The Death-Overs Ledger: The Workload Log Behind Bangladesh's Bowling Collapse That Nobody Reads

In the last three matches, Bangladesh's death-over economy has climbed from 9.4 to 12.1. Same bowlers, nearly the same line and length, different results. I have marked that 18th-over spell in red ink in my notebook, because the problem is not the ball — it is the schedule.

From the stands I watched the same yorker specialist deliver at two different heights in back-to-back overs: first at the toes, then at waist height. Commentary calls it "execution." I call execution an outcome; before an outcome there is a cause. And the cause is usually written in the log, not in the highlight reel.

Context: What I measure, and why

When I say death overs, I mean overs 16 to 20. That definition is the inheritance of the model I built for the BPL in 2026, when I hand-coded 1,240 shot events across 72 matches, cross-referenced with distance and PPDA data. The habit persists: before any claim, I write down the sample, the provenance and the coding rules. A metric without a baseline is just a rumor with decimals.

The sample here: 38 Bangladesh T20 matches over the past 14 months, with every death-over spell logged separately. For each spell I record four things — which over, which spell in that match, days of rest since the previous match, and travel days. Ball speed, yorker-success rate and dot-ball ratio get their own columns. Without those four columns I do not reach a conclusion.

Let me say it plainly: almost all talk about Bangladesh's death-bowling problem is really the last two overs of a highlight. The match story, though, is written by the 16th over. Across the 38-match log I found a pattern — in spells where a bowler had fewer than four days of rest, the economy averaged 11.8; with six days or more, it was 8.3. That is a gap of roughly three and a half runs per over. In a T20, that is the difference between winning and losing.

Core analysis: Three numbers that must be read separately

Column one, spell number. I record which spell of the match a bowler is on, because returning at the death means not just bowling — it means carrying the fatigue of the earlier spell. In the 38 matches, death spells that were a bowler's "third or later spell" had an economy of 12.4; first or second spells ran at 9.1. The practical meaning: bringing back a bowler who already bowled two overs in the powerplay costs about 3.3 runs in economy. That cost never shows on the scoreboard; it shows in the log.

Column two, travel. The gap in death economy between home and away series is 1.9 runs per over in my log. The reason is not mysterious — away means flights, time zones, broken sleep, and lost training sessions. One match hides this; a block of six matches turns it into a clear slope. This is where "home advantage," for me, stops being only crowd noise and becomes rest arithmetic.

Column three, the powerplay-death trade-off. A bowler who has carried a heavier powerplay load concedes about 1.4 more per over at the death. If a side burns its experienced pacer to put out the powerplay fire, expecting a cool head from the same man at the death is not realistic. That single line sits at the centre of Bangladesh's setup collapse, in my reading.

Now the working question: what happens when all three columns are stacked together? My log holds 11 spells in which all three conditions occurred at once — under four days of rest, a third spell, and a heavy powerplay load in the previous match. Those 11 spells combined for an economy of 13.7. By contrast, where none of the three was present, the figure was 7.9. This is not a moral story; it is a classification. If a side selects bowlers without reading workload, the selection itself is a bet against the data.

I think back to 2026, when I flagged the rise in Germany's PPDA before their pressing broke down in the group stage and circulated a note. The 2026 group stage taught me that chaos has a schedule. In cricket that lesson still holds: a collapse feels sudden, but it has a timetable. Bangladesh's death-over collapse now has a clear timetable in my log.

Contrarian angle: Workload is not the only cause

Here I have to guard against my own trap. Correlation is not causation, and I deliberately step slowly. Rising death runs have causes beyond workload — dew, the age of the ball, match-ups, field settings. In a T20, once dew settles, the ball cannot be gripped; the economy rises from a sweaty hand, not a tired one. I have started logging a dew factor separately, because without removing it I would be inflating my own thesis.

The second objection is sharper: perhaps the problem is not bowling workload but batting setup. If a side loses wickets through the middle overs, its bowlers get a defensive field at the death and concede more — which is only natural. To rule this out I added a match-state filter: I looked only at matches where the two sides' situations at the 16th over were comparable. Even after that filter, the rest effect survives at 2.1 runs per over — smaller, but not zero.

Third, I admit it: 38 matches is a small sample, and league-tournament workload cannot be compared directly with international congestion. When the stadiums went empty, I recalibrated what home meant — and in the same way, I say that when the sample changes, the threshold must change too. All I claim is this much: to explain the rise in death economy, look at the rest calendar before the quality of the ball.

Takeaway: What to watch in the next series

In the next series I will watch three things. One, the 16th-over economy — if it crosses 10.5, the setup crisis is confirmed. Two, the number of sub-six-day-rest spells per bowler. Three, the balance of over allocation between powerplay load and death load. I do not chase upsets. I chart the conditions that invite them. And I built the baseline before I trusted the outlier. If the numbers hold, changing the batsmen will not change the result — changing the calendar will.

The market moves fast; the baseline moves first.