Where the Auction Light Never Falls: BPL's Retention Window and a Ledger Counted by Hand

The retention list arrived at twenty-two minutes past eleven at night. Six names, one PDF, and zero explanation. Sitting at home in Khulna, I was hunting ball-by-ball numbers for three of those six across the last two seasons — what they had produced per ball in the death overs, what they had done in chases once the required rate climbed past nine. No provider sells that number for BPL domestic batters. So the question I had spent half an hour with had no answer in any feed.

What stopped me was not the list but a name missing from it. In my ledger, the best death-over strike rate in the league over the last two seasons belonged to a left-handed middle-order batter who went unpicked in each of the last two auctions. The man retained at a top-bracket fee had a raw strike rate near the league average. Two numbers, two opposite stories, and between them an empty cell where nobody had written the question.

Context

Cricket's so-called transfer window is not football's. There is no release clause, no January panic-buying. There are three separate processes — retention, the draft or auction, and the NOC, the no-objection certificate. The IPL auction makes every number public: on December 19, 2026 in Dubai, Mitchell Starc went to Kolkata Knight Riders for 24.75 crore rupees, and in the same auction Pat Cummins went to Sunrisers Hyderabad for 20.5 crore. On November 24, 2026 in Jeddah, Rishabh Pant went to Lucknow Super Giants for 27 crore and Shreyas Iyer to Punjab Kings for 26.75 crore.

The BPL runs on a different scale. Squad budgets here are a fraction of an IPL figure, retention and a board-controlled draft replace the open auction, and most importantly there is no published method for valuing a domestic player. Where the IPL makes both price and performance public, the BPL lets you guess the price and hides the performance. That vacuum is where my work lives.

The overseas market is messier still. In January and February, the ILT20, the SA20 and the BPL sit on each other's shoulders. A Caribbean all-rounder cannot sign for three leagues in one month, so who gets an NOC first directly changes match results. That information is not gathered anywhere either; I have had to read each board's separate announcement and copy it into the ledger by hand.

Core

In 2026 I did this work first in football — 24 matches at Khulna District Stadium, a paper grid, and a homemade model built from shot angle, distance and defensive pressure. It placed a 23-year-old winger above the league's leading scorer. The 900-word piece got 60 shares. I kept the notebook anyway, because a league nobody counts turns counting into a habit. I built the model by hand, because the league deserved to be counted.

Now that notebook is on cricket. 92 matches across the last two BPL seasons, 11,847 legal deliveries hand-coded from scorecards, and 214 batters who faced more than 60 balls — that is my sample. Cut-off date, February 15. No provider would chart it, so the counting became a kind of prayer.

My main metric splits in two. First, phase-adjusted strike rate — overs 16 to 20, and chases where the required rate sits above nine. Second, pitch bias — Khulna and Sylhet decks are flat and high-scoring, Dhaka's slow surface is not, so I looked at ground-adjusted runs rather than raw ones.

Three findings. First: a domestic batter fourth in the league on raw runs was first on phase-adjusted strike rate. He was not retained. Second: overseas power-hitters' numbers were inflated by flat decks. The same batter struck at 162 in Khulna and Sylhet and 118 in Dhaka — a 44-run gap per 100 balls. A franchise that reads only total runs is really picking its fate off the pitch report. Third, and most uncomfortable: a bowler played one season, took 14 wickets in 11 matches at an economy of 7.8. In the next two auctions nobody could even raise his name, because an 11-match sample cannot persuade any model. Mine flagged him as insufficient data — true, and cruel. Every number is a person who never got to explain themselves.

The same gap runs through the Dhaka Premier League and the National Cricket League. Hundreds of matches are played there each season, and there is no public performance database. Men who have scored for six or seven years never once appear in a shared list.

Part of my method is deliberately slow. After every match I write a short note beside the scorecard — which bowler bowled which over, which way the wind blew, how long the lights stayed on under DLS. None of it fits a metric, but later it is the only witness I have. Sitting at Sheikh Abu Naser Stadium in Khulna, I have watched the same batter become two different people in afternoon sun and evening dew. The scorecard does not know that difference.

I have to say plainly what my model cannot see: field placements, pitch wear, a bowler's NOC calendar, injury history, dressing-room politics — none of it is in there. My noise log, where I file statistics that look meaningful and explain nothing, is topped by total runs and most sixes.

Contrarian

The story everyone tells about auction price and performance is conveniently simple. My ledger says that across the last three BPL windows, only 38 percent of players on top-bracket fees stayed in the league's top thirty percent on phase-adjusted metrics the following season. The other 62 percent were paid for something that was not batting — agent networks, the NOC calendar, the shape of a wage bill, a franchise's marketing arithmetic. Price and skill are not the same thing; price is a decision made at a small table where a cricketer may not be present.

The second trap is the imported Moneyball story. Dropping a generic analytics template onto Bangladesh's domestic game goes wrong, because seasons are short, matches are few, wickets behave differently, and NOC rules keep the entire overseas pool unstable. A franchise that copies an outside template loses its own conditions. There is another layer I rarely put in print: the ball-by-ball feed is sold on to commercial markets, where a player's worth becomes a live line nobody asks to explain.

Takeaway

In the coming window I will watch three things. One, the names dropped from retention lists — I will build their performance cards myself. Two, the NOC timing of overseas players: in the February-March pile-up, who gets released first will decide several teams' fortunes. Three, whether any franchise publishes its own valuation method — because until one does, players will be sold in a market whose pricing rules they are not allowed to read. So the question is simple: before we count the players, will we ask them which number actually knows them?