Nine Fields, Eight Blank: The Six Variables That Actually Carry a Cricket Analysis

Nine fields. Eight came back blank. The one that was filled contained two words — cricket_asia. In May 2026, sitting ten kilometres from the MCA ground in Pune, I logged Rising Pune Supergiant's innings not in a scorebook but on a slip of paper. 128 for 8 at Hyderabad, against Mumbai Indians' 129 for 8 — a one-run defeat, and the last match of that city's short-lived franchise. Four weeks later at The Oval, Mohammad Amir's 3 for 16, Rohit Sharma out for 0, Virat Kohli for 5. Across that whole summer I re-plotted 43 overs of field placements from broadcast replays into a school notebook — rings, gaps, release angles. Not one field in that notebook stayed empty, because an empty field meant I had not watched, and I had watched. Five years later a sheet came back from a data desk with nine rows and eight N/A values. Title: N/A. Source: N/A. One-sentence summary: blank. Information points: empty array. Entities: empty array. My first reaction was that the sheet was useless and should be bounced back. My second reaction was more honest: it was the most useful document to reach my desk that month, because it had drawn up the list every cricket analysis needs — the list of load-bearing variables. The spreadsheet was not the answer; it was the first question with cells. This piece is a map of that empty sheet. I keep a field map for the moments the scoreboard forgets, and here the scoreboard itself is missing. So the map is of the machinery behind the scoreboard — the frame without which any cricket number is just a number. One filled cell is more dangerous than eight blank ones Blank fields are honest. They say: we do not know. The ninth field was not honest. It carried a domain label, and a domain label is the most innocent-looking thing on the sheet and the most damaging, because it licences inference. Watch the chain build. The label says cricket_asia. Step one: this concerns Asia. Step two: Asia means spin-friendly surfaces. Step three: it will be a night match, dew will arrive, bowling second will be harder. Step four: therefore the side that chose to bat first was right. Four steps, zero evidence. And if step four reaches a headline, nobody asks who took step one. From nine years of watching this game closely, one thing is clear to me: the biggest error in cricket analysis does not come from too little data. It comes from filled data. A thin dataset stops you. A filled one pushes you forward, and moving forward is the danger. That is why this report is itself a result, even if it is not a cricket result. It produced the list. The list has six items, and each one comes with a case file — because in cricket a principle only counts when there is an over, a ball and a decision standing behind it. Format is the first wall every number hits Elite cricket has four principal forms: Test, ODI, T20 and The Hundred. Carrying data across them is not merely wrong, it is dishonest, because the number changes meaning. A strike rate of 145 is elite in T20 — match-winning rather than match-saving. The same 145 is above average in ODI, where risk is priced across fifty overs differently. In Test cricket, 145 usually describes an event rather than a standard, because the real currency there is the leave. The same applies to bowling. An economy of 8.5 is acceptable in ODI middle overs, poor in T20, and meaningless in Test cricket, where the gap between wickets matters more than the rate. I have a live example of this error, and it is mine. December 2026, campus shut, home in Pune, nineteen years old. India's second innings at Adelaide collapsed to 36 all out in 21.2 overs — Josh Hazlewood 5 for 8, Pat Cummins 4 for 21. That night I did not react. I logged every dismissal across three columns: delivery type, length, and the batter's trigger movement. Pink ball, dusk session, all of it on a separate sheet. That autumn the empty-stadium IPL ran in the UAE, and I logged the commentary too, to see what broadcasters fill crowd silence with. The answer was almost always narrative, rarely technique. Two things went to the top of my sheet that year: commentary is a secondary source, stump mic and wagon wheel are primary. And no dismissal is ever explained by a mood. Every wicket has a ball, a length and a decision. January 2026, Brisbane. India 329 for 7, Rishabh Pant 89 not out, a series won by a side nobody had modelled. If I paste the Adelaide sheet onto the Brisbane sheet without a format note, both become worthless. That is the first rule of format: numbers do not travel. Match nature: the same fifty, a different price Format decides how a number is read. Match nature decides whether the number can be trusted at all. A bilateral series game, an ICC knockout, a franchise league fixture and a warm-up do not carry the same pressure, so they do not produce the same sample. A T20 league fifty and a World Cup knockout fifty are not the same statistical unit, even though both say 50. The one-run final of May 2026 is the cleanest case. Hearing "one run", people conclude it was close, therefore it was about nerve. My notebook does not have that story. It has overs. The over where the required rate crossed the wicket-risk line is over 17. What the fielding side posted on the boundary after that was not a gamble; it was arithmetic. They took the risk and got the return, because the batting side had no power left in those overs. Had the same innings happened in a league's 14th match, the analysis of the defeat would be entirely different. In a knockout, one wrong over cannot be repaid. In a league, it can be repaid in three days. That difference — what a defeat costs — is set by match nature, not by form. So when the intake sheet returned blank, including the field for match nature, I stopped the analysis there. Without nature, no cricket conclusion is defensible. Pitch and venue: the largest hidden variable If I could keep only one variable, I would drop runs, wickets and strike rate and keep the pitch. In international cricket, home-ground and pitch effects are roughly the dominant confounder in any result. Adelaide settles it. Pink ball, dusk session, seam movement after the ball softens. In a 36-run innings every dismissal has a different length but the same schedule. Hazlewood takes 5 for 8, Cummins 4 for 21. That story is not about who lacked courage. It is about which over, at which length, the ball began to move. I wrote that innings in ball numbers — ball 14, ball 31, ball 47 — and that is my real note. The fielding side was not guessing either; it was changing its length because it knew the surface. The second half of venue is character, not just surface: grass cover, dryness, crumbling, wind, cloud, humidity, and DLS if rain intervenes. Miss one of those and your conclusion is half a lie. The intake sheet had no venue field at all. That does not mean the pitch was neutral. It means we do not know. This is where I have to police my own reflex — "it is Asia, so it will spin". That reflex is true less often than it looks. Several Asian grounds are now paced-bowling strongholds, and night dew can neutralise spin from the first innings. A metric lies when you refuse to write down its conditions. Name, role, then metric When a player's name appears, my first question is not a metric. It is a role. Opener, anchor, finisher, pace, spin, all-rounder, keeper — only once that is settled does it become clear which number matters. Thirty runs at a strike rate of 160 is a finisher succeeding; it is also an opener failing, if those thirty runs took thirty balls. February 2026, Under-19 World Cup, New Zealand. India beat Australia by eight wickets at Mount Maunganui and eighteen-year-old Shubman Gill finished with 372 runs and the Player of the Tournament award. I logged his scoring zones by hand across seven matches, then rebuilt the sheet in Excel after starting a BSc in Sports Science at Savitribai Phule Pune University that July. The sheet showed something I had suspected: Gill's cover driving was governed by where the ball pitched, not by where the field stood. That was the first time I had a variable I could defend. One cover drive, one length, one boundary — and beyond that boundary the metric lies, because on a slower ball the same drive becomes a mishit. Two further things attach to a name and cannot be checked without one: the age curve and injury history. A fast bowler's pace traces a curve between twenty and thirty, and ignoring it turns a three-year-old number into today's truth. Date: an old number is a museum piece Cricket's data landscape moves monthly — ICC rankings, auction cycles, central contract lists, injury bulletins, NOCs for overseas leagues. Any of it can be unusable within six months. Rankings need particular care. The ICC maintains three separate tables — Test, ODI and T20I. A team's position in one says nothing about another. If an article says "ranking" without naming the table, the sentence is half-finished. On auctions, the mega- versus mini-cycle changes what a price means. A mega-auction price reflects franchise squad-slot economics, not international strength. Retention lists, RTM and purse allocation are resource-allocation problems, and I read them as such. India's real selection crisis is not a talent shortage; it is scarcity pricing — too many credible roles chasing too few slots. Without a date, that entire layer collapses. The intake sheet had time sensitivity explicitly marked as unassessed. That is not a small gap; it is an axe at the root. Source: the field that decides whether you are reading or hearing An ICC media release, an ESPNcricinfo staff report and a traffic-driven social account are routinely treated as equivalent, and that is the most common failure in cricket journalism. They carry different evidential weight and different accountability. The source field returned N/A. That does not mean the source was bad. It means we do not even know whether there was one. That is the most dangerous state, because blank fields are usually waved through as normal. One point needs stating plainly: the absence of an integrity signal is not evidence of integrity. No anti-corruption news does not mean a clean match. It means we do not know. The same rule applies to rankings, to form, and to team environment. The contrarian angle: an empty sheet is honest, a filled one is suspect Here I have to argue against my own instinct. An analyst's reflex is to demand more data. This report's finding runs the other way. The sheet held no data and still turned out useful, because it asked the question a filled sheet never asks: which field was load-bearing? The real failure mode in cricket analysis is not thin data. It is the chain of inference that starts from a filled cell and never gets stopped. One domain label becomes, four steps later, a defence of a captain's decision. Nobody demands evidence at any step, because each step looks reasonable next to the one before it. So my rule is this: name the metric's failure mode in one line before deploying it. Strike rate lies when compared across formats. Economy lies when the phase of innings is missing. A pitch report lies when it is written from first-innings experience and applied to the second. The sheet also exposed taxonomy drift: it said cricket_asia where the canonical label is Cricket. Small, but revealing — a routing label was fusing geography with sport. Geography is routing metadata, not content. It is a postcode, not an address. And the third point, the one that stings: if this blankness happened once, fine. If it repeats twice in the next five reports, the fault is not in the source. It is in the system. What to watch: a dated, checkable condition My forecasts stay inside six months, because that is as far as a model in this sport can honestly reach. So here is a dated, checkable condition. Before any cricket analysis proceeds, it should pass a minimum-viability gate: at least five atomic information points, at least one named entity — player, team or organiser — a confirmed format, and a dated source. If any one is missing, hold the conclusion. Holding is not a failure of analysis; it is a condition of analysis. Over the next five reports I will count three things: how many fields came back populated, how many entities were named, and whether the source carried a date. If two consecutive empty payloads arrive from the same source, I will not file it as editorial sparsity. I will file it as a fetch or parse defect, and it goes to the data team. The question is not which match was hiding inside that empty sheet. The question is whether, next time a sheet comes back full, we will ask what the format was — or jump straight to the verdict because this time the cells were filled.