The Invisible Ledger of Dot Balls: The Runs Nobody Counts in an Asian T20 Season

Hook: The Numbers the Scorecard Hides

Last season, in an Asian franchise league match, the first innings produced 211. The scorecard lit up the stadium and the commentary box called it a batting carnival. That night I opened my ball-by-ball ledger and found a different number: the winning side had bowled 43 dot balls in 120 deliveries; the losing side had bowled 31. In a 211-run game the gap was not built by sixes. It was built in the zeroes — balls that never touched the bat, a keeper shifting two feet to his right, a dive at long-on that kept two runs down to one. None of those moments ever reach a thumbnail.

I opened the spreadsheet and let the World Cup confess its exaggerations. In 2026 England's Under-17 side scored 28 goals at the FIFA World Cup in India, but their xG was 22.4 — an overperformance of +5.6. I told clients the scoring rate was not sustainable. A year later in Russia, Spain carried 1,029 passes, 74 percent possession and 2.4 xG against Russia's 0.6 xG and 31.2 PPDA. I recommended under 2.5 goals and Russia +1.5. It finished 1-1, Russia winning 3-4 on penalties. I write the same way about cricket now: regress the loud number first, then open my mouth.

This piece comes out of that habit. My private log holds ball-by-ball records from 74 T20 matches in one Asian franchise season. One thing stands out. In a batting era, run-prevention labour is worth more than ever, and the scorecard is precisely where that labour is least visible. The timeline was loud, so I regressed it until the noise fell away.

Context: The Ledger, the Sample and the Definitions

My method is simple. Fix the sample first, pre-register the threshold second, look at the data third. This season I keep three separate columns.

The first is dot balls. Not all dot balls are equal. Four dots in the first two overs of a powerplay and four dots in the 16th over are the same number with completely different meaning. The first is a new-ball swing hunt; the second is a death-overs shackle. So I assign a weight: 1.0 in the powerplay, 0.8 in the middle, 1.4 at the death. Those weights are derived from my own log, not from any betting market correlation.

The second is keeper interventions. Dives, stumpings, leg-side takes, chases toward fine leg — the orphans of the stats sheet. We remember Rishabh Pant or Mushfiqur Rahim for a smart catch, but when a spinner bowls four dot balls because the keeper stood tight to the stumps and denied the batter room to stretch, the spinner gets the credit in the column.

The third is bowling workload. Overs, spell length, rest days between matches, travel distance, flight time, death-over burden — separate cells. Injuries are rarely sudden events; the accounting is usually already written in the ledger. My rule: no verdict on a bowler without a minimum nine-match slice.

An example where all three columns must be read together. In the tenth match of the season a left-arm quick took two wickets for 24 in four overs. The highlight reel shows the two wickets. My ledger shows 13 dots in 24 balls, but also that in his third spell — his sixth match in ten days, after a Dhaka-Colombo-Mumbai flight the previous night — his pace fell from 137 to 130 and his slower-ball spin rate dropped eight degrees. The runs were cheap; the process was cracking. He conceded 28 and 34 in his next two matches. The injury news arrived four weeks later. The spreadsheet had already spoken. Nobody was listening.

Core Analysis: The Invisible Ledger of Dot Balls

Phase-based value: a powerplay dot is not a death-overs dot

Across 74 matches I placed phase run rates beside dot percentages. Powerplay (overs 1-6): run rate 8.2, dot percentage 41. Middle (7-15): run rate 7.4, dot percentage 32. Death (16-20): run rate 10.6, dot percentage 24.

The plain reading: death-overs dots are rare, so each one is worth more. In my weighted model, a side that took 41 percent of powerplay dots added six percentage points to its win rate. A side that took more than 28 percent of death-overs dots added nineteen percentage points. Recovering a dot at the death buys three of the game's most expensive commodities at once — delivery speed, decision pressure on the batter, and the fielding unit's belief.

A warning, though. That nineteen-point figure is correlation, not causation. A side with a good bowling unit both bowls more death dots and wins more matches, because both flow from the same quality. Treating it as causation is how teams collapse, chasing yorkers and serving half-volleys instead.

The keeper column: where runs are never born

Across 74 matches I counted 2,314 keeper interventions — roughly 31 a match. Only 21 percent of them appear on the stats sheet as a catch or stumping. The other 79 percent are invisible.

Three pathways show how they suppress runs. One: standing tight to the stumps. On slow pitches, a batter who fears the keeper standing up is hesitant to go back. Across nine spinner spells I measured the sweep rate against leg-spin falling from 23 percent to 14 percent when the keeper stood up.

Two: leg-side work. A leg-bye off a left-hander's pad is not four, but a keeper's dive stops it at two. I logged 89 such leg-side dives, an average of 1.2 per match. The season's race for the playoffs was decided by two points between third and fourth. That is 17 runs across 14 matches — enough to flip two results.

Three: the chase toward fine leg. Wides and byes still cost runs, but the keeper's pursuit breaks the batter's rhythm. That effect lives nowhere in the numbers.

Fielding: three columns the stats sheet orphans

Run-outs, boundary saves and dropped catches are the most neglected columns in cricket data. Run-outs are counted but rarely valued. I logged 117 run-outs in 74 matches (1.58 per match). Forty-one came after the 15th over, and 34 of those were direct hits. A direct-hit run-out is worth a wicket — because that batter would not have been out ten balls later, he would have become a century.

Boundary saves: 412 in the season, each saving about 1.7 runs, roughly 700 runs saved. The tenth-placed side averaged 164 per match; the fourth-placed side averaged 173. A nine-run gap. Spread 700 runs over 14 matches and you get 50 a match — the gap between the top three on the fielding leaderboard was built here, not on the batting leaderboard.

The workload ledger: count rest, not overs

This is where the concern sits. In my season log, 11 fast bowlers crossed 48 overs across 14 matches. Six of them lost more than five kph of average pace in the final four matches.

Counting overs alone misleads. I build a Spell Load Index (SLI) for each bowler: overs per spell, the gap inside a spell, rest days between matches, and travel time.

Four spells of six overs in nine days with two rest days apart means a higher SLI than 16 overs across nine days, because muscle recovery happens in short gaps, not big ones.

Two groups emerged. Group A: average rest under three days, 17 bowlers. Group B: more than four days, 22 bowlers. In their final five matches Group A's economy was 8.9, Group B's 7.4. At the death the split widened to 10.9 against 9.1.

Yet the season-long economy gap between the two groups was only 0.3. Everyone looks fine in the first eight matches. The wear accumulates and breaks in the last five. Select a side by looking at the back end and you will be misled by the season average. There is an extra layer that all-format accounting misses: a bowler who returns from a Test for his country and plays a franchise match four days later carries a different format's load into his shoulder. For a bowler like Shakib Al Hasan, who crosses formats constantly, every spell is more than the sum of its overs — it is a different length, a different field, a different bounce, and the cost of adapting is also muscular.

Venue, pitch and the weight of the boundary: mapping home advantage

Home teams won 53 percent of matches this season. In the previous three seasons the figure was 58, 54 and 56. In a fast-moving T20 game the edge is not shrinking; it is moving.

It is also uneven. At three venues home teams won over 65 percent; at three others they won under 45 percent. Altitude, wind direction, pitch type, even floodlight angle all enter the account. At 5,000 feet the ball curves more, swings less, and a quick outfield reduces the chance of a boundary save.

I split home advantage into two parts. Physical: familiar pitch, familiar grass, familiar air. Cultural: the crowd, the microphone, the shouts behind the bowler's arm. The second can outweigh the first, in either direction. At international level, a major side travelling to a smaller side's ground sometimes gets decisions it does not get at home. Crowd noise, headlines and boardroom weight combine into an invisible penalty. I do not call this a conspiracy; I call it a slope — a small effect, one or two deliveries a match, five or six points a season.

When stadiums were empty during the pandemic season, I watched home win rates fall from 53 to 49 percent. When the stands empty, nobody can pocket the advantage, which means the difference was never in the ground. It was in the people.

Spin versus pace: the dot-ball split

Spinners averaged 7.1 an over, seamers 8.4. By dot percentage, spinners took 34 percent, seamers 30. The middle overs are where spin is most valuable, and the middle overs are where a match's tempo is set. If a side takes 45 runs for two wickets across seven middle overs, 185 becomes hard to chase. Without middle-overs control, teams concede 55-60 in the last four. That difference exists nowhere else but in presentation: more dots, more delay.

One odd pattern: spinners bowling the 16th over took dots at 29 percent, seamers only 22. The reason is simple. Late in an innings the batter wants only length; a spinner who errs is hit for six, a spinner who nails it gets a dot. Higher risk, higher reward.

Contrarian Angle: Correlation Is Not Causation, and the Sample Boundary Is Not Permanent

Every number here has a limit, and not stating it would break my own rule.

The first limit. In a 74-match sample each venue holds roughly nine matches. Saying "death-overs dots add nineteen percentage points" is easy on nine matches, but the effect size belongs to the pooled 74, not to any single venue. On venue slices I either publish nothing or move to a 27-match, three-season sample.

The second. My weights (1.0, 0.8, 1.4) come from my log. Different pitches, different balls, different boundaries will move them. Anyone borrowing the number without testing it borrows my error.

The third, and the most important. Camera angles shape which keeper interventions and boundary saves enter my log. A save outside the frame is missing from my ledger too. My ledger is not complete; it is a photograph, not the negative.

The fourth. Pressure in franchise cricket is organisational, not personal. A side whose owners are patient does not change personnel when it loses in June. A side that drops a bowler after five wicketless matches starts over every season and suffers the same decay in the last four.

The fifth. Not every football-derived idea fits cricket. Dropping a PPDA-style metric directly into T20 causes trouble.

Takeaway: What I Will Watch Next Round

Three things. First, middle-overs dot percentage (overs 9-15) against the opposition. If a side holds above 35 percent for three straight matches, that number carries more weight than its death-overs economy.

Second, rest gaps. Where Group A bowlers sit, the rate of change in economy over the next three matches will be sharpest. That is my pre-registered signal.

Third, third-man and boundary saves. A side heading toward 400 saves will sit near the top of the run-prevention table even if it is not near the top of the points table.

A closing note: the ledger does not care about the thumbnail. Sixty-six years taught me patience; the data taught me why it pays.