the ball — yet this never shows up in bowling statistics, because reading economy in isolation conceals the team's aggressive failure.

This is my strongest methodological objection to current discussion: we judge bowling by economy and batting by run rate, but the T20 middle overs are a joint account — one side's run rate and the other's economy are two faces of the same coin. As long as they sit in separate columns, nobody owns the outcome.

1. VENUE, DEW AND TOSS

No analysis is complete without venue control. First-innings averages in Bangladesh across 2026-25:

Mirpur (Sher-e-Bangla): 142.

Chattogram (Zahur Ahmed Chowdhury): 156.

Sylhet International: 161.

Bangladesh's middle-overs run rate is 8.14 in Sylhet and 6.79 in Mirpur. The team can partly hide the problem in Sylhet; it cannot in Mirpur.

Dew is a major variable. When dew arrives in the second innings, spinners lose grip and the ball comes onto the bat. Twenty-three of the 47 matches featured meaningful dew. In those, Bangladesh's middle-overs run rate was 7.94; in the other 24 it was 6.71.

So the middle-overs problem is also environment-dependent. But here is the danger: tosses cannot be controlled. What can be controlled is preparation for batting first after losing the toss. And that is precisely where Bangladesh are weakest — in 14 matches batting first at Mirpur, their middle-overs run rate was 6.38.

1. OPPOSITION-CONTROLLED COMPARISON

Looking only at your own numbers always misleads. So I ran an opposition-controlled sample: only matches against the eight Test-playing nations.

In that sample (31 matches), Bangladesh's middle-overs run rate is 7.18, dot rate 40.6 percent, boundary rate 11.2 percent.

Against Test nations, Bangladesh lose a middle-overs wicket every 29.4 balls — roughly one every five overs. India's figure is 38.2 balls, Australia's 41.6.

A subtle interaction appears here: Bangladesh bat slowly in the middle overs but also lose wickets more often. Slow batting usually exists to protect wickets. Here that is not happening. The strategy is internally contradictory — simultaneously conservative and unsafe.

That combination is what worries me most. A bad strategy can be fixed. A contradictory strategy is harder, because fixing it first requires admitting that two different objectives were pursued at once.

CONTRARIAN ANGLE: CORRELATION IS NOT CAUSATION

Now my objection to my own numbers.

First objection: sample size. Forty-seven matches, and the middle-overs sample drops to 31-38 once opposition controls are applied. At that size, the uncertainty interval around a 1.81-run-per-over gap is roughly plus or minus 0.4 to 0.7 runs. The direction is clear; the magnitude is not.

Second objection: selection instability. Across these 47 matches Bangladesh used 26 batters. Middle-overs statistics largely reflect who got opportunities and who did not, not only who performed. A 300-ball sample for one batter comes from six to eight innings; that is a shadow, not a trend.

Third objection: pitches. At Mirpur, first-innings average across 2026-25 was 142. Scoring at 9.12 an over on such a surface means 82 runs in nine overs — and chasing that costs wickets, which Bangladesh's middle-overs batting depth cannot absorb. India's number is therefore evidence of India's resources, not India's strategy.

Fourth objection: toss, dew and light are all outside team control, and all three directly affect middle-overs run rate. The dew-controlled gap is 1.23 runs per over — roughly two-thirds of the 1.81 I measured.

Fifth objection: there is a number here I cannot explain, and publishing it is my job. In 2026 the middle-overs run rate was 7.09; in 2026 it was 7.53 — an improvement of 0.44. Over the same period, powerplay improvement was 0.44 and death-overs improvement 0.61. Nearly identical gains in all three phases. Gains that uniform are usually a sign of accumulated experience, not strategic reform. The team is maturing, not changing.

One more point is needed, because the ledger metaphor is often misused. Cricket's ball-by-ball record is genuinely like an immutable ledger — each delivery a block, each over a chain, each new ball carrying the hash of the last. But as with a blockchain, what is written in the ledger cannot be altered; it can only be interpreted. My ledger of 11,084 balls is true. The conclusion "the team is bad" is mine, not the ledger's.

I also admit this: watching the middle overs from the ground, my eyes say the team is stuck. Watching counters from behind a camera, I think the balls were never in the batters' shot zones. Both can be true. The difference is that the eye's claim cannot be verified, and the column's can.

TAKEAWAY: THREE SIGNALS FOR 2026

Signal one: not the powerplay, but overs seven to fifteen. If Bangladesh want to invest in one phase before the 2026 T20 World Cup, this is the window. The target should be explicit: lift the middle-overs run rate from 7.31 to 8.20. That requires cutting the dot-ball rate from 39.8 to 34 percent — converting roughly 35 passive dots per match into singles.

Signal two: strike rotation while wickets remain. Of all the dot balls Bangladesh played in the middle overs across 47 matches, 44 percent were passive. That class is tactical, and tactical things change in practice. The question here is about habit, not talent.

Signal three: batting first after losing the toss. In 14 such matches at Mirpur, the middle-overs run rate was 6.38 — the team's weakest condition, and the condition the 2026 World Cup will most resemble.

A transfer analyst's first duty is to reconcile the story with the fee. In this ledger, the story is that Bangladesh are not suddenly losing in T20 cricket; they are slowly getting stuck. And what happens slowly shows up only in ledgers.

Before the 2026 World Cup I have one question left: if those 766 runs really are being lost in the middle overs, who gives them back — a new batter, or a new strategy? After watching Japan versus Belgium from the press box, I learned that pressure is just distance with a stopwatch. For Bangladesh the distance is 1.81 runs, and the stopwatch is running until February 2026.

APPENDIX: A REPRODUCIBLE TEMPLATE

Every number in this piece can be reproduced, and that is the central claim. The method is laid out in seven steps so anyone can run the same accounting for any team.

Step one: define the sample. Write the time window (here January 2026 to December 2026), the match count (47), and the opposition class (all teams versus Test nations only).

Step two: split the phases. Powerplay (1-6), middle (7-15), death (16-20). Do not move the phase boundaries — move them and prior accounting becomes incomparable.

Step three: code six variables per ball — bowler type, line-length zone, batter's hand, shot type, contact quality, field pressure.

Step four: build the expected-runs model from three layers — global base rate, venue base rate, innings-stage base rate.

Step five: split dot balls into three classes — earned, passive, error. Skip this and the analysis becomes meaningless.

Step six: publish uncertainty intervals. Below 30 matches, label a figure as a signal, not a trend.

Step seven: publish the rejected explanations too. If you do not list the causes you tested and dismissed, readers will assume you never tested them.

I first used this template in Chattogram for a football match, where a 4-2 win was really a 1.7 to 2.3 deficit. In cricket the numbers differ; the method does not. Because if the method is sound, the ledger works even when the sport changes.