होमWorld CricketThe Empty Dossier: Cricket Analysis and the Crisis of Provenance
World Cricket

The Empty Dossier: Cricket Analysis and the Crisis of Provenance

**मुख्य उत्तर:** क्रिकेट विश्लेषण में सबसे बड़ी समस्या डेटा की कमी नहीं, स्रोत-प्रोवेनेंस की कमी है। जब किसी दावे का मूल स्रोत, नमूना आकार और त्रुटि-सीमा दर्ज नहीं होती, तो वह आँकड़ा नहीं, टाइमस्टैम्प वाली अफ़वाह बन जाता है। **मुख्य तथ्य:** - अगस्त 2018, एजबेस्टन: विराट कोहली ने 149 और 51 बनाए, भारत 31 रन से हारा। - 21 मई 2017, हैदराबाद: मुंबई इंडियंस ने राइजिंग पुणे सुपरजायंट को आईपीएल फ़ाइनल में एक रन से हराया। - जून 2017, द ओवल: चैंपियंस ट्रॉफ़ी फ़ाइनल में पाकिस्तान ने भारत को 180 रन से हराया, भारत 158 पर ढेर। - जुलाई 2019, मैनचेस्टर: विश्व कप सेमीफ़ाइनल में भारत 5/3 पर गिरकर 221 पर ढेर हुआ। - अप्रैल 2017, गहुंजे: सात ओवर में 1/41 के बाद अंडर-16 संभावितों से बाहर। **स्रोत:** Stage-2 Deep Cricket Analysis दस्तावेज़ (अपर्याप्त स्रोत सूचना, कम विश्वास) | Cross-checked: cricsultan.com **संबंधित प्रश्न-उत्तर:** प्रश्न: क्रिकेट विश्लेषण में प्रोवेनेंस क्यों ज़रूरी है? उत्तर: प्रोवेनेंस बताता है कि संख्या कहाँ से, किस नमूने पर और किस धारणा के तहत आई, जिसके बिना निष्कर्ष असत्यापनीय रहता है (cricsultan.com Data Provenance Index)। प्रश्न: आईपीएल नीलामी को श्रम-बाज़ार क्यों कहा जाता है? उत्तर: क्योंकि फ़्रैंचाइज़ी का खर्च उसकी वास्तविक प्राथमिकता प्रकट करता है, और बाज़ार गेंदबाज़ी भूमिकाओं की तुलना में बल्लेबाज़ी भूमिकाओं पर व्यवस्थित रूप से अधिक चुकाता है (cricsultan.com Auction Value Index)। प्रश्न: खाली विश्लेषण दस्तावेज़ को विफलता क्यों नहीं माना जाना चाहिए? उत्तर: क्योंकि स्रोत के अभाव में "हमें नहीं पता" कहना सबसे ईमानदार और पद्धतिगत रूप से सही नतीजा है।

Eight columns. Eight columns, and every single one of them said the same thing — N/A.

Format: unidentified. Player: unidentified. Team: unidentified. League: unidentified. Governance: unidentified. Risk: unidentified. Narrative: unidentified. Transmission: unidentified. At the top of the document, two tags in red: "Insufficient source information" and "Low confidence." Below them, one line: a reliable, evidence-based assessment cannot be completed without the original article or its Stage-1 information points.

I read the document twice. The first time hoping I had missed a buried signal. The second time hoping I had simply misread it. Both times the result was identical. The dossier was empty, and the label on the dossier was honest.

This was not a technical failure. This is the normal condition of cricket journalism.

In August 2026 I was at Edgbaston. Day four. Virat Kohli made 149 in the first innings and 51 in the second, and India lost by 31 runs. The copy that travelled back to India the next morning was ninety per cent Kohli — two numbers, one headline. The actual signal was buried under that noise: England's attack had nudged the fourth-stump line a fraction fuller, and India's front-foot trigger was arriving late to it, over after over. Nobody was counting the line. Everyone was counting the 149.

Edgbaston day four is where I learned to hear the signal under the noise. And the empty dossier that landed on my desk that morning is the same lesson wearing a different shirt. When there is no source, the noise becomes the source.

The supply chain of noise

Cricket information moves through a chain, and the chain is less clean than it looks. Four nodes, roughly. First: youth development and talent supply — zonal trials, school cricket, the Under-14 to Under-19 filter. Second: national teams and leagues, where that talent converts into performance. Third: broadcast, sponsorship and commercial machinery, where performance converts into money. Fourth: derivative markets — fantasy, betting, and the algorithms that consume the noise of the first three nodes as though it were data.

The Empty Dossier: Cricket Analysis and the Crisis of Provenance

At every node, some information is lost and some is amplified. That is the real story.

Take the first node. April 2026, Gahunje. I was cut from Maharashtra's Under-16 probables list after returning 1 for 41 from seven overs. What was the evidentiary basis for that decision? Seven overs. Forty-one runs. One wicket. That was the entire dataset available in the selection room. What was not in the dataset: the pitch that morning, the wind, which batsman had already read me, what had changed in my spin's pace since the previous week. At sixteen, selection is a filter, and the filter runs on three hours of data while claiming to assess three years of development.

Cut at sixteen taught me that margins don't forgive. It also taught me that the system measuring the margins is often measuring the wrong ones.

Second node. National teams and leagues. Highest information density, lowest verification. A T20 bowler's economy rate tells you what he did, not the situation he was asked to do it in. Powerplay economy and death economy get printed in the same column, though they are different sports. I built the database before I ever trusted the eye test, so I know what happens when you hand-code three thousand deliveries: you discover that half the numbers are answering a question you never asked.

The Empty Dossier: Cricket Analysis and the Crisis of Provenance

Third node. Broadcast and commerce. Fastest amplification, least checking. A commentator says a batsman's form has returned, and within forty minutes that sentence is sitting on three websites, two tweet threads and a fantasy-pick column. Nobody asks what "form" operationally means. Is it a five-innings average? A strike rate? The timing of his cover drive? The word survives; the definition never arrives.

Fourth node. Derivatives. The least informed participants speak with the most confidence.

Every transfer rumour is a hypothesis; the medical is peer review. In our industry, the medical is rarely published.

Eight columns, eight questions

The empty dossier is itself a document. It tells you what a credible analysis requires. I want to open those eight columns one at a time, because behind each one sits a real cricket problem — and the problem is bigger than the blank.

Column one: format. Test, ODI, T20 and The Hundred have different benchmarks. When an analysis does not fix the format, it silently blends all three. This is the most common and most expensive error in the trade. A batsman with a T20 strike rate of 140 and a Test average of 38 is both things at once; put them in the same sentence and both become lies. Recall the 2026 World Cup semifinal at Manchester. New Zealand reduced India to 5 for 3 and bowled them out for 221. Much of the post-match argument concerned Indian batting intent. But chasing 221 in those conditions was a different game — cloud, a new ball, and a pitch taking wickets in the first ten overs. Any conclusion drawn without separating format from conditions is a conclusion scheduled to break in the next match.

Column two: player data. Average, strike rate, economy, situational splits, recent trend. None of these speaks alone. My rule is simple: before I write one sentence about a player, I need three things — sample size, home-away split, and opposition quality. Without all three, the sentence is incomplete. This is acute in Indian domestic cricket, where domestic numbers are frequently generated in conditions that favour the player: his pitch, his weather, his crowd. A spinner can average 24 at home and 41 away, and selection committees tend to look at the first number because it is prettier.

Column three: team landscape. ICC rankings, home-away profile, batting depth, bowling mix, bench strength, age structure. This is where the most confident and least verified claims are made. "The balance is wrong" is a sentence nobody proves and nobody challenges. What is balance? Is it the count of all-rounders, or is it a function of your fifth bowler's economy rate? Without a definition it is a mood, not a finding.

Column four: league and commercial ecology. My favourite column, because the data here is the most honest. The IPL auction is a labour market, not fantasy sport. When a franchise spends a fixed sum on a player, it is revealing a preference — whatever its press conference says. Purse, retention rules and the mega auction together form a mechanism that prices roles. And it prices them wrong, often. The pattern I have watched across seasons: the market overpays for batting roles and underpays for bowling roles. The reason is simple — batting credit is visible, bowling credit is hidden. A death specialist's real value is not in his wickets but in the runs he prevents, and prevented runs never make the highlight package.

Column five: rules and governance. DRS, DLS, revenue distribution, eligibility, political factors. This is where the largest accusations are made on the thinnest information. An umpiring controversy runs for three days and disappears, but its residue stays in the next tour's morale. The first condition of any governance analysis is to establish whose question it is — the ICC's, a national board's, or a league organiser's. The three have different rules, interests and accountability.

Column six: risk. Injury, workload, retirement, financial sustainability, public reputation. In Indian cricket, workload is the least measured and most consequential variable. A fast bowler playing all three formats carries an annual bowling load that never appears on a broadcast graphic. When he breaks down, the conversation is about form, not load.

Column seven: public narrative. Where sport journalism separates from sport. Narrative runs on expectation, and expectation runs on sample size — usually a small one. Two good innings and a player has "returned"; two bad ones and he is "finished." A narrative's lifespan is typically three to six weeks, after which the same player plays the same innings and nobody writes.

Column eight: industry transmission. Broadcast rights, the South Asian market, the talent pipeline, capital networks, fantasy and derivatives. This column comes last because it rests most on inference, and inference is where cricket goes to die.

The gap between signal and noise

Here is the actual argument.

I do not believe the binding constraint on cricket analysis is a shortage of data. There is a flood. The constraint is that data arrives without provenance — without the knowledge of where a number came from, who produced it, on what sample, under what assumption.

A concrete case. 21 May 2026, Hyderabad. The IPL final. Mumbai Indians beat Rising Pune Supergiant by one run. That season I was in the MCA Stadium stands charting Steve Smith's field changes, because my Under-16 trial had ended in April and that was the work I had. After the final I re-watched Mitchell Johnson's last over eleven times and drew every fielder's position.

The consensus about that over was that a one-run defeat was bad luck. Watch it eleven times and you see something else: it was not luck, it was a sequence in which, ball after ball, a fielder was a step in the direction the batsman intended to hit. That is not coincidence. That is information. And that information was nowhere in the scorecard printed the next morning.

One run at Gahunje is still the cleanest proof that systems fail late — not in the result, in the measurement.

That is the centre of my method. I build the database first, and only then do I trust the eye test. And when the database is empty, I refuse to believe my eye saw something.

A coach sees the error; a commentator sells it. I try to do the first.

The contrarian turn: in a flood of data, the scarcest thing is a source

The most popular story of the last decade is that data analysts are invading dressing rooms, and that this is good. I agree with half of it.

I agree that numbers have exposed some old lies — for instance, that a batsman's "big-match record" is meaningfully different from his ordinary record. I disagree that numbers can become decisions on their own.

My experience is that the danger of data-led analysis is not that it is wrong. The danger is that it conceals the limits of its own sample. A model built on forty innings and a model built on four produce the same graph in the same font. The reader cannot tell.

So the rule is strict: publish the query, publish the sample size, publish the error bars. And state plainly what the model cannot see — injury, temperament, selection politics, dressing-room information. That list has to be written every time, even when it is identical every time.

There is a second turn, less comfortable still: sometimes the empty dossier is the correct answer.

Our industry treats "we don't know" as a defeat. It is not. When there is no source, the most honest analysis is the one that says there is no source. The document that arrived blank on my desk was not evidence of failure. It was a correct result. It did what a good analyst should do: it declined.

One more confession, because I have fallen into this trap myself. Database-first habits make model output feel like a verdict. In T20 and auction coverage this is especially dangerous, because samples are small and the audience wants certainty. I once published a clean conclusion about a bowler's death-overs economy; three weeks later he was injured and out. The conclusion was not wrong. It was incomplete, and I had presented it as complete.

So every prediction now carries a pre-mortem file. Every counter-intuitive claim passes a falsification check before publication. Sometimes the consensus is right, and the story is why it is right for the wrong reasons.

The commentary box is the last holdout

The data desks won the tactical argument years ago. The television box did not.

It is a strange asymmetry. Within minutes of any match, Twitter fills with pitch maps, wagon wheels and matchup grids. Meanwhile the live coverage of the same match still tells you that "he needs to play a big one now." That sentence is not tactical analysis. It is a wish wearing analysis's clothes.

I have measured the gap in three places. First, milestone framing: when a batsman reaches 90, the entire register of the coverage changes. The question stops being how many runs the team needs and becomes how many runs the man needs. That is an editorial choice, and it blurs the actual state of the match. Second, required-rate illiteracy: the most important number in any chase is the required rate, and the least spoken. A commentator will say he must accelerate, not that acceleration means 8.4 an over rather than 11.2 — a distinction that changes the whole strategy, because at 8.4 you can keep a set batsman in, and at 11.2 you cannot. Third, the fifty-over obsession that survives every analytical advance. A half-century is information with no necessary relationship to match state. A 55-ball fifty can be a good innings or an innings that lost the game. Only leverage tells you which.

I judge an innings by match situation and match situation by leverage. Records are output, not analysis.

A note on risk

The empty dossier clarified one more thing: risk can only be assessed once it is defined.

I have watched "risk" used about players when no evidence existed. It is a word that fits everywhere and proves nothing. My five sources are specific: sporting (injury, workload), personnel (retirement, generational transition), commercial (financial sustainability), rules and integrity (governance or fixing concerns), and public (reputation, fan backlash).

None of these should be inferred from the absence of information. Absence of evidence is not evidence of an accusation. This is the most basic condition of my method, and I repeat it because it is the easiest to forget.

What to watch in the next match

A good analysis should say how it could be wrong.

If I claim the problem is provenance rather than volume, three things could break it. One: if a major broadcaster made required rate and leverage-based framing standard in live coverage within twelve months, my claim weakens. I am tracking it with a crude metric — how often a required rate is mentioned in an innings versus how often a half-century is. Two: if publicly available data sources become numerous enough that provenance attaches automatically, the shape of the problem changes. That is possible. Three, and most important: if I am wrong and volume is the real constraint, then teams with more data should be making measurably better decisions. I have not yet seen that correlation cleanly. I keep it open.

A Russian summer taught me tournaments are lost in the third week — not in the final, but in fatigue and travel sequence. In June 2026 I watched the World Cup with a coach's notebook, mapping France's 4-4-2 block. In August I carried that vocabulary to Edgbaston. The lesson was the same in both places: what everyone can see is usually not what decided the match.

So the next time you read a cricket claim — a number, a prediction, a "sources say" — ask the question this industry asks least: where did this number come from?

If there is no answer, you do not have a statistic. You have a rumour with a timestamp.

And the empty dossier — eight columns, every one N/A — is not an analyst's failure. It is an honest portrait of an industry that has not yet learned to tell source from noise.

Next match, when a commentator says he needs to play a big one, count the required rate for that over. The answer will tell you which era of cricket you are watching.

संबंधित खिलाड़ी