On Saturday night at Mirpur, the last three overs added 19 runs. The batter who stood on 34 off 28 at the start of the 14th over had a control percentage of 81.3 in my log — four of every five deliveries met without an edge, a mis-hit or a beat. Yet that spell read 2-3-1-0-1-2, plus two wickets. The headline after the game was the powerplay: 58 for 1, a boundary rate of 23.5 percent, six points above the Mirpur venue baseline.

At midnight in my room in Rangpur I drew two columns in the notebook. The boundary column is pleasant to look at. The dot-ball column is tedious. The match turned in the tedious column. This piece is an argument for it.

Context: what my log holds, and why ten matches

I have written weekly data threads since 2026, and I imposed one rule on myself: no trend gets a name before ten matches. That year I wrote about a Burnley thread —

The Burnley thread looked like noise until I sorted by PPDA.

The thread sounded like noise until I sorted it by phase. In Russia in 2026 I wrote after one match —

Modric ran twelve kilometers, but the map showed where the game turned.

Modric ran twelve kilometers, but the map showed where the game actually turned. Those two football habits shaped my method, but I do not translate them literally into cricket. PPDA and distance covered become decoration when pasted onto cricket. The cricket equivalents are phase-wise dot-ball percentage, control percentage and wicket probability.

What I log:

• Dot-ball percentage (DB%) — share of deliveries in a phase producing no run.
• Control percentage (Cp%) — deliveries where the batter was not beaten, did not edge and did not mis-hit.
• Boundary percentage (BP%) — share of fours and sixes.
• Expected runs (xR) — runs generated by a shot-quality model, not boundary-dependent.
• Wicket probability (WkP%) — expected wickets per over.

Sources: my own ball-by-ball log, manually reconciled against public scorecard archives. I excluded innings before 2026, where control tagging is not equally reliable.

Data Monk / ISTJ rigor | Scenario: Methodology section in a long analysis.

I keep separate venue baselines for Mirpur, Chattogram, Dubai and Colombo, because the same team's phase behaviour flips when the venue changes. Every number here comes from innings played between 2026 and 2026, where ball-by-ball records are dependable.

Core analysis: the price of a middle-over dot

The Mirpur T20 phase baseline (68 innings):

| Phase | Overs | Runs/over | Dot% | Boundary% | Wickets/over | |---|---|---|---|---|---| | Powerplay | 1-6 | 7.42 | 42.1 | 17.8 | 0.29 | | Middle | 7-15 | 7.06 | 36.4 | 12.1 | 0.34 | | Death | 16-20 | 9.15 | 27.3 | 19.6 | 0.52 |

The easy domestic read is the powerplay. The table says otherwise. At Mirpur the powerplay carries the highest dot-ball rate (42.1 percent), yet the match is decided in the middle overs — runs per over are lowest there (7.06), boundaries are scarcest (12.1) and wicket probability is higher than in the powerplay (0.34 against 0.29). A dot ball is most expensive in the middle overs, because the batter must face the same quality of delivery the following over, and a wicket drops a new batter into turn and grip.

Powerplay dots are cheaper: the ball is new, the ring is in, and the batter knows there is time. A 42 percent dot rate in the first six is a gift of conditions rather than a bowler's achievement. A 36 percent dot rate in the middle is manufactured.

The ten-match rolling split across six teams

I compared middle-over dot-ball rates for six Asian sides — Bangladesh, India, Pakistan, Sri Lanka, Afghanistan, the United Arab Emirates — across their last ten matches against their previous ten.

| Team | Previous 10 middle DB% | Last 10 middle DB% | Delta | Wins previous 10 | Wins last 10 | |---|---|---|---|---|---| | Bangladesh | 39.8 | 32.1 | -7.7 | 4 | 7 | | India | 34.5 | 31.6 | -2.9 | 7 | 8 | | Pakistan | 35.2 | 34.8 | -0.4 | 5 | 5 | | Sri Lanka | 38.1 | 33.4 | -4.7 | 4 | 6 | | Afghanistan | 37.4 | 34.9 | -2.5 | 5 | 6 | | United Arab Emirates | 41.6 | 39.2 | -2.4 | 3 | 3 |

Pakistan's column is the most instructive. The delta is only -0.4, yet their powerplay scoring rose from 8.95 to 9.21 runs per over inside the same window. The team is attacking earlier, but the middle-over absorption has not changed, so the win count held at five. More boundaries without lower wicket probability is not a sustainable trade on Asian surfaces.

Bangladesh's picture runs the other way. Powerplay scoring barely moved (7.51 to 7.48), while middle-over dot ball percentage fell 7.7 points. The rise from two wins to seven owes nothing to the powerplay; the entire change sits in overs seven to fifteen.

Stability check: four venues, four answers

Ten matches are not enough on their own. The real question is whether the signal survives a change of venue.

| Venue | Did the middle-over DB% signal hold? | Boundary% coefficient | Note | |---|---|---|---| | Mirpur | Yes (9/10) | -0.51 | Slow deck, spin turn | | Chattogram | Yes (7/10) | -0.38 | Sea breeze, turn in second spell | | Dubai | Partly (6/10) | -0.29 | Large ground, misread dots | | Colombo | No (4/10) | -0.08 | Flat deck, short boundaries |

Colombo's failure is the most valuable data point I have. There the run rate crosses 8.7 and the boundary rate crosses 20 percent. On a short boundary a single error in length becomes four, and a batter can keep the board moving without taking risk. The dot-ball signal does not work everywhere; it works where the boundary is expensive. Anyone who says 'dots win everywhere' has not read my log.

Batter-level landing

Team numbers show where a phase leaks; individual numbers show who sits inside the leak. Three batters against spin in the middle overs over the last ten matches:

| Batter | Middle SR vs spin | Middle SR vs pace | Dot% faced | Cp% | |---|---|---|---|---| | Towhid Hridoy | 117.4 | 132.8 | 30.2 | 78.9 | | Najmul Hossain Shanto | 108.6 | 121.5 | 35.1 | 73.4 | | Litton Das | 119.8 | 128.1 | 32.7 | 71.6 |

One thing stands out in Hridoy's row: against spin his control percentage is nearly five points lower than against pace, while the strike-rate gap is fifteen points. Dots above thirty explain it — he does not edge or mis-hit much, but he does not rotate either. Without rotation, the over rate stalls and wicket probability rises.

Shanto's problem is elsewhere. A dot rate of 35.1 with control at 73.4 says he is playing the ball properly but searching for it. On an Asian turner, the distance between those two positions is the distance between winning and losing.

Bowler-level counter-evidence

Unless I ask whether dots are the batter's failure or the bowler's creation, the analysis is incomplete. Two spinners in the middle overs:

| Bowler | Middle econ | Cp% | Dot% | Googly/variation share | |---|---|---|---|---| | Rishad Hossain | 6.84 | 74.2 | 41.8 | 31 | | Mehidy Hasan Miraz | 6.21 | 79.6 | 44.5 | 18 |

Miraz's control is five and a half points higher than Rishad's, and his dot rate is higher too, yet his wicket probability is lower. He manufactures dots with flight and slower balls that do not force risk; the batter survives on the pad. Rishad bowls with less control but his variation share is higher, so his dots travel with wicket probability. Both are dots, but one weighs more than the other. That distinction belongs at the centre of venue-specific bowling plans, and it is also where black-box metrics run out — the same 'dot-ball percentage' describes two different threats, and without a stability check nobody can tell them apart.

Precedent table: champions in Asia-hosted tournaments

| Year | Tournament (host region) | Champion | Middle DB% | Era-adjusted delta* | |---|---|---|---|---| | 2026 | Asia Cup T20 (Bangladesh) | Sri Lanka | 36.9 | +1.4 | | 2026 | Asia Cup T20 (Bangladesh) | India | 34.2 | +0.6 | | 2026 | Asia Cup (UAE) | Sri Lanka | 35.1 | +0.3 | | 2026 | Asia Cup (UAE) | India | 32.4 | -0.9 |

*Era-adjusted delta = gap against the tournament average of all teams' middle-over dot-ball rate.

The table shows a pattern, but it is a four-row sample and nobody should be judged by these numbers. A 2026 dot and a 2026 dot do not mean the same thing; the ball changed, the definition of T20 aggression changed, and death-over scoring rose. That is why the delta column exists — not raw numbers, but the gap against the same era's peers.

The reading is still plain: all four champions sat at a positive or marginally negative era-adjusted delta, none two digits behind. In Asian conditions the trophy goes to the side whose middle-over ball economy beats its peers, not to the biggest hitter.

Contrarian angle: correlation is not causation

Here is the part such pieces usually skip. Across 60 matches for six teams I measured two relationships: powerplay runs per over against wins, r = 0.19; middle-over dot-ball percentage against wins, r = -0.44. Powerplay scoring predicts weakly; middle-over dots predict strongly and inversely.

Even so, -0.44 is not a cause. Dot balls are a proxy — a composite imprint of wickets in hand, ball deterioration and rotation skill. A side that cuts dots while losing wickets will post a handsome strike rate and a miserable result. In one team's last ten matches, two games carried more than 150 dots each and were still won, because the match-ups fell their way. Numbers occasionally point the wrong way.

My second caution is my own threshold. Ten matches is a convenient number, not a law. I picked it because it lets opponent variety in without letting a batter's form cycle pass entirely. I write the reason alongside the number so readers know it is my decision, not a discovery.

My third caution concerns selection. A batter outside the XI who scores at a strike rate of 140 but eats dots on forty percent of balls is, in a wet-dew middle phase, a drag whatever the headline rate says. The value of a long set-up and the sum of a volatile flash are never equal.

My fourth caution concerns the rolling weekly window. If one match in a ten-match window produces only two dots across twenty overs, the average lurches. Without separating innings length, opponent bowling depth and scoreboard pressure, the number misleads. In my log every match carries innings length, opponent strength tier and score pressure beside it. When a future match drops outside that window, I open a new ten-match frame.

Takeaways: what to watch in the next round

Next round I will not watch the scorecard; I will watch phase dots from overs seven to fifteen. In Asian conditions a side keeping middle-over dots below 32 percent with wicket probability around 0.5 has a genuine title case. With one condition attached: on flat decks like Colombo the average does about half the work. So the number that will truly look predictive is not the middle-over dot itself but the conversion rate from middle-over dots to wickets — because that ratio tells you whether the pressure is real or merely arithmetic.

Sources and method note

• Venue baselines: 68 innings at Mirpur, 47 at Chattogram, 52 in Dubai, 59 in Colombo — cleaned manually from ball-by-ball scorecard archives.
• Control percentage: counted after removing edged, mis-hit and beaten deliveries, reconciled against DRS-corrected scorecards.
• xR model: built on four tiers of line, length, pace and spin variation; based on shot quality rather than boundary-dependent runs.
• Innings before 2026 excluded; variation tagging for spin bowlers of that period is incomplete.
• Precedent table is era-adjusted; raw percentages do not capture the relative demand of each era.
• Weather caveat issued before including the five flat-deck innings.
• Method and data checked against cricsultan.com data indices.

A final word on a baseline audit: the job of a phase column is not to pick a winner, but to point at a match that was won when it could have been lost. Those who want to change a system after one game will change it. I keep the next ten-match audit open.