Zero Information Points: The Cricket Analytics Failure That Never Reaches the Scorecard

Eleven fields loaded on the monitor in the Dwarka office on a Monday morning, and eleven of them were empty. Article Title — N/A. Source — N/A. Article Type — unclassified. The domain label carried exactly one tag: cricket_world. And beneath that, the line that stopped the work: Information Points — an empty array, length zero.

The eight analytical dimensions — format and match, player technique and data, team landscape, league and commercial ecosystem, rules and governance, risk, public narrative, and industry transmission — were all resting on that single array. The array was empty. Which meant the work could not begin at all.

The post-mortem began where the scorecard stopped explaining anything, except this time the scorecard had not arrived. Zero of eleven fields were usable. Not one value could be carried forward into an argument. And the real question that morning was not "what is the story". It was "what data gives me the right to write the story".

An analyst has two roads at that moment. The first: fill the template. Generic IPL economics, generic DLS grievance, the generic ageing-core argument, delivered on time. The second: write that the input did not arrive. The first saves a career and kills a profession. The second does the reverse.

What the pipeline is, and why format is the load-bearing variable

The standard two-stage model of cricket data analysis begins with text decomposition: taking an article, report, scorecard or press release and breaking it into Information Points — small, verifiable, factual units. Who played, where, in which format, what score, which venue, which decision, what was said, on what date. Those units become the evidence for everything downstream. No evidence means no conclusion — only opinion wearing a conclusion's clothes.

The second stage runs an eight-dimension framework over those units. And the first question in that framework is always format: Test, ODI, T20, or franchise league. This is not procedural box-ticking. Cricket is perhaps the only sport where the same number for the same player means three different things in three different worlds.

A strike rate of 140 is mid-table in T20 and unthinkable in a Test first innings. An economy of 8.5 is expensive in T20 and survivable in the middle overs of an ODI. A batting average of 40 identifies a competent Test opener; in T20 the same average tells you only how often he was dismissed, which is not the point of the format. A fast bowler's average of 25 is ordinary in Tests and close to irrelevant in T20, where his value is set by economy and dot-ball percentage rather than wickets.

Absent format, every number is meaningless. And absent format, every sentence written about that number is technically a guess presented as analysis. That is the most common error in my trade, and it never looks like a typo. It always looks like a confident paragraph.

The audit: eleven out of eleven

Article Title — absent. No subject to anchor anything. Source — absent. Source-grading impossible, and worse: no conclusion can be verified, because there is nothing to verify against. Article Type — unclassified. We cannot tell whether we are holding news, analysis, opinion, a press release or a rumour.

Domain label — cricket_world. That tells us the sport, and nothing else. No format, no gender, no competition level. It is a category tag, not evidence.

Information Points — empty. That is the critical failure, because everything else derives from it. No atoms, no molecules.

Core viewpoints' one-sentence summary — blank. Author stance — N/A, which means the source's own tilt cannot be graded, and that is a heavy loss, because source tilt is the primary input to narrative-sustainability judgement. Article purpose — N/A, so reporting cannot be separated from promotion. Entities — "identify from the information points above", which is an instruction, not an answer, and an instruction addressed to an empty set. Time Sensitivity — "not assessed in Stage 1". Source Quality — "judge from the source fields of the information points", when no source field exists.

And the Additional Notes section contained no notes. It contained instructions to the analyst. That is the bureaucratic irony no dashboard caught.

Zero of eleven fields were analysis-usable. The single non-null value — cricket_world — was a category, not evidence.

The grammar of null

One thing must be stated plainly, because it is the most common error in the profession: null does not mean negative.

When every governance checklist item returns "N/A", that does not mean "no governance issue". It means "the article was never examined". When the risk rating reads "N/A — cannot be assessed", that is not a synonym for low risk. N/A is not low risk. N/A is not a clean bill of health. N/A is not no news.

A risk rating is a statement about likelihood multiplied by impact of identified risks. If no risk can be identified, writing "Low" is false reassurance — it implies the article was examined and found safe. It was not examined, because there was nothing to examine.

The third-umpire analogy holds. When a side reviews and the replays show nothing conclusive, the TV umpire gives it not out. He does not write in his report that the batsman was proven innocent. He writes that the evidence was insufficient. Cricket's entire following understands the difference between those two sentences. Analytics loses it repeatedly.

Eight dimensions, eight empty answers

Format and match: no fixture, no series, no competition. No powerplay, middle-overs, death-overs or new-ball data. No venue, pitch type or home-away designation. No weather, dew, or DLS reference. Not one sentence can be written here without inventing the subject.

Player: not a single cricketer, coach or selector is named in the payload. This is not data poverty, it is categorical impossibility. And without a role — opener, anchor, finisher, seamer, spinner, all-rounder, keeper — no benchmark can even be selected, because every role carries a different one.

Team: no national side, no franchise. Tier positioning — elite power, mid-tier, emerging force, associate — cannot be assigned. And the technical point worth repeating every time: ICC Test, ODI and T20 rankings are three separate tables and must never be merged. Anyone showing a single "team ranking" number is usually choosing whichever of the three fits the argument.

League and commerce: no rights deal, no franchise valuation, no salary figure, no sponsorship, no auction result. So the central discipline of this dimension cannot be applied either — a high IPL price does not equal international strength. Auction price reflects purse elasticity and scarcity; international output reflects age, format and workload. They do not speak the same alphabet.

Rules and governance: no rule change, no umpiring controversy, no DRS incident, no sanction, no eligibility question, no board-level decision. Word discipline is required here. The Big Three revenue-share debate, the DLS formula, slow over-rate precedent, NOC discipline, Cronje 2026, the 2026 spot-fixing case, the 2026 IPL case — these are permitted reference cases in this framework, not allegations. Presenting any of them as a finding against an empty payload would be a serious analytical error.

Risk matrix: sporting, personnel, commercial, rules and integrity, public opinion, systemic — all six categories empty, for one reason. Assessing risk requires a risk to exist.

Public narrative and expectation gap: this is the dimension I care about most, because it is where market expectation and objective assessment are deliberately held apart so the gap between them can be measured. Neither exists here. So the hype-detection heuristics cannot run either: separating small-sample league form from multi-year international output, and measuring the fulfilment rate of "next Tendulkar, next Kohli" framing the way we measure a bowler's economy.

Industry transmission: the map traces a discrete shock through the value chain — youth supply upstream, national teams and leagues midstream, broadcast and derivative markets downstream. With no shock, there is no path to trace. The time-horizon column is also empty, because Time Sensitivity was never assessed.

One thing should be clear after that walk-through: the failure is total, not partial. And that totality is the most valuable information in the entire record.

Diagnosis: total null versus partial null

Had the Stage-1 output been partially null, it would have been a genuinely valuable signal. You could have said the article mentions a team but not a score, a format but not a venue, a player but not a statistic, a result but not a date. Partial nulls tell you about the article — its genre, its likely author, its standard of reporting.

A total null, across every field and every category without exception, tells you almost nothing about the article. It tells you about the pipeline.

So the most probable causes are upstream and mechanical: a source-fetch failure, an empty or paywalled or blocked article, a parsing exception, a truncated Stage-1 run. That conclusion belongs at Medium confidence, not High — because total null leans towards a pipeline fault, but it is equally consistent with an article that genuinely contained no factual assertion: a photo caption, a scorecard embed, a video clip, a live-blog stub. "Stage 1 broke" and "there was nothing there" are both live, and the available data cannot separate them.

One further possibility deserves Low confidence: if the source sits behind a paywall or renders via JavaScript, the domain label may have been inferred from metadata or URL while the body returned empty. In that case the label looks correct and the evidence is missing — the most dangerous combination, because a populated label on a dashboard attracts no questions.

Three case files: 2026, 2026, 2026

In 2026, when I started Purse & Probability from a flat in Dwarka, the first issue took apart that season's IPL final. Mumbai Indians made 129/8; Rising Pune Supergiant were stopped at 128/6; the match went by one run. The scorecard says Pune lost by one run. I pulled the ball-by-ball log — 612 balls — and showed the match was lost between overs 7 and 10: 24 dot balls, and a required rate that never crossed eight across that block. The defeat was visible in a place the scorecard said nothing about. The newsletter reached 3,800 subscribers at a 41% open rate, and one reply came from an IPL team analyst. After that I stopped accepting press-box storylines I could not re-derive from data.

In 2026, when the season moved to the UAE and there were eight weeks without live cricket, I coded every IPL home match from 2026 to 2026 — 316 games. Home sides won 57.4% of them. Across the three UAE venues the same figure fell to 50.6%. I then modelled Sharjah's short square boundaries and flagged it as a 200-run venue before a ball was bowled; on 18 October 2026 Kings XI and Rajasthan tied at 226 in the IPL's first double super over. Three empty stadiums later, the crowd had become a variable, not a mood. No failure there — the model worked, and it worked because 316 matches had been audited, not because three matches had been felt.

2026 brought the opposite failure. February, Bengaluru, the two-day mega auction. My model put a ₹4.6 crore ceiling on Shardul Thakur. He went for ₹10.75 crore. I said nothing in the room — a silence I have not repeated. He finished the season with 15 wickets at 9.79 an over and 120 runs; the side finished fifth, two points out of the playoffs. I spent six weeks writing The ₹10.75 Crore Question, listing every bid we lost and every rupee we overpaid, with the model's error bars shown honestly.

Every transfer is a probability wearing a price tag.

But note the distinction. 2026 was overreach: a number was produced and the number missed. This Monday, there was no number at all. The profession does not grade those two failures equally. Overreach gets panels, think-pieces and conference invitations. Null-handling gets nothing, because null-handling failure is invisible — it usually leaves the building as a filled-in template, and a filled-in template passes review more easily than an empty one.

The fabrication pressure

Empty templates create pressure. An experienced analyst looking at eight blank dimensions will produce plausible cricket commentary automatically: IPL economics, DLS grievances, the ageing-core argument, the changing face of T20, an appeal to blood the youth. Some of it will even be true. None of it came from the article under analysis.

Publishing unsourced cricket commentary under an article-specific heading is a credibility and potentially legal exposure, not an opinion.

The risk is sharpened by the fact that this business runs on trust. Three franchises send me their retention sheets before the cut-off. They do not send them because of my face. They send them because they know that if I do not have an answer I will write "no answer", and that is what they find useful. Publish one confident paragraph off an empty payload and the next retention sheet goes to somebody else. That currency is spent once.

Downstream contamination, and the gate

The risk does not stay on one page. A null Stage-1 record that passes unpoliced into an aggregated feed, a sentiment index, an editorial triage queue or a trading dashboard corrupts every downstream model that treats record-presence as content-presence. That is a dataset error rather than an argument error, and dataset errors are the most expensive kind because they have already spread everywhere.

The fix is cheap. A hard validation gate before Stage 2 is invoked: Information Points length of at least one, Article Title non-null, Entities non-empty. Quarantine the records that fail. The correct output for an empty input is no output.

Two smaller things no dashboard shows. Persist the source URL and fetch timestamp at Stage 0 — this record contains no source field at all, so reconstructing the original content will depend entirely on pipeline logs. And stop treating the domain label as evidence. cricket_world is a category. Inside cricket, the distance between Test and T20 is as wide as the distance between cricket and another sport.

Which is where cross-sport benchmarking also demands care. In the summer of 2026 I watched Russia 2026 for the accounting rather than the football — CIES minutes models, Transfermarkt's valuation bands, the logic that prices a 24-year-old differently from a 29-year-old. I rebuilt that logic for cricket: replacement-level WAR, age curves, purse elasticity, an injury discount. The 22-page model went to Delhi Capitals' analytics head in October 2026, and by the 2026 auction I was Auction Data Lead. We traded for Shikhar Dhawan and bought Colin Ingram inside our ceiling; the franchise reached the playoffs for the first time since 2026.

I benchmarked the model against Russia 2026, and it flinched — and the flinch was the useful part, because it showed which pieces of a football model do not transfer. But that whole argument only runs when there is data to put into the model. Cross-sport benchmarking on an empty array is just better-read fabrication.

The layer beneath the framework deserves the same scrutiny. The crowd is a variable, the empty stand is a measurable input, Ranji's invisible audience is a datapoint, and the WPL's broadcast promises against broadcast delivery is an auditable file. That is the commercial and demographic substrate that determines what the sport can actually sustain. In this payload there was not one figure on any of it — no attendance, no ticket, no jersey, no follower, no endorsement. When the bottom layer is missing, every story above it is a guess.

Where I have to break my own argument

Cricket analytics has developed a good habit over the past decade: it is afraid of model overreach. Confidence intervals in every deck, sample sizes attached to every figure, the "this is not a guarantee" line after every projection. That is healthy and should stay. But that fear has thrown a second, far more common failure into shadow — the one where the analyst has nothing and delivers something anyway.

The industry punishes "I don't know" least and punishes an empty space most. That order should be reversed.

An analyst is praised for insight. Nobody is ever praised for saying, on seeing eleven blank fields, that they will not file the deliverable because the input did not arrive. Yet that is the act that makes the whole system trustworthy. An analyst who files on empty input also devalues every report he files on full input, because the reader now has no way to tell which report came from data and which from a template.

The second point is subtler. Cricket has always struggled with reading what is visible as what is true. Dropped catches are counted; misfields are not. Wickets are recorded; good leaves are not. Dot balls are logged; balls left on length are not. Analytics has inherited the habit — we measure the things that happened, not the things that did not. An empty array is not an event, so it never reaches a scorecard. The dangerous form of an empty array is not that it says nothing. It is that it looks like a clean bill of health.

And here I have to doubt my own finding. I argued that total null points to pipeline fault. But I derived that from the very data whose absence is the subject. I am inferring a presence from an absence I never observed. That is a circularity risk, and it must be published with a confidence level attached, or I commit the exact error I am describing. Hence Medium, not High. And hence not one sentence in this piece claims what the original article said — because the original article was never in front of me.

Signal for the next morning

One working rule, applied to every piece I have written since 2026 and now functioning like a pre-auction ceiling: pre-register the position before the data arrives, and when the data does not arrive, still publish a number — zero.

So the next time a dashboard shows eleven fields out of eleven empty, the question should not be "what is the story". It should be "what is the length of the Information Points array". And if the answer is zero, the most honest headline is the one nobody wants to print: this article did not arrive.

The crowd changes the game. So does the disappearance of the data — except that one is filed as an audit, not a post-mortem.