The 22nd-Over Audit: Spin Workload in the Dhaka Premier League, One Number and Its Bill

The 22nd-Over Audit: Spin Workload in the Dhaka Premier League, One Number and Its Bill

On April 11, 2026, I opened my laptop on the steps outside the Sher-e-Bangla National Cricket Stadium in Mirpur for one reason: the left-arm spinner who came on for the 22nd over was the same bowler I had logged in that exact phase for the previous three rounds. The man sitting beside me said the captain had done the right thing. The scorecard closed his spell at 10 overs, 2 wickets, 38 runs. A clean, tidy, quotable number.

My file was collecting a different number — his cumulative overs across rounds one to five, and what his last two overs of each spell had cost. In round eleven, same bowler, same phase, near-identical Mirpur surface: 9.2 overs, 0 for 64. His middle-phase economy, 4.54 across the first five rounds, climbed to 6.91. What the scorecard does not show is what he had already spent to be available for that 22nd over. There is always one lonely number hiding inside the noise, and in this file it hides in the 31st over.

The event universe I counted

The Dhaka Premier Division Cricket League is the country's premier 50-over domestic stage and the place where Bangladesh's home-grown spin pipeline is built. Twelve teams, a round robin, then a Super League; 70 to 80 matches per edition. Fixtures sit five to eight days apart, usually three matches in seven days. Venues are largely the three grounds inside Mirpur, plus Fatullah and Savar. Much of the national spin pool bowls here regularly — Mehidy Hasan Miraz, Taijul Islam, Nasum Ahmed, Rishad Hossain among them — and their domestic overs live in this file.

I logged ball-by-ball records for 198 of the 234 matches across the 2026, 2026 and 2026 editions, or 84.6 percent. The remaining 36 had no video or inconsistent phase detail in the scorecards. The log holds 47,520 legal ball events, plus separate ledgers for 2,114 wides and 311 no-balls. My file runs like a blockchain: every entry timestamped, every correction filed as a new entry, no ability to delete an older number. That is deliberate. In 2026 I worked two weeks on a broken phase definition, and erasing it would have destroyed the evidence.

Back in 2026 I built a 412-player spreadsheet nobody asked for, and it became a witness. The same habit keeps these overs recorded by name, because a transfer window is a spreadsheet with a pulse and a deadline.

I call the metric middle-over spin share, or MOSS: overs bowled by spin between overs 16 and 40, divided by the total overs in that phase, expressed as a percentage.

What the numbers said

Start with the good news. Teams holding a MOSS of 42 percent or higher across the first five rounds won 61.8 percent of those matches. The same rate in 2026 and 2026 was 60.1 and 59.7 percent, so the pattern is stable rather than a one-season accident. In domestic 50-over cricket, spin pressure through the middle phase does a job.

The bad news arrives between rounds nine and twelve, when those same teams' win rate falls to 38.4 percent. Spin economy climbs from 4.52 to 5.79 runs per over, an increase of 1.27. Over the identical window, pace economy moves from 5.41 to 5.72, an increase of 0.31. That gap between two fractions sits at the centre of my calculation. If drying pitches were the whole story, both types of bowler would be paying the same price.

The age split sharpens it. Spinners under 21 averaged 8.1 overs in the first five rounds and 8.4 in the last four — a nominal rise. In the final two overs of their spells, runs per over jumped from 5.3 to 7.8. Teams did not reduce those overs; they kept handing a teenager the ball for the 30th over, in the exact place where he had already bowled 16 overs in the previous four days.

Rotation produces the cleanest separation of all. Teams using four or more spinners in the first five rounds saw their spin economy move from 4.08 to 4.61, a rise of 0.53. Teams riding two or three names went from 4.79 to 6.21, a rise of 1.42 — roughly three times as steep.

The human side

Stopping at the numbers would be a mistake. Four spinners dropped out mid-season in 2026, and three of them had averaged 8.4 overs or more across the five rounds before their exit. Four is a terribly small sample, and I am not claiming causation is proven. But I know one of them: a 20-year-old left-armer from a district outside Dhaka on a match-fee-only deal. Forty-eight overs in six matches. A side strain. Season over. No list carries his name next edition. His seven scorecard lines sit in my file and nowhere else.

A wage ledger sits in the same file — every delayed payment I could verify in Bangladeshi domestic and franchise cricket from 2026 to 2026. The result is uncomfortable: unpaid wages were not an outlier, they were the baseline. A player on a match fee earns his overs and pays for his rest, and that market structure manufactures the workload.

Let me build the mainstream claim first

The oldest line about the DPL is that spin wins and the Mirpur pitch favours spinners. I am not dismissing it. There is a relationship between MOSS and winning in the first five rounds, and the Pearson correlation I calculated is 0.41. In the noise of domestic cricket, that is not a small number, and I have no argument with Mirpur's character.

But 0.41 means the bulk of the variance is still explained by something else — more than 83 percent of it. The second number I kept finding was not spin's quality but spin's depth. Among the high-MOSS teams of the early rounds, those with four or more usable spinners survived to the end; those who loaded the quota onto two shoulders broke. The teams winning more than half their matches in the last four rounds actually carried a lower MOSS, 34 percent, with more spinners in the squad, an average of 4.8. Down the stretch, spin was not winning matches. Squad depth and over allocation were.

This is where the market label points the wrong way. 'Mystery spinner' is a manufactured term: six wickets in three matches raises the price, while the overs stacked on his shoulder appear on no list. A teenager's strike rate spikes overnight and nobody invoices his workload. No auction document asks who will bowl the 31st over.

What the number cannot tell you

I keep a ledger of my own limits. I have no injury data, because domestic cricket does not publish it. Measuring workload by overs and rest is a proxy, not physical load. Release point, rotation speed and right-hand/left-hand batter splits are absent from my file. While reconciling 198 scorecards I found 23 ball-type classification mismatches, corrected against video. The largest correction of all covered 31 matches in the 2026 season, and I no longer use my results from that window. That is the rule of the file: I trust numbers after they survive a pivot table and a bad night.

What I will watch next season

The question needs to change. Today it is: how many wickets does this spinner take? It should be: who bowls your 31st over? A side that can answer in round eleven is the side that survives the Super League. Clubs that treat the spin quota as a round-by-round budget rather than a match-by-match one will shrink that 1.27-run bill. And a small request for the board: publish workload and injury data. A name deleted from a scorecard cannot be recovered, but a timestamped over ledger can remember it.