The Six-Over Ledger: A Hand-Coded Audit of BPL Powerplays, and the Three Columns Left Blank

Hook

I opened the hand-coded ledger again, and the margins disagreed.

In a Mirpur match last season the broadcast graphic showed 58/1 at the end of the powerplay. My own coded column for the same innings was recording 47/1. An eleven-run gap. The scoreboard had not made a mistake, and the graphic did not know it had made one either; the discrepancy was hiding inside a single boundary — a six, not a four — filed in the wrong over. Anyone who has scored for a living knows these nights do not let you sleep. The reason is simple. Move one boundary to the wrong column and you have not only moved the runs, you have moved the ball count of the over. Then the strike rate moves, the economy moves, the dot-ball percentage moves, and eventually the average for the whole season shifts a few steps.

That night I sat down and reconciled the ledger. The logic was plain: if one column can sit in the wrong box, seven hundred columns can. And if seven hundred columns sit in the wrong box, then the thing I have been shouting about as a "trend" is not a trend. It is a typing error.

What follows is the result of that reconciliation. This is not a match report. It is an accounts sheet, in which three of our favourite powerplay beliefs — "intent", "the second-innings advantage", and "start well, finish well" — are weighed separately.

Context: how the ledger is written, and why the labour is necessary

My hand-coded ledger holds 462 matches from four BPL seasons. A separate notebook holds 22 Chattogram Abahani matches, 588 shots, 197 of them on target. That one is a football notebook from 2026, but the method is identical: every ball, every event, every decision in its own column, and a timestamp on every column.

In the BPL ledger I filled 41 columns per ball. Over number, ball number, batter, bowler, line, length, shot type, field location, runs, match state, wicket or no wicket, dew present or not, and the date and venue. Then a separate column: version number. When I corrected a match later, I did not delete the old figure. I wrote the new one beside it and dated it. Delete the old figure and you destroy the evidence; keep it and you can see the history of your own mistakes.

Why the labour? Because BPL scorecards are correct on paper but not identical in every database. Some matches do not record the Super Over separately. Some matches have the Duckworth-Lewis revised target folded into the original column. Some matches have abandoned innings recorded as zero runs, as though the teams were dismissed without batting. Fourteen months of silence taught me this: an empty column is not a zero. A zero means the side batted and did not score. Empty means we do not know what happened. Put the two in the same box and the average becomes a lie — and it is that lie we write next season's analysis from.

The BPL market is not small. Overseas stars arrive every season, franchises change hands, coaches change. There is no central ledger for that change. So the questions that matter most — how quickly powerplays actually go at each venue, how much the second innings gains when dew falls, what share of balls young batters actually receive in the powerplay — can only be answered by hand.

Core: six overs, four seasons, one set of numbers

First, the population. Of 462 matches in my ledger, 431 have complete ball-by-ball coding for the first six overs. The other 31 are partial — either rain, or a lost broadcast feed. I did not fold the partial columns into the averages. They sit separately, because mixing a partial sample into a full one manufactures a trend without your noticing.

Across 847 powerplay innings from those 431 matches, my ledger reports:

• Average powerplay run rate: 7.42 • Dot-ball percentage in the powerplay: 46.8 • Average wickets lost in the powerplay: 1.6 • First-innings powerplay run rate: 7.24 • Second-innings powerplay run rate: 7.61

The first thing the eye catches is the second-innings gap — 0.37 runs per over. Over six overs that is roughly two runs. Against a match total it sounds small. But two runs in a powerplay are not two runs; two runs in a powerplay are one extra boundary, and behind that boundary the field changes, the bowler changes, and the count of wickets in hand for the last five overs changes.

Split by venue the picture sharpens. Powerplay run rate at Mirpur is 7.18. At Chattogram 7.66. At Sylhet 7.91. The gap is not accidental. At Mirpur the new ball seams, and spinners get a grip in the first ten overs. At Sylhet the ball comes onto the bat and the outfield is quick, so adventurous shots are rewarded. Chattogram sits in between — wind, coastal humidity, but the ball slows once the spinners settle.

Here is where I want to push back on how we usually talk about this. We say "batting in the powerplay at Mirpur is difficult". Difficult is not an analysis; difficult is an experience. The numbers say the dot-ball rate at Mirpur is 49.1 per cent, against 44.6 per cent at Sylhet. The difference is not hard versus easy. The difference is that at Mirpur batters are forced to let more balls go — and whether that restraint pays back with interest in the death overs is the real question.

The dew column has to be read on its own. Of the 431 matches, I could confidently flag dew in 38 — the ball skidding off the pitch on a straight drive, or a spinner's fingers wiping the ball before release. In those 38 matches the second-innings powerplay run rate was 8.02. In matches without a dew flag it was 7.48.

So should you always bowl first on winning the toss? I calculated the relationship between toss decisions and match results in the BPL and it sits close to nothing. Add dew and the relationship lifts slightly, but never strongly enough to call the toss the master key. Dew travels with two companions: humidity and night temperature. When I added temperature and humidity to the model, the separate effect of dew shrank.

There is a signal here that speaks directly to franchise management in this country. Big clubs buy young players and send them out to smaller sides — the parent team gives no opportunity, but the ownership does not change. In that arrangement young batters do not receive powerplay balls, because the powerplay overs are allocated to experienced overseas openers or the franchise's star. Among young batters with five matches or fewer in my ledger, powerplay balls were 11 per cent of their total balls faced. Among those with more than twenty matches, that share was 28 per cent.

The question is structural, not about talent. Who bats in which over is decided by a hierarchy of experience, and in that hierarchy the young batter is always last. Counted that way, the country does not produce powerplay specialists. It produces rescue workers for the last ten overs.

Contrarian: the story of intent, and the column of wickets

Now the part where I break my own favourite idea.

One sentence is close to sacred in our coverage: "You have to attack at the start." The implication is that a higher powerplay run rate means a higher chance of winning. I tested the relationship in my ledger. Between powerplay run rate and winning, the relationship is weak — around 0.19 in my calculation. Between wickets lost in the powerplay and losing, it is considerably stronger. The strongest relationship I found was between wickets in hand at the fifteenth over and the final result.

What follows should change the direction of our argument. The powerplay is not a window for scoring runs. The powerplay is a window for managing risk. The side that makes 45 in the powerplay and loses two wickets, against the side that makes 38 and loses none — the second side wins more often in my ledger, clearly more often at a venue like Mirpur.

The reason is unglamorous. Lose two wickets in the powerplay and the middle overs force caution, and cautious batters eat dot balls against spin. Accumulate dot balls and your freedom in the last five overs shrinks; shrink the freedom and the big shots do not come. In my ledger, sides that lost no wicket in the powerplay scored 9.68 runs in the last five overs. Sides that lost two or more scored 8.31.

A warning here applies to me as much as anyone. Correlation is not causation. Anyone who sees the higher second-innings run rate and concludes that bowling first wins matches is making a mistake. The cause is probably not the toss, probably dew; and the cause of the dew is probably not dew either, but the interaction of night humidity with the ball-change regulations. I have not finished separating those. I have a partial answer, and with a partial answer I will not claim a complete one.

One thing I can state with confidence is the arithmetic of empty columns. In some BPL databases the powerplay runs of abandoned matches sit recorded as zero. I reconciled them by hand: in those matches no ball was bowled, or very few. Leave those zeros in the average and the powerplay run rate falls artificially. Once I set those matches aside in my ledger, the average moved from 7.34 to 7.42. The gap is small; the direction matters. Our problem is not false information. Our problem is that we treat missing information as zero.

Takeaway: what to watch next round

Next season I will watch two things separately. The first is the seventh over. Which side keeps attacking in the over after the powerplay tells me more than the powerplay run rate itself. The second is which young batters are receiving powerplay balls, and what their strike rate is outside the powerplay. Put those two columns side by side and the picture of our talent management develops itself.

The ledger is open, the version number is written, the dates are written. Anyone who wishes can check it. The ledger is patient; the news cycle is not. But if we spend another season writing only about intent and leaving the wicket column blank, then six overs in we still will not know where the match actually bent.