The 66-Match Spreadsheet: Bangladesh's T20 Result vs Process Ledger, and the Signals for 2026

Nassau County Stadium, June 10, 2026. South Africa 113/6, Bangladesh 109/7. A four-run defeat. Wherever the broadcast cameras turned, the story was identical — a close fight, a late heartbreak, "one shot short." I spent that night in the corner of the press box filling a different column on my sheet: powerplay runs, powerplay boundaries, and the deviation in per-ball strike rate. In those six overs Bangladesh had made 34 without a single boundary; on the same pitch South Africa's powerplay was 43. At least seven runs of that four-run defeat were hiding in the one phase where Bangladesh was weakest across the entire tournament.

The scoreboard writes a sentence. A dataset writes a paragraph. After returning from the 2026 T20 World Cup I sat down to update my old 66-match spreadsheet — every Bangladesh T20I from the 2026 World Cup to the 2026 World Cup, each innings cut into three phases: powerplay (overs 1–6), middle (7–15), death (16–20). What the paper told me was no longer "one shot short." The spreadsheet didn't lie — it was the scoreboard that had been vague all along.

Context first, or the numbers hang in the air. Bangladesh's T20 identity over the past decade has stood on three pillars: spin-friendly slow pitches, a bowling-led middle phase, and a risk-averse powerplay. In the 2026 group stage that formula partly worked: a two-wicket win over Sri Lanka, a 25-run win over the Netherlands, a 21-run win over Nepal on DLS. The only group defeat was by four runs to South Africa.

Then came the Super Eight. A 28-run loss to Australia, a 50-run loss to India, an eight-run loss to Afghanistan on DLS. Three defeats on three different pitches, against three different opponents, but from the same structural gap. This article is an attempt to measure that gap — not in the language of emotion, but in the language of columns.

One thing needs clearing up. This is not a match report, and it is not a ledger of individual failure. The question is system-level: is Bangladesh's T20 model a winning model, or a surviving model? A surviving model gets you to knockouts; it does not bring trophies. 2026 to 2026 — three World Cup cycles, the same result pattern.

Method: how 66 matches became a number

Publishing numbers without method is incomplete work. In 2026, on a Dhaka digital desk paying BDT 18,000 a month, I hand-charted all 66 matches of the Bangladesh Premier League — shot location, body part, defensive pressure, goalkeeper position. In week six I rebuilt the sheet in Python. That table showed Abahani Limited Dhaka outperforming their expected goals by 11.4; the league table showed them as champions. Nobody in Bangladeshi football had printed those two numbers side by side.

That habit now travels with me into cricket. In T20 I use two equivalent metrics. The first is an expected-runs value for every ball, weighted by pitch condition, bowler type, match-up and match situation. The second is a pressure index that reads dot balls and the number of balls between boundaries together. Combining the two, I assign every innings a par score, then set it against the actual result.

June 27, 2026, Kazan. Germany 0–2 South Korea. I logged 2.31 xG for Germany against 0.78 for Korea and posted a thread before the final whistle arguing the champions had lost a match they controlled on every metric except the scoreboard. Here I need to be explicit: football's xG and cricket's par score are not the same thing. I use this only as a heuristic — in both, the central question is identical: how far apart can result and process be? The mapping is stated, the claim is restrained.

Finding one: the powerplay is the room Bangladesh rents every day

Across the 66-match dataset, Bangladesh's powerplay run rate was 7.1 per over. Over the same period the top ten T20 sides averaged 8.2. Bangladesh trailed by roughly 6.6 runs per powerplay — a deficit close to one wicket, every match. In the 2026 World Cup it widened to about 1.1 runs per over.

This number is harder than it looks. A powerplay run rate is not purely a function of batting skill; it is a function of intent. Showing respect to two seamers and making 34 in six overs is a decision, not an accident. My ball-tracking shows Bangladesh batters were finding the field, not breaking the line. Conserving wickets and conserving runs at the same time — in T20, the sum of those two is almost always negative.

Finding two: spin in the middle is defence, not attack

Overs 7 to 15. This is Bangladesh's traditional strength. In my sheet their bowling economy in this phase was 6.9 — among the best three in the tournament. The problem is on the batting side: in the same overs Bangladesh scored at 7.0 while the leading sides averaged 8.6. In the middle phase Bangladesh gained roughly nothing with the ball and lost about 1.6 runs per over with the bat.

This is where my first counter-intuitive suspicion was born. The "spin-friendly pitch" we take pride in — whose advantage is it really? The pitch the spinners bowl on is the pitch the batters bat on. A slow surface suppresses everyone's powerplay, but it suppresses relatively more the side that is already behind on intent. Bangladesh is using its strongest weapon to enlarge its own weakest one.

Finding three: the death-overs number is bleak, and it is the most reliable signal

Overs 16 to 20. No mystical explanation here. Across 66 matches Bangladesh's death-overs boundary rate was the lowest among the top ten sides — 1.4 boundaries per over against 2.9 for the top four. At the 2026 World Cup their strike rate in the last five overs was 119, with a wicket falling every 4.7 balls.

This death-phase deficit is statistically the firmest, because pitch luck matters least here. Yorkers at 120–130 kph and hard-length boundary hitting are environment-neutral skills. A side that cannot clear the rope in the last five overs cannot reach 175. And without 175, surviving the Super Eight on India-Sri Lanka's flat decks in 2026 means relying on the bowling every single match.

Finding four: a winning bowling attack hidden inside a losing side

Now the number that flips the story. At the 2026 World Cup Bangladesh's bowling economy was fourth in the tournament, 7.1 per over. In the powerplay they were among the most frugal sides, under 7.0 in the first six. In the death overs their economy was 9.2, top five.

In other words, Bangladesh is a side whose bowling has been final-grade for a decade and whose batting is qualifier-grade. Across the 66-match window they won barely more than a quarter of their matches, while my par-score model says they should have won close to 40 percent. That gap is what I call a winning bowling attack hidden inside a losing side. The lesson from Kazan: when result and process diverge, look at the process first, then admit the process's limits.

Why nobody wants to see this gap

In South Asian cricket there is an institutional explanation for the gap. Selection architecture still leans toward all-rounder balance, because a specialist batter's failure invites questions while an all-rounder's failure is filed under "role balance." So the squad fills with players who bat adequately at number four but are not explosive at number eight. T20 mathematics does not pay the same currency for those two roles.

Second, pitch preparation is a decision, not natural fate. Preparing a slow home pitch is a short-term winning formula, because opponents skid in the first two overs. But the same formula habituates your own batting intent downward, and on neutral World Cup grounds that habit gets exposed. Every selection is a ledger, and behind every argument there is a decimal point — it is just that nobody wants to print the decimal.

Response, not proof: a holdout test

I tested my own thesis against myself, because the first rule of a data method is to register the hypothesis in advance. I split the 66-match dataset in two: the powerplay-deficit pattern I saw in the first 40 matches (2026–2026) was checked against the remaining 26 (the 2026 cycle). The deficit held, but shrank: from 1.1 runs per over to 0.8. The problem is persistent, not static — there is a trace of improvement, which matches the improved middle-overs batting of 2026.

I concede the limits. My pressure index is not match-up specific, so the friction of a left-hander against a leg-spinner does not register. My pitch-condition data is my own log, not third-party verified. In small samples the death-overs rate swings on a single ball, so a decision needs at least three tournament windows. Ignoring those conditions and telling stories with data would stop me being a data journalist and turn me into a data guesser.

The counter-intuitive angle: the problem is not a shortage of power hitters, it is how power hitters are used

The easy decision is: "bring in more explosive batters." I tested that fairly first. Across the 66 matches, when Bangladesh's powerplay run rate was above 8, the win rate was 63 percent; when it was below 7, 21 percent. The correlation is clear. But correlation is not causation.

Look closer: in the same 66 matches, in nearly every game where Bangladesh batted first and posted more than 175, the powerplay run rate was above 8. The two are the same event under two names. Saying a side that posts 175 also has a good powerplay tells us nothing new. The real question is why a side does not attack. The answer is in selection, in intent, and in the arithmetic of fear.

Here I should state my own bias plainly. I do not believe Bangladesh lacks power hitters. I believe Bangladesh has built a structure in which a batter finds it more profitable to be a safe batter than a power hitter. The franchise market is now so large that a responsible 45 off 35 balls is a safe personal asset, while 40 off 20 carries more risk. The personal market and the team's mathematics are not looking in the same direction — that is the central conflict of Bangladesh's T20 cricket.

The same scepticism applies to the story of Bangladesh's top-ranked powerplay bowling. My data suggests a large part of that low economy is the slow pitch — the surface deserves more credit than the bowlers. On neutral, flat grounds that number rises by about one run per over. In the 2026 venue schedule, slow pitches will be hard to find.

Signals: three data points I am watching for 2026

First, the powerplay intent index. I will watch whether the boundary rate in overs 1–6 rises, and whether it rises on neutral grounds. If it only rises at home, it is not an index at all.

Second, the death-overs boundary rate. This is my most reliable predictor, because it is the least contaminated by pitch luck. If that rate does not move from 1.4 toward 2.0, final-four probability stays near zero regardless of squad composition.

Third, the age distribution of selection. I will log each series to see whether the quota arithmetic among Bangladesh's core 30-plus batters changes in the 2026 cycle. Bangladesh's strike rate in the last five overs of the 2026 Super Eight was 119 — a number that is not comfortable on any pitch, in any weather, with any luck.

I have kept my own scoring pipeline, published the dataset with the code attached, and attached a reproducibility link to every claim. Because what I learned in the press box is this: if you do not build your own numbers, someone else's numbers will sell you a story — and it will not always be your story.

Since that night at Nassau County I have built a habit. When a match ends I no longer write a sentence about winning or losing. I write three numbers: powerplay run rate, middle-overs strike rate, death-overs boundary rate. Read together, they tell you more about Bangladesh's T20 side than the scorecard ever will. Only one question remains: when the ball rolls on India-Sri Lanka's flat decks in February 2026, will Bangladesh's players be playing for the scoreboard, or for the numbers?