The Middle-Over Tax: Where Bangladesh's T20 Runs Go Missing in the Ledger

Over the past fourteen months I have logged 23 T20 matches ball-by-ball at the Sher-e-Bangla Stadium in Mirpur. Every delivery gets its own line: over number, bowler type, batter's hand, field setting, runs, and whether it was a dot. One pattern refuses to leave this ledger. Bangladesh's strike rate in the powerplay (overs 1–6) is 128.4; between overs 7 and 15 it falls to 104.1; then it leaps back to 148.6 in overs 16–20. A team surrenders roughly 24 points of run rate between the first six overs and the middle nine. I call that gap the middle-over tax.

What makes it interesting is how quiet the tax is. There is no single dropped catch or run-out to serve as a highlight moment. There are just dot balls, one after another, singles that go nowhere, and a crowd doing arithmetic in the last five overs. The right question is not "Bangladesh bats slowly in the middle". It is: on which pitch, against which bowling matchup, and under which conditions is that slowness most expensive?

Let me be explicit about method, because this is where most T20 talk collapses. I never use a single innings as evidence. My ledger has three tiers: exploratory (5–11 match samples), gated (12–24), and audited (25+, cross-checked by two independent scorers). The central claims here are gated; the Mirpur window of 23 matches, combined with Sylhet and Chattogram, gives 41 matches that have not yet reached the audited tier. Where the sample is thin, I gate the claim and tell the reader so.

A second definition matters: phase-adjusted strike rate. Raw strike rate swings wildly by over-block, pitch, and match state. So I compute a run-expectancy (RE) value for every ball—same over, same wickets lost, same bowling type—and then ask how far a batter or team is ahead of or behind that expectation. In T20 the middle-over RE curve is rarely the steepest, because the fielding side pushes two fielders deep and chokes off dots and singles. In Bangladesh's case, actual scoring runs 11–14 per cent below our RE. That shortfall is the size of the tax.

The origin story of the ledger is relevant. I opened the xG ledger in 2026; the 2026 World Cup wrote its own audit. The habit from those 64 football matches—xG, PPDA, sample size on every page—later taught me, while reading Italy's pressing code, that a metric only means something when a legible trigger chain sits behind it. After logging 92 Bundesliga matches behind closed doors in 2026, I learned that empty seats do not merely change the noise; they rewrite the home-advantage coefficient. I am now applying the same discipline to cricket, though not the same units—here my core units are run expectancy, dot-ball rate, and phase strike rate.

Working on Bangladesh's home conditions carries an extra burden that copying a foreign model cannot solve. The Mirpur pitch changes as a tournament progresses—more turn early, less later; evening dew makes second-innings batting easier; and crowd composition matters more than seat count. So I co-design my coefficients with local coaches, scorers and fans rather than importing them.

Now the evidence chain. Of the 23 Mirpur matches, 14 saw the side batting second play at least 22 dot balls in the middle overs (7–15); the first-innings average was 17. Filter out dew-affected games and the gap widens. Dot-ball density, not missed sixes, is the real signal.

Matchup dependence is a major driver. When Bangladesh's middle order is right-handed, opponents almost routinely pair a left-arm spinner with an off-spinner. Against that pairing, right-handers in my ledger post a phase-adjusted strike rate of 101.7, against 123.9 versus right-arm pace. The cause is not bowling quality but the angle of turn: when a left-arm spinner attacks the stumps, a right-hander's strong side closes and a dilemma opens between the sweep and the late cut. That dilemma produces dot balls.

Field setting is the next driver. If the opposition parks fielders at long-on and deep midwicket in the middle overs, a batter must hit square to rotate strike. Mirpur's outfield is slow, so square-hit balls die between two fielders. My log shows run conversion—touching the ball versus actually running—at 68 per cent in the middle overs against 79 per cent in the powerplay. The team is making contact but not making runs. That is the tax in essence.

Pitch age matters too. Within the same venue, the difference in spin bounce between the tournament's first week and its last is obvious in my log. As the pitch wears, the ball arrives slower, and the only route to middle-over runs becomes the boundary—the hardest route on this surface. The same batter who struck at 130 against the same bowler a week earlier drops to 108. That is not a form story; it is a venue-age story.

Scoreboard pressure is the final driver. A wicket in the middle overs usually pushes the next batter into "get set first" mode, and that setting-in period is where dot balls are born. In my RE model, expected run rate over the two overs after a wicket is about 7.2; in reality teams produce 5.6. That 1.6-run gap often becomes 12–16 runs by the end—enough to change a match.

The bowling side matters equally. When Bangladesh's best spin pair bowls the middle overs, opponents' phase strike rate sticks at 112; when our seamers bowl overs 7–15, opponents' run rate climbs past 140. The middle overs are simultaneously our batting wound and our bowling opportunity. The question is whether we use the opportunity on time.

A caution is essential here. Taskin Ahmed's new-ball spell and Rishad Hossain's middle overs are different jobs. In my ledger Taskin's first two overs carry an economy of 6.1, but his sample in overs 7–15 is so small—six matches—that I cannot call it audited; it stays exploratory. Deciding "Taskin for the middle overs" off two successful spells is exactly the error a ledger exists to prevent.

By venue the picture complicates further. Sylhet's ball comes onto the bat quickly, so square hits travel for four and the tax is far lower, around nine points. Chattogram offers little turn but uneven bounce, making strike rotation hard, so its tax sits in the middle, near 16. Mirpur's 24-point tax is therefore not a national trait—it is a venue-specific levy charged only on the right pitch.

As a Transfer Market Administrator, my daily job is valuing cricketers—who sells for what, whose bowling matchup is durable. That work taught me the middle-over tax is not an individual weakness but a system weakness, and system weaknesses cost money when you build a squad. A franchise that buys a specialist middle-over spinner and a strike rotator is paying to close that 24-point gap. A franchise that buys only powerplay hitting keeps paying the tax.

The same logic bites harder for the women's team. In my small window, Bangladesh Women's middle-over phase strike rate sits roughly 15 points below the men's, even though they play fewer than a third of the matches. The sample is small, so I keep the claim exploratory, but the direction is clear: where opportunity is scarce, the cost is booked before the tax is even assessed. No camp fixes that without structural redistribution.

Here is where I slow down. Correlation is not causation, and one hazard dominates my ledger: the toss. In 15 of the 23 Mirpur matches, the toss-winning side chose to field, because dew eases second-innings batting. So "slow middle overs" and "batting second" arrive almost together. Counting middle-over dots alone might lead me to blame the batting order when the real variable is the conditions.

A second trap is the half-empty stadium. Post-pandemic we saw that changing crowd composition changes home advantage; in cricket that effect is not linear. In my log, when Mirpur attendance drops below 50 per cent, the home side's strike rate falls about seven points—but pitch behaviour and toss conditions explain more of it than team pressure does. I will not say "the crowd is the cause". I will say the crowd is a context, not a single explanation.

The most important caution concerns sample size. Twenty-three matches is enough for one cycle, not for a two-season comparison. Saying "Bangladesh is structurally slow in the middle overs" on the basis of this piece would exceed my gated claim. What is safe to say: within this window the tax is consistently present, and its main drivers are identified.

In the next round I will watch two numbers. First, the dot-ball percentage in overs 7–15—if it passes 40, the side is not trying to accelerate, only to survive. Second, the phase strike rate of the side batting second after the toss—if it stays under 110 even in dew matches, the fault is not the conditions but ours. If those two numbers agree, the middle-over tax will no longer be an explanation. It will be a verdict.