Overs 7 to 15: The Index Geography of Bangladesh's T20 Batting Collapse

Sitting in the Mirpur Sher-e-Bangla galleries over many seasons, one repetition keeps returning to my notebook. At the end of six overs the scoreboard offers hope — 54/1. Then the next eight overs swallow the innings whole, and by the 15th over it reads 97/5. Over the past two seasons I have tracked 23 Bangladesh T20 innings ball by ball. In the phase after the powerplay the dot-ball rate was 36.4 percent; between overs 7 and 15 it jumps to 51.7 percent. Run rate falls from 8.1 to 6.4, yet the wicket rate barely moves — one every 3.8 overs. The batters are not getting out faster; they are stopping. My notebook now has a name for it: middle-phase leakage. When I launched 'Expected Truth' from Khulna in 2026, my first condition was simple: write down the sample window before looking at the data. In football, PPDA measures pressure, but in cricket the batter has already made most of the decision before the ball is delivered, so a direct translation fails. I built the index on three pillars — dot-ball density; strike rotation, meaning the count of ones and twos per over; and boundary dependence, the share of total runs coming from fours and sixes. The sample rules were fixed in advance: full 20-over innings, unchanged top-seven batting line-up, rain-affected or follow-on matches excluded. Twenty-three innings survived; eleven were discarded. Pre-registration exists precisely so the index cannot be bent after seeing the result. Context matters. Since 2026 fielding sides have used wicket-to-wicket lines for spinners in the middle overs, and on two-paced pitches the window for strokeplay narrows as the ball ages. At Mirpur the effect is sharper because the ball grips after ten overs. Bangladesh beat Nepal by 21 runs at the 2026 T20 World Cup — check the ICC match centre scorecard and you find that in four of those five games, middle-ten-over scoring stayed under sixty. The core data sits in three layers. Broken strike rotation is the first and largest gap. In the powerplay Bangladesh batters take an average of 3.4 singles and twos per over; between overs 7 and 15 that drops to 2.1. Dots rose, shots did not — and the innings jams in the space between the two. Between the 30th and 40th ball the run rate dips hardest, because batters wait for boundaries instead of rotating strike. Boundary dependence is the second silent witness. Across these 23 innings, 64.8 percent of all runs came from boundaries; in the powerplay that share was 71 percent. In the middle overs they try to survive, but the survival mechanism is not yet built — once boundaries stop, the run wheel stops. The 14th over is the clearest evidence: Bangladesh average 6.4 in that over, while in overs 16 to 18 the rate climbs to 9.8. The repair work starts far too late. Left-right match-ups form the third layer. Opponents routinely bring a left-arm spinner on in the middle overs; Bangladesh's run rate against that skill set is 5.9, against right-arm spin it is 7.3. The issue is not spin alone — it is the absence of a ready reverse-sweep and late-cut option. Litton Das and Towhid Hridoy fare comparatively better in this phase because they can play off the crease; others stay stuck on the line. Read Taskin Ahmed's and Mehidy Hasan Miraz's numbers separately and you see them hold this phase with the ball, while the reverse happens with the bat. An index means nothing without a comparison group, so I placed India, England and Australia beside it — their dot-ball rate in the same phase sits between 41 and 44 percent, strike rotation between 3.1 and 3.6. The gap is Bangladesh-specific, and explaining it through a talent deficit leads down the wrong road. A caution is necessary here. The numbers didn't break the model; they exposed where the model was blind. A large slice of the sample was chasing innings, where the run rate naturally falls when wickets go down because the team recalculates the win. My first version had no such variable, and the index narrated a pure batting-failure story. Split the innings and batting-first leakage is 51.7 percent while chasing it is 42.1 percent. Half the fault is strategic. A second blind spot surrounds ball-change rules. Since 2026 commentators have discussed its effect, but the sample size is still too small. Reaching a verdict off one series or one spell is mere outlier chasing — and I don't chase outliers; I follow them until they confess. The third gap is dressing-room shaped: data shows who is stuck, never why. Coaching plans, batting-order experiments, a young batter's confidence — those are measured only through ground reports and interviews. For the next series I have set a checkpoint: if the dot-ball rate between overs 7 and 15 falls below 46 percent, I will treat it as model revision, not a miracle. Only when strike rotation reaches 2.8 per over will I say the leakage has closed. Expected truth is not a verdict; it is an ongoing investigation. The single remaining question — will Bangladesh's batting unit change its strokes, or merely wait for the clock to change?