In Tournament Cricket the Real Battle Is Overs 7 to 15, Not the Powerplay

29 June 2026, Kensington Oval, Barbados. In the T20 World Cup final, South Africa needed 30 off 30 with six wickets in hand and Heinrich Klaasen on strike. I had two windows open on my laptop — a live scorecard on one side, my own phase map on the other. Klaasen was 52 off 27, striking above 190. Over the next three overs, Jasprit Bumrah, Hardik Pandya and Arshdeep Singh conceded twenty runs between them and took five wickets. India won by seven runs, 176/7 against 169/8. Open the ball-by-ball scorecard, though, and the first thing that catches your eye is not the drama of the last three overs. It is that South Africa were roughly twenty-five runs behind India across overs seven to fifteen. The hero of the final over won the match; the match itself was constructed in those eight middle overs.

I joined the sports desk at The Daily Star in Dhaka in 2026, when bowlers were judged on economy and batters on strike rate. Nineteen years later I work on ball-by-ball cricket data from London. Along the way I picked up a habit: before I tell the story of a match, I separate the phases. There is no magic in it, only a suspicion that what happens last is not always what caused it. In 2026 I wrote about expected goals in football; by 2026, watching Russia sit behind the ball against Spain, I learned that a model's estimate and a model's explanation are two different objects. Football's structure does not transfer cleanly onto cricket, because cricket runs on balls, innings and over limits. So I came back to cricket's own units: the value of a ball, the state of the game, the probability of a wicket.

The backbone of this piece is a small dataset. I split ball-by-ball data from 141 matches across three ICC men's tournaments — the 2026 and 2026 T20 World Cups and the 2026 ODI World Cup — into phases. In T20 there are three: powerplay (1-6), middle (7-15), death (16-20). In ODIs there are four: powerplay (1-10), middle-early (11-25), middle-late (26-40), death (41-50). For each phase I computed run rate, dot-ball percentage and boundaries per ball, then asked which phase's run differential aligns most closely with the final result.

The answer surprised me at first. In T20, a team's middle-phase run differential correlates more clearly with winning than its powerplay differential does. Teams leading by ten runs at the powerplay won about 61 percent of those matches. Teams leading by ten at the middle phase won more often than that, and past twenty runs the win rate climbed past 80 percent. At the death the relationship weakens again, because runs flow so fast that small gaps vanish in an over. The phase where scoring is slowest is the phase where the most durable advantage is built.

The reason is not mysterious. In the powerplay there are two new balls, two openers and fielding restrictions — low risk for the batter, high expected runs. In the death overs the bowler already knows he will leak, so the plan becomes: give runs, take wickets. The middle eight overs sit between those extremes. The ball is older, the seam has dulled, spinners have more overs to bowl, the field spreads out, and clearing the boundary needs direction as much as force. That is why the middle phase carries the highest dot-ball rate. In my sample, the middle overs produced roughly 4.1 dots per over against 3.3 in the powerplay and 2.6 at the death.

That is where a pattern emerges that I call the two-paced trap. On tournament surfaces, especially used pitches in semi-finals and finals, the new ball comes on a touch quicker and then the surface slows as the innings wears on. Short-format talent scouting, however, is effectively built for the powerplay and the death. Squads are assembled at both ends. The batter a middle phase actually needs — someone who rotates strike on a holding pitch, who has the patience for the scoop, sweep and single — is the last name on the sheet. Bangladesh's recent T20 thinking looks like a textbook case: plenty of power, thin rotation through the middle.

Spinners own this phase. Afghanistan reached their first ICC semi-final at the 2026 T20 World Cup, and the engine was the middle-overs control of Rashid Khan, Mohammad Nabi and Gulbadin Naib. Look at 22 June 2026 in St Vincent, where Afghanistan beat Australia by 21 runs. Australia's powerplay was fine. Across overs seven to fifteen, Afghan spin pulled the rope so tight that the last five overs began with a required rate above eleven. No single delivery was extraordinary. The collective plan was. Tournament T20 is not a murder; it is a squeeze. The side that wastes one fewer ball per over in the middle is the side holding two extra balls at the end.

In ODIs the arithmetic is even sharper. At the 2026 World Cup, the run differential between overs 11 and 40 tracked results far more reliably than the powerplay. On 19 November 2026 in Ahmedabad, India were bowled out for 240 and Australia made 241 for four. India's powerplay was sound, but between overs 11 and 40 they scored roughly 160 across nearly 30 overs. Australia's strategy was pure patience: accept being behind in the powerplay, grind through the vast middle at five to five-and-a-half an over, keep wickets in hand, then explode. To me this is the cleanest ODI model, because the format gives you time — the condition is that you do not waste it.

A counter-question follows, and skipping it would leave this analysis unfinished. I am claiming that teams who win the middle phase win more often, but is that cause or consequence? Arrows can be drawn both ways. A side ahead in the middle is usually already controlling the game; a side that has lost wickets is the one slowing down. Game state is a hidden variable. When I separated first innings from second, the relationship held but weakened, and once rain-affected matches were stripped out the sample shrank from 120 to 48. There is a second trap: a chasing side often slows deliberately in the middle, terrified of losing wickets. So part of the middle-phase differential is bowling skill and part is batting fear.

I therefore file my own conclusion as a probabilistic lean, not proof. Before applying it I attach conditions: the pitch and whether it has been used; the spin-over depth of both sides; and the timeline of wickets falling. When those three align, the middle-overs signal is reliable. When they do not, it is noise. In football I used to say expected goals never lies, it only waits. I am still searching for cricket's equivalent. A dot ball in cricket is not a missed shot in football. Sometimes five dots are just a quiet start; sometimes they swallow the tempo of an entire match.

One habit from that 2026 desk has survived: I write a prediction before a match and check it afterwards. In 2026 I predicted, then verified, that home win rates in empty football stadiums would fall from 43 to 33 percent. In cricket my next verification will concern the relationship between spin-over depth and squeeze capacity, priced not just in field restrictions but in the limits of the game itself. And that is where the next tournament becomes interesting. At the 2026 T20 World Cup in India and Sri Lanka I will not be tracking opener hundreds. I will track three quiet middle-phase indicators: dots per over, the ratio of singles to boundaries, and the share of deliveries spinners land on hard length. Any side fielding at least one patient accumulator and two bowlers who can control overs eleven through sixteen is a side I want to follow to the end. One question I will leave hanging, because I do not have the answer either: if the slowest-scoring phase is the most valuable one, why is our scouting language still built almost entirely from powerplay and death-over images?