The Death-Over Dot Ball: The Numbers a T20 Scorecard Never Shows

Hook

March 23, 2026, M. Chinnaswamy Stadium, Bengaluru. Bangladesh needed two runs off the last three balls. India won by one run.

What nobody wrote that night: in the 17th over of Bangladesh's innings, four of the six deliveries were dots. Four. The very next ball went for a boundary, so the over entered memory as a good over. The scorecard keeps no trace of those four dots, because zero has no weight — until you lose by one run at the end.

I opened a blank spreadsheet that night in Rangpur. I opened a blank spreadsheet and let the Bangladesh Premier League teach me — but on that first night there was nothing to learn, only empty cells. I left the column blank and went to sleep. Ten years later, I think that empty cell is what taught me to ask questions.

Context: a hand-coded model and its appendix

In 2026 I was forty. By day I audited rice-mill accounts in Rangpur; by night I hand-coded an expected-goals model. That season I coded 132 matches and 3,410 shots myself, because no public xG existed for the Bangladesh Premier League. Abahani Limited's title run showed a 9.4-goal gap between my model and their actual goals. Within a week, three betting syndicates emailed me.

The real lesson, though, was elsewhere. The xG model was crude, but the missing cells confessed more than the goals. I stopped writing match reports and started writing methodology notes: every claim carried a sample size, a weighting choice, an error margin. And I started keeping an appendix where I write down everything my model gets wrong. I trust a number only if it comes with a list of the ways it lies.

In 2026, in Russia, I ran the same method in another sport. Across 64 matches I logged PPDA and set-piece xG, and I published a pre-tournament piece arguing Germany's press had already decayed. They went out in the group stage; 40,000 people read it. But my model still ranked them third-favourite, so I hedged my own thesis in the final paragraph and lost the argument anyway. By Russia 2026, I was watching Germany twice: with eyes and with PPDA.

That two-track habit is the method behind this piece. Cricket has no xG, but dot balls, phase-by-phase run rate and death-over ball counts build the same kind of structure. The difference is only this — in football the raw material is the shot, in cricket it is the ball.

Core: the dot ball is the real currency of T20

In T20, runs are not the currency. The dot ball is — because a dot delivers no run and consumes one of only 120 balls. A six sets the height of an innings; a dot ball sets its lifespan.

Across the thousands of matches I have watched with a spreadsheet open beside me, boundary-led innings and dot-led innings look different. On March 18, 2026, at the R. Premadasa Stadium in Colombo, the Nidahas Trophy final, Bangladesh made 166. India won off the last over, Dinesh Karthik scoring 29 off eight balls. All the talk afterwards was about Bangladesh's bowling. But in the middle of India's innings there was a cluster of consecutive dots created by Bangladesh's spinners — and then the cluster broke in the last three overs. The next morning everyone wrote about India's hero; nobody wrote about that cluster.

In my hand-coded sample I separate dots by phase, because the same dot ball means three different things:

• Powerplay (overs 1-6): dots are cheap. The batter is hungry; playing out a ball or two is part of the plan.
• Middle (overs 7-15): dots are most expensive here. This is where an innings wants to breathe, where the set-up happens. 50/2 in the powerplay versus 90/3 in the 15th — the second looks ahead on paper and is structurally far behind.
• Death (overs 16-20): dots here are usually burn balls, the batter swinging through air. That is not batting failure; that is the fielding side winning a ball.

A powerplay dot, a middle-over dot and a death-over dot are never the same event, yet the scorecard renders all three identically. That is my central objection to the scorecard.

The second thing is stranger. A single dot does no real harm. In my sample, the probability of a boundary on the ball immediately after a dot is roughly 14% — the batter loses patience and hunts a shot. What damages innings is a cluster of dots, especially when it lands between the 16th and 18th overs. There, ring fielders step six yards in, the batter's options contract, and three dots in one over open the door to a hit-wicket or a run-out in the next. That is not luck. That is ball-by-ball design.

One pattern I keep finding: the team that wins a close match is often not the team with the most boundaries. It is the team with the fewest dots between overs 16 and 20. Boundaries tend to arrive in matches already won; dots arrive in the overs that decide matches.

Contrarian: the dot-ball metric is itself a disguise

This is where I stop, because the most elegant metric usually leads to the dumbest decision.

There is a relationship between a good dot-ball percentage and good batting, but not causation. Two sides can have near-identical death-bowling dot rates and still be separated by the depth of their batting order. If the specialist finisher bats at seven, the captain can afford to eat a dot at 17; if the all-rounder bats at six, one dot tilts the whole next over. I get nervous when someone says a team scored fewer dots, so they had a better plan — the causal arrow may run entirely the other way. Let me be explicit: a lower dot-ball rate is desirable, but it is not the most reliable measure of skill.

The real skill is not making the ball small. A side going at 7.2 in the middle overs with two wickets down is playing a different sport from a side going at 7.2 with six wickets down. Same number, different game.

The dot-ball narrative is a middle-class metric. It flatters teams that are already organised. It does not tell you who wins the tournament.

Takeaway

So the thing to watch is not the dot-ball percentage. Watch the dot-ball cluster length — the longest run of consecutive scoring-free balls inside overs 16-20.

• Bangladesh at home: clusters are short but frequent. That is control without kill — a structure problem, not a skill problem, and structure problems are fixed with the batting order, not with nets.
• Away, on quicker pitches: clusters lengthen and the finish fails. That is not mental fragility. That is the middle overs being eaten before the death overs arrive.

If a team wants to win the next tournament, the number to chase is not strike rate. It is this: how many consecutive balls can you survive without scoring, before the innings starts shrinking?

Watch the blank spaces on a scorecard, not its bold numbers. A side scoring at 8.4 in the middle overs with five wickets down is not ahead of a side scoring at 7.6 with two down — whatever the board says. The scoreboard is only a receipt. It tells you how much you spent. It never tells you how much you wasted.