Dew, Dot Balls and Death Overs: A Data Autopsy of Bangladesh's Home-Ground Record, Written from Khulna

Last season at Khulna's Sheikh Abu Naser Stadium, I counted something by hand and wrote it down. The first four balls of the 17th over and the first six of the 18th — ten straight deliveries, not a single run. The batter changed his guard twice, adjusted his stance once, and still played and missed or defended. The scoreboard finished on 231, and by the next morning the talk was about a finishing failure. 231 is not information. The information is the decision chain behind those ten dot balls, and why it began in the 15th over rather than the 17th. That single question forced me to rebuild the whole season's notebook.

The problem underneath is fundamental. Cricket has no standard metric called pressure. Football has PPDA, xG, pass networks. Cricket writes that a bowler was under pressure or a batter looked nervous — those are weather reports, not indicators. In 2026, in the thread I wrote about Germany's 0-2 loss to South Korea at the Russia World Cup, PPDA was 6.2, with 18 shots and 2.4 xG conceded against only 0.8 xG generated. It worked because pressing in football is continuous; the ball is alive almost every second. In cricket the ball dies six times an over and is reborn. Transplanting a pressing metric literally becomes data export, not analysis. That is my own admission of a mistake, and it belongs on the record.

So I stopped borrowing from outside and broke pressure down from within into three countable events. First, dot-ball clusters: three or more consecutive dot deliveries. This is the pulse of an innings, because it is here that the batter's decision count stops, not just the run rate. Second, the wicket-taking ball share: what percentage of deliveries in each over genuinely created a wicket probability, not merely a wide ball. Third, the boundary suppression rate: not conventional economy, but by what percentage the four-and-six probability was reduced. The index those three produced I called the Pressure Event Index — PEI. Before the model had a name, I counted chances by hand; a 200-match notebook, every ball in its own column, every column defined identically each time.

I deliberately kept the counting method simple so dossiers stay comparable. Five columns per match: over, ball number, delivery type, batter's shot area, outcome. Across 200 matches that is roughly 24,000 deliveries tagged by hand — six days a week, two hours a day. People ask why I count by hand when tracking data exists; the answer is that tracking data tells me where the ball went, while hand-counting tells me why the batter went there. Two different questions, two different answers. Without one, the other is incomplete.

The baseline looks like this. In my sample, the powerplay (overs 1-6) run rate is 7.4, but in the middle overs (7-15) it drops to 5.1, and the dot-ball cluster rate climbs from 23 percent to 41 percent. The damage is not caused by missing boundaries; it is caused by clusters forming — one dot ball wastes one ball, but four consecutive dot balls change the plan for the entire over. Next over the bowler gains courage, sets the field, and the batter abandons his own shot map. This is where run rate and innings tempo take separate roads.

For comparison, Pakistan's middle-over cluster rate in my sample is 37 percent, India's 34 percent, Sri Lanka's 43 percent, and Bangladesh's 41 percent. That comparison only means something when read against the opposition's average PEI — otherwise fewer clusters may just mean easier bowling, not better batting. The split clarifies further. At Dhaka's Shere Bangla National Stadium, daytime middle-over cluster rates sit near 41 percent, but at night they climb to 49 percent. Because dew falls after dark, the ball arrives wet, and the spinner loses grip. Anyone picking a night-side team from daytime numbers knows half the truth.

This is where environmental correction enters. In 2026, analysing 83 Bundesliga matches in empty stadiums, I built a coefficient: home win rate fell from 43 to 33 percent, goals per game dropped, and away teams needed 0.15 xG added to their equation. In cricket I installed the same method, but separately, because cricket does not have one environmental variable — it has four: dew, humidity, pitch age and the toss decision.

In my calculation, the night-time dew adjustment coefficient comes to 0.14 extra runs per over for the chasing side, and 0.6 runs of damage per over to spinners' economy. I publish that number in advance, not afterwards — because bookmakers adjust first and analysts notice later. Checking Khulna's notebook across 38 night matches of the 2026-25 domestic season, the chasing side won 21 of them — 55 percent. Without the coefficient I would have explained that 55 percent as the death-over skill of Bangladeshi batters. That would have been wrong, and the error would have compounded weekly.

Now to the specific place: the death overs. Between overs 16 and 20, Bangladesh's biggest problem is not wickets falling but clusters forming. On 2026 figures, 38 percent of Bangladesh innings reached the 20th over having already lost six wickets, and in those matches the dot-ball cluster rate was 44 percent. Inside that sits a fixed pattern: if the first two balls after the 14th over are dots, Bangladesh's run rate falls by an average of 1.8 across the next three overs. This is not mood-reading; it is column-by-column comparison across 210 innings. The eye test is a witness, not a judge; the model keeps the transcript.

But here I stop, and here I stand against myself. The easy conclusion is that our death overs are poor, or the reverse — that we are unbeatable at home because dew helps us. Both are lazy. Correlation is never causation. Dew sits behind Bangladesh's home success, but two larger variables sit above it: opposition quality and the resource gap. When Shakib Al Hasan, Mustafizur Rahman and Taskin Ahmed land in the same BPL squad, the bowling quality that emerges does not travel to the away leg of a series. Reading only home win rates, I would not be explaining the ball; I would be explaining the team's name.

So I added a template exception. The standard dossier now carries a new column called opposition-strength adjustment. The lesson of the 2026 empty stadiums does not translate letter for letter. In football the crowd was absent but the opponent was league-standard and comparable. In a Dhaka domestic match, an international spinner and an under-19 spinner do not bowl in the same over block; the quality gap is enormous. Environmental correction means turning the environment into a controlled variable, not an alibi.

One more line belongs on the record. I stopped reading transfer stories when I learned to read risk profiles; cricket follows the same rule. A score of 231 is not a risk profile. On 18 June 2026 at Mirpur, Mustafizur Rahman took 5/50 on his ODI debut against India and followed with 6/43 in the second match — 11 wickets across his first two ODIs, a debut record at the time. For me the number matters as repetition rather than record; his dot-ball cluster rate in those two matches exceeded 50 percent, meaning the success was not sudden magic but method. Shakib Al Hasan scored 606 runs and took 11 wickets at the 2026 ICC Men's Cricket World Cup in England, the first player to reach 600-plus runs and 10-plus wickets in a single World Cup. What connects those two facts is repetition — tournaments do not punish method, they reward it.

So what will I watch in the next series? Three signals. First, whether a dot-ball cluster forms in the first ten balls after the powerplay. Second, which spinner a side bowling first at night introduces in the 12th over — the dew coefficient bites hardest there. Third, who bowls from the 16th over: an experienced hand, or someone outside the calculation. Answer those three and the match's real owner is known before reading the scoreboard. 231 may rise again, and it should. Only this time at least we will know whether it was a finishing failure, or the arithmetic of those ten dot balls.