The Ball-by-Ball Ledger: Not 68 Death-Over Runs, but Middle-Over Dot Balls

I opened the match log before I trusted the memory. The crowd is still humming about a T20 match from the round just finished; the scorecard says 68 runs in the last five overs. The stands remember two sixes, one long-on drive, and the finisher raising his bat. But inside the ball-by-ball log, 31 of those 68 runs came from just eight deliveries; the other 22 balls produced 37 runs—including nine consecutive dot balls.

A cricket log is really an immutable ledger: one delivery, one block. Once a ball is added, the delivery can no longer be edited. People remember the six; they forget why the batter at the fifth ball of the over pushed at one and edged it. The log does not forget. So I freeze the raw numbers before the narrative can harden.

This piece is not built on a single match. I logged 1,240 legal deliveries by hand across the first twelve T20 matches of the current regular season—each ball's outcome, line-and-length zone, bowler type and batter's hand in separate columns. The habit that formed while working on the 2026 behind-closed-doors study still holds here: write the sample size next to every claim, and write "suggests," never "proves."

The method has three layers. It starts with phase splits—powerplay (1–6), middle (7–15) and death (16–20). Then dot-ball clusters, meaning how many consecutive deliveries produced no run. Finally the wicket window—how much a wicket in a given over breaks the innings' tempo. Twelve matches is a small sample, and I concede that up front; what emerges is a tendency, not a verdict.

The numbers say this: across these twelve matches, the average death-over run rate was 9.4, against 7.1 in the powerplay. Anyone who reads only the death-over run rate and calls a side a "strong finisher" has walked into an average trap. The standard deviation in the same death phase is roughly 4.8—match-to-match variation so wide that the mean is close to meaningless.

The ball-by-ball log shows the stable thing sits elsewhere: the dot-ball rate in the middle overs. Teams that kept dot balls below 40 percent between overs 7 and 15 reached the last five overs at an average of 114 with three wickets in hand. Teams above 50 percent in that phase entered the death at 89 with five wickets gone.

The real pattern is not in the death overs; it is in the eight overs before them. Without strike rotation, however many sixes land in the last five overs, that is not structure—it is excitement.

Bowling changes tell the same story. When spin was introduced for the first time between overs 7 and 15, the return was 6.2 runs and 0.41 wickets per over. Bringing pace back in the same window gave 8.1 runs and 0.29 wickets. Small sample, yet the signal is clear: turning the ball in the middle phase is the lock on an innings, and death-over heroics are often just the picture after that lock is broken.

Another layer of the log is the dot-ball cluster. Across these twelve matches, whenever a side played six or more consecutive dot balls, at least one wicket fell within the next two overs in 73 percent of cases. It could be coincidence; but the same picture across nine separate innings is not something to wave away.

One sourced fact, kept on record: browsing international T20 records in the ESPNcricinfo Statsguru database, the average run rate in overs 16–20 is broadly about two runs higher than in the powerplay. The closing overs are always loud—but loudness and structure are not the same thing.

The obvious explanation is that "the side scoring more in the death overs wins." Across twelve matches there is a relationship between death-over run rate and victory, but it is weak. Of the sides that scored more than 55 in the last five overs, three still lost—because they had fallen so far behind in the middle phase that the late storm only trimmed the margin. The pattern appeared only after I stopped asking who won.

Correlation and causation are different animals. A death-over six looks like the cause of a win to the eye; in the ledger it is often the consequence. A finisher's strike rate is a dependent variable—what he does is decided by whether the two batters before him filled the innings with dot balls.

The second uncomfortable angle is the bowling change. Captains save their best seamer for the final over and strip spin out of the middle phase; the run rate jumps, yet the log marks it as an "attacking field." The actual error there is tactical, not numerical.

In the next round I will not watch the scorecard; I will watch one column of the log—the dot-ball percentage between overs 7 and 15. If a side can push that below 45 percent, then whether they score 50 or 80 in the death overs is not the yardstick. Memory will speak of the finisher; the ledger will speak of the nine dot balls before him.