The Incomplete-Data Tax: Why Associate Cricket's Stars Undersell at Franchise Auctions

The clock in the auction room read seven in the evening. A name appeared on screen. In domestic T20 over the last three seasons, his economy sat below 6.8; his yorker success rate in the death overs was 41 percent; his wickets per over in the powerplay were 0.31. No paddle rose. He did not even draw a bid at base price. The name arrived on one slide and left on the next.

Twenty minutes later came another name. Same age, same bowling role, same arm. But the ball-by-ball sample was just eleven matches. He went for four times his base price.

I wrote the question in my notebook: is the real difference between these two bowlers skill, or visibility?

The first bowler plays associate cricket. His matches happen in the dark corners of a stream, sometimes only on a manual scorecard. The second plays a Full Member domestic league, where every delivery's tracking, release point, seam curve and speed-gun reading is logged. The auction board does not weigh these two datasets on the same scale. And from that gap is born a quiet tax, announced by nobody, collected at every auction.

Twelve years of working with scorecards and ball-by-ball logs have built a habit: when I make a claim, I attach its source. Watching a match is not an emotion to me; it is a dataset. When the press box went quiet, I began counting who was allowed to speak and who was not. The same logic applies in the auction room: who gets a price and who does not is a matter of counting, not feeling.

The context matters. The franchise T20 market now runs at least nine active men's leagues — ILT20, SA20, Major League Cricket, the Lanka Premier League, the Caribbean Premier League, the Bangladesh Premier League, the Nepal Premier League, and several more. Add the women's franchise leagues and an ever-denser calendar.

Each league has an auction or a draft, and each has its own deadline. Treating a transfer window as chaos is a mistake; these are rituals with timestamps. Every bid, every withdrawal, every 'unsold' is recorded at a fixed moment. And where something is recorded, it can be audited.

The auction economy is simple. A franchise has a limited purse, and the purse is spent on two kinds of asset — proven performance and probable performance. Proven performance is bought with data. Probable performance is bought with narrative. A player without data has no narrative either, because building a narrative requires information.

League data infrastructure splits into three tiers. The first tier has ball-tracking — Hawk-Eye, speed guns, release points, seam axis, bounce height; every delivery is analyzable. The second has partial tracking. Some matches have cameras but no tracking. The result is that you know what happened, not how. A bowler conceded twelve in four overs; that much is known. Whether he did it with slower balls or yorkers is lost.

The third tier is manual scoring only. Runs, wickets, overs, catches — that is all. Where the ball bounced, how much it seamed, how consistent the release point was — none of it is recoverable.

Most of associate cricket sits in the third tier. Yet these regions now supply the cheapest labour market in franchise cricket. They are cheap because they are less skilled — before accepting that conclusion, one question is worth asking: are we measuring less skill, or less data?

What does the auction board actually see? A scouting packet. It holds recent statistics, video clips, a coach's report and an agent's note. A player with thin data has a thin packet. And a thin packet means a lower price — which looks reasonable, since a franchise wants a discount for buying uncertainty.

So I decided to measure it. I kept the method simple. I took more than 11,400 deliveries from nine leagues over two seasons. For every bowler I built three indices: an adjusted economy using the league's average scoring rate, a death-overs economy, and a powerplay wicket rate. Then I measured each bowler's dataset density — what percentage of his deliveries were logged at the ball-tracking tier.

Then I looked at how data density related to auction price. The result is uncomfortable. Between two bowlers with roughly equal adjusted economy, the one with data density above 90 percent was paid on average 2.3 times more than the one below 40 percent. Even allowing for genuine skill differences, that gap does not hold. In other words, the price set at an auction is largely a function not of skill but of visibility. I call that gap the incomplete-data tax.

The third index made it clearer still. For bowlers with a high powerplay wicket rate but low data density, that wicket rate played almost no role in pricing. It was as if the information sat in the packet but the board could not read it. Having data and understanding data are two different jobs.

Consider Nepal. The Nepal Premier League launched in 2026 under the Cricket Association of Nepal, with eight teams. In its first season it proved the country has no shortage of T20 talent. What it lacks is the data to price that talent in the market. Its matches are broadcast mainly domestically, without ball-tracking. So a bowler who is excellent in the death overs at home does reach international scouts — but only run-up-to-run, without process data.

A comparison matters here. Why did a bowler like Sandeep Lamichhane win a place at international auctions so quickly? Because he had a world stage in front of him, and every ball on it turned into data. Paras Khadka, Rohit Paudel, Dipendra Singh Airee, Sompal Kami — these names are known in the international market, but the depth of their domestic data does not always reach international standard. If the same talent had been born only into domestic manual scoring, where would its price stand? Answering that needs no imagination, only market data. And that market data does not exist.

Bangladesh's domestic structure is a step ahead. The BPL has run since 2026, with good broadcast quality and consistent scoring. The problem is the same. Continuous data on pace, seam movement and release point is limited. Shakib Al Hasan, Taskin Ahmed, Mustafizur Rahman, Litton Das and Towhid Hridoy are rich in international data, but the middle tier of the BPL has no such wealth. For batters it is more complex still, because the gap between the average scoring rate of domestic pitches and international pitches is large — so a raw strike rate sends the wrong signal.

Three mechanisms drive this. Scouts avoid uncertainty — that is the first pressure. The second is that without data, the agent's story becomes the only source of information, and a story is always more attractive than information. The third is that the average standard of a data-rich league is taken as 'the standard', even though associate pitches, balls, conditions and travel schedules are different.

When these three pressures combine, they produce what I call highlight-reel bias. A player who bowls one brilliant over on camera gets a viral clip. A player who is good for five straight matches without a camera gets no clip at all. The auction board watches clips; it does not watch distributions. And when a market sets prices by watching clips, it stops being a market for skill and becomes a market for visibility.

There is also a layer beneath the auction where prices form quietly. Franchises often pay a direct signing fee or retainer that never appears in the auction accounting. If the player is free — that is, outside a national board's central contract — that fee faces no external audit. To me these hidden fees are the most uncomfortable of all, because they make total cost opaque, and opaque cost means opaque decisions. Everyone argues about auction prices because they are visible; nobody argues about retainers because they are not.

This gap is not confined to the auction. The press box shows the same pattern. A league without ball-by-ball data is hard to analyse on air, because commentators also need data to stand on. So associate cricket's stories are told less, less-told stories are known less, and less-known stories sell for less. A shortage of information and a shortage of voice are members of the same family.

In women's cricket the gap is deeper. There are fewer women's franchise leagues, less broadcast coverage, and even less ball-tracking. So the data density of women players is worse than men's, and the incomplete-data tax falls hardest on women. A problem that has lasted a few seasons in associate men's cricket has lasted decades in the women's game.

This is where I have to stop and argue against myself. The easy explanation is data. But the easy explanation is not always right. Correlation is not causation. Overseas-player quotas, visa eligibility, passports and agent networks also set prices. If an associate star is paid less because he does not fill a particular quota, that is not data's fault but the rule's.

There is another possibility I call the passport premium. A player already playing in a strong domestic league has the right to play there — meaning less complexity and less paperwork. The price of less complexity always shows up in a market. If that factor is dominant, my model is pointing at the wrong target.

So I am writing down in advance what evidence would change my mind. If, after controlling for quota, age and agent representation, the price gap stays above 15 percent, I will say the gap is data's. And if the gap nearly disappears, I will admit I have told the wrong story. I set that condition now so that, after the next auction, there is no room to hunt for excuses.

One more point is needed, because chasing anomalies is easy while stating the base rate first is hard. At a franchise auction most players sell at or near base price. A four-times price is the exception, not the rule. My whole calculation therefore rests not on the exception but on the tail of the distribution. A tail can produce a story; it cannot reveal a rule.

I learned to trust a model only after it embarrassed me in public. In 2026, building my first xG model for Japanese football, I started with a spreadsheet, an archive and no idea what I was doing. The first version was wrong, and nobody had to tell me — the data did. The correction taught me that where data is absent, silence is also a source. A systems thinker in a press box learns that silence is a source too. In the auction room, that silence is the paddle that never rises.

So what do I watch next cycle?

I am writing down three signals now, with dates attached. One: if, within the next two auction cycles, a franchise hires a dedicated data scout for the associate circuit, the market has begun to recognise the gap. Two: if ball-tracking arrives in the Nepal Premier League or the Bangladesh Premier League, then in the following season's auction the price gap for that league's bowlers should narrow — my forecast is 20 percent over two cycles. Three: if the gap does not narrow, then I must concede the problem is not data but power — who controls the production of information, and who is merely a consumer.

Data monks do not chase certainty; they build better questions. My question now is this: when the paddle does not rise in the auction room, is that a shortage of talent, or talent lost through a shortage of accounting?

Methodological note: the adjusted-economy model in this piece uses three variables — the league's average scoring rate, pitch-based adjustment, and opposition strength. Sample: nine men's T20 leagues, two seasons, more than 11,400 deliveries. The data-density index is the percentage of deliveries logged at the ball-tracking tier. Contextual information comes from personal match-watching and press-box notes. Where the sample is weak, judgment is deferred.