47 Percent Dot Balls: Pakistan's Middle-Overs Ledger and One Misread

June 9, 2026, Nassau County International Cricket Stadium, New York. India were bowled out for 119; Pakistan needed 120. Handed a target that small in a T20 World Cup group game, Pakistan stopped at 113 with seven balls left. The scorecard says a six-run defeat; my ball-by-ball log says something else. Between overs seven and fifteen, Pakistan's dot-ball rate was 47 percent, meaning they took zero runs off roughly half the deliveries in those nine overs. Jasprit Bumrah's spell read 4 overs, 14 runs, 3 wickets. That one spell left the largest residual in the match. The middle phase, overs seven to fifteen, is the least discussed and most decisive phase of T20 cricket. The powerplay keeps fielders up, the death overs push bowlers into risk; across the middle nine the field spreads, spinners bowl, and a new batter has to rotate strike one and two at a time. On slow, low subcontinental surfaces this phase grows heavier still, because the ball arrives late and a batter building an innings simply banks dot balls. I have coded ball-by-ball since 2026. Coding 380 Belgian second-division matches taught me that a scorecard never measures pressure. For cricket I built a cricket-native proxy: the Dot-Ball Pressure Index, or DBPI. It pairs the number of dot balls per over with the strike-rotation rate, the singles taken per dot ball. Importing football's PPDA here would break the measurement; cricket has no notion of possession, only of deliveries. My three-season rolling baseline, 2026, 2026 and 2026, puts Pakistan's middle-overs strike rate at 118, 121 and 113. The dot-ball share moved from 41 percent to 43 to 47. The sample is 41 matches and 2,870 balls; my confidence level is 90 percent, not 100. Over the same window India's middle-overs strike rate is 134 with 34 percent dots. The pressure Wanindu Hasaranga and the Afghan spinners create in this phase is the regional benchmark. A regional comparison matters here, because treating one team's weakness as the championship standard produces error. Under Asia Cup conditions, Dubai's slow pitch, evening dew, Sri Lanka's spin-friendly surfaces, middle-overs strike rates drop four to five points by default. In my sample, Asia Cup matches average 124 in the middle phase against 131 in bilateral T20I series. The tournament's conditions are themselves a constraint, and the difference has to be measured inside that constraint, otherwise the comparison is simply unfair. My own ACL tore in 2026, and I rebuilt myself as a ledger of lost minutes. Cricket runs the same logic: absence is data, not narrative. Shaheen Shah Afridi's ledger therefore needs its own line. In July 2026 he injured the ligament in his right knee during the Test against Sri Lanka in Galle and missed the 2026 Asia Cup. He returned for the T20 World Cup, then hurt the same knee again on November 13, 2026, taking a catch in the final against England in Melbourne; much of 2026 passed in load management. I trust the model, then I audit it until the residuals confess. That audit shows Pakistan's powerplay wickets per over fell from 0.21 to 0.14 across those 14 months. At the innings break, DBPI whispered that Pakistan had a 62 percent chance of chasing 120. The model was wrong, and the error is instructive. The easy explanation is that Shaheen was missing, so Pakistan felt pressure in the middle. The data does not support it. Shaheen's impact sits in the powerplay and at the death; between overs seven and fifteen he rarely bowls more than two. The middle-overs dot-ball problem is not a bowling problem, it is a batting-rotation problem. Keep the match state in view: 44 needed off 30. In that equation 35 percent dot balls is tolerable, 47 percent is not. Preserving wickets was not irrational; the failure was that nobody switched on the second half of the plan, meaning nobody took the risk of two or three boundaries in a single over. There is another trap I try to avoid myself: granular overfitting. The micro-pattern that Pakistan struggle against Bumrah's slower ball only becomes a conclusion if it survives three phases and a rolling baseline. Twenty-four balls from one spell cannot explain a national batting culture. Batter-level ledgers make it sharper. The Rizwan and Babar pairing loses few wickets in the middle but rotates slowly, fewer than two singles per six balls on average. When a wicket falls, that structure breaks as a unit, and the dots pile up. The problem belongs to the rotation design, not to one man. Load cannot be skipped either. From 2026 to 2026 Shaheen Afridi played international cricket plus franchise leagues almost year-round; recovering that load after a knee injury is not simple. When tournament expansion comes up, I usually ask the question with player-load data: add fixtures and the fitness-dependent strike rotation in the middle phase breaks first, because it demands the most repetition. One thing could prove me wrong. If dot balls fall from 47 to 40 percent in the next series while the strike rate does not rise, the issue is intent, not rotation. Then my v1.0 is wrong, and I have no objection to writing that down. So what is the metric to watch next series? Pakistan's strike-rotation rate between overs 12 and 16, and the willingness to take at least one six-hitting risk per over. Cutting dot balls in the middle does not mean more aggression; it means raising the singles rate without falling into the four-and-six trap. This is v1.0; when the next series produces data I will write v1.1, and then place Shaheen's load management and the batting rotation in the same ledger. The question stays open: chasing 120, is a 47 percent dot-ball rate a failed plan, or one exception named Bumrah?