From Scorebook to Auction Table: The Invisible Value of Data in Bangladesh Cricket

A 2026 afternoon on a club ground in Mymensingh. A folded notebook in the coach's pocket, nothing but a wristwatch on the players. That season I logged 1,440 balls from 12 BPL matches by hand — which ball came in which over, which batter stood under pressure, which bowler bowled the dead overs. One name surfaced whose scorecard was entirely silent: average 28, strike rate 132. On paper, ordinary. But in the last five overs his strike rate was 187, and across those 12 matches nobody scored more death-over runs than he did. At the next auction he went unsold. Nobody saw his last-five-overs number, because that number was never written down.

The notebook was my first model, and Mymensingh was my first laboratory. From that ground I learned that cricket loses information in two places — outside the field, and inside the numbers.

The Bangladesh Premier League began in 2026. Since then a distinct economy has settled over Asian franchise cricket — auctions, drafts, retention, trades. After the IPL launched in 2026, the model spread to the Pakistan Super League, the Lanka Premier League, ILT20. It is not a year-round negotiation like football's transfer window; in cricket a player's price is set once a year, within a few hours. Whatever data sits on the table before those hours is what fixes the price.

That is exactly the problem. Asian domestic cricket has almost no public ball-by-ball archive. Not a single ball of the BPL, the Dhaka Premier League, or the national league is stored anywhere. So players are valued by memory and reputation, not by sample and evidence. Whatever a selector happens to remember becomes the player's market value. Bangladesh gained Test status in 2026 and won its first Test in January 2026 in Chattogram against Zimbabwe — yet the complete ball-by-ball dataset of that historic match cannot be found anywhere today.

In this window, then, the real story is not in the retention rules but in the absence of data. Which team pays what for whom depends on three things — squad gaps, the wage cap, and matchups against specific opponents. Of these, the third is the weakest, because matchups need samples, and samples need stored data. In a league where not every ball is recorded, matchup analysis is only guesswork.

In T20 cricket, batting average and strike rate alone can never measure value. Average tells you how often a batter has returned not out; strike rate tells you how many runs came per ball. Neither says in which over, or under what circumstance, those runs came. There are two routes to 60 off 40 — on a flat pitch in the powerplay, or chasing 150 in the closing overs. Their values are never equal.

So I split the innings into three phases — powerplay (1–6), middle (7–15), death (16–20). Each phase gets its own strike rate, its own boundary percentage, its own dot-ball percentage. A batter with an overall strike rate of 130 but 170 in the death overs is worth far more than his overall number suggests. The reverse is also true: 140 in the powerplay but 110 at the death — that batter is a burden in T20.

In that notebook of 1,440 balls, one pattern kept returning. A bowler with an overall economy of 8.2, which does not look bad. But split apart — 6.1 in the powerplay, 11.4 at the death. Use him in the powerplay and he is gold; use him at the death and he is a loss. Buy him at auction as a death bowler and the team loses; buy him as a powerplay bowler and the team wins. The same player, two different prices — the difference lies only in the splitting.

The ground must be split too. The pitches of Sylhet and Dhaka are never the same. The same strike rate carries two meanings on Sylhet's spin-friendly surface and Dhaka's batting-friendly one. If I do not produce a venue-adjusted strike rate, I will value a bowler under his own home advantage and a batter under his own disadvantage. This is the most common error in Asian domestic valuation.

Matchup models are subtler still. An off-spinner's record against left-handers, or a leg-spinner's numbers against right-handers — these decide team composition. But here lies the sample trap. In one league season a bowler may face perhaps 30 balls in a specific matchup. No decision survives on 30 balls.

When the sample is small, silence is the better answer. In 2026, when home advantage collapsed in empty stadiums, I watched 306 matches and saw the coefficient fall from 0.41 goals to 0.17. My manager wanted a quick fix. I waited for a 20-match sample, because below 20 the number is only noise, not signal. The broken model taught me more than the accurate one ever did.

So beside every number I write a confidence interval. A strike rate of 130 is not enough; it must read: strike rate 130, confidence range 118 to 142, sample 320 balls. A number without a range is a claim; a number with a range is evidence. In an auction where buyers are not used to seeing confidence intervals, the biggest risk is taken by those who treat a single number as truth.

Russia 2026 had become a database before it became a memory — 64 matches, 1,842 shots. That was football, but the lesson holds in cricket too: every row was a small argument against chaos. A season in which data is not preserved returns only as a story, never as evidence. I did not discover expected goals; I submitted to them, one page at a time — and that submission taught me that every ball in cricket demands the same care.

So in my method no single number ever stands alone. To value a batter I combine at least four layers — phase-based strike rate, boundary percentage, dot-ball percentage, and wicket-equity (how much impact his dismissal had on the match, in context). For a bowler — economy, strike rate, powerplay-death split, and catch-dependency rate. Together these four build a range, not a single figure. When one metric wobbles, the other three hold its balance.

Bowling workload and injury history are also variables of value that almost nobody checks at auction. A bowler who sent down 300 overs in a season carries far higher injury risk the next — yet his price rises at auction because his recent numbers shine. Data cuts both ways here: the same number lifts his value while hiding his risk. That hidden risk is what wrecks many teams' wage bills.

There is an uncomfortable truth here. At auction, price is set by scarcity, not performance. If only two specialist death bowlers exist in the market, their price will far exceed their true contribution. If ten middle-order batters are available, their price drops. Market price and true value are never the same. The team that understands this difference wins matches beyond its budget.

And this is where places like Mymensingh matter. Handwritten scorebooks on local grounds, the behaviour of local pitches, records of informal matches — all of this is raw material. The answers to national-scale questions hide inside this raw material. But the problem is that nobody records it. The talent built on the ground disappears before it ever reaches the auction table.

Bangladesh's domestic structure holds a particular contradiction. On one side, Dhaka Premier League clubs play a long season; on the other, none of those matches has a central database. So a player who scores 600 runs in one season has his next season's value set by a newspaper headline. Unless this gap is closed, no bridge will form between domestic performance and national selection.

The IPL now runs a separate data pipeline for every ball — tracking cameras, ball-tracking, field-placement mapping. There, player valuation is far less memory-dependent. The rest of Asia's leagues lag well behind that infrastructure. If Bangladesh, Sri Lanka, or Pakistan's domestic competitions began storing data at the same level, the auction map would change within five years.

And here the question of data permanence arises. For Asian domestic cricket, the real solution may be a blockchain-like, immutable ball-by-ball ledger — where every ball, every source, every verification is permanently recorded. No one could later alter or erase a number. When data becomes immutable, it stops being memory and becomes evidence.

I record the source beside every note. Where I logged a ball from, who saw it, how reliable it is — without this audit trail a number is only a claim. The provenance of data and its verification should be the foundation of cricket valuation, or else we will mistake reputation for data.

So for the next window I build a scenario tree, not a direct prediction. Branch one: if retention rules tighten, teams lean toward domestic talent, and demand for phase-based data rises. Branch two: if the overseas quota grows, domestic middle-order players come under pressure, and matchup data enters decisions. Branch three: if the wage cap falls, budget-friendly specialist players rise in price. In each of the three branches the winner is the team that already knows which piece of data it needs.

A small team beating a giant — behind that story hides financial inequality, and it is not sustainable. When a domestic side defeats a big one, we call it a fairy tale. But the numbers ask: how solid was that win? Can the same side hold the same standard next season? If not, it is not a fairy tale but a one-off spike in an irregular sample. Romance presses on numbers, but numbers do not believe in romance.

We too easily find causes in auction outcomes, but correlation is not causation. A team paid a high price for a player, therefore that player is the best — that conclusion is wrong. Perhaps the team had a gap for that specific role, or a rival's pressure pushed the price up. Price and ability — the link between them is weak, because one is measured by the market and the other by the field.

Another trap — data itself can deceive. Piling up runs in dead matches, taking wickets in result-less games, inflating strike rate against weak opponents — these make numbers look good but do not raise value. A team that buys on numbers alone falls into this trap. A team that tests a player in his specific role avoids it.

So at the next auction table I will sit down with one question: in what circumstances were this player's numbers made, and will those circumstances return in my team? As long as every ball of Asian domestic cricket is not preserved, a gap will remain between price and value. The team that learns to measure that gap first will be ahead not only at auction but on the points table. Chaos is never invincible in cricket — it has merely not yet been collected.