Beautiful Report, Empty Data: How to Read a Football Analytics Sheet in V.League

On a July evening in 2026, at a coffee shop on Nguyen Thi Minh Khai Street in Saigon, a data analyst from a V.League club slid a twenty-two page document across the table towards me. The cover was printed in colour. Bar charts, shot-zone heat maps, formation diagrams, a complete source list. I turned the pages, and on page nine I stopped. Every data cell was empty. Not one metric, not one player name, not one value. All that remained were the template strings the software generates by itself, along the lines of “identify from the information points above”. He was not deceiving anyone. He had simply uploaded an empty file, and nobody on the coaching staff had reopened it before it was sent on. I retell this because it is far from rare. Vietnamese football has entered a phase in which a good-looking data report can pass through three layers of review without anyone touching its interior. The problem is not that the numbers are wrong. The problem is that the numbers do not exist, yet the form of the report is persuasive enough on its own. Since 2026, each V.League season has seen a few more clubs hire analysts. Broadcasters have put expected goals on live graphics. Viewers have grown used to abbreviations that previously appeared only in the internal documents of European academies. The new habit formed quickly, but the discipline that should accompany it has been far slower. The first xG table I ever kept was handwritten on a bus, back when nobody called it data. I recorded every shot attempt by fourteen clubs in a notebook, defined the shot zones myself, counted the defenders in front of the ball myself, and scaled the values by distance and angle myself. That method was slow and manual, but it taught me something modern software does not: if the cell is empty, no conclusion may be produced. The standard pipeline of a football analysis report has four stages: collection, cleaning, modelling, interpretation. Errors in the first and second stages usually show up immediately, because the tables go blank. The most dangerous error sits in the third stage, when the model still runs, still exports a file, still formats correctly, but has nothing inside to compute. Engineers call it a silent failure. In football, silent failure has another name: trust placed in the wrong place. The nine checkpoints below are the filter I use when reading any analytics report, whether from a club, a broadcaster, or a colleague in the trade. The first checkpoint is tactics and technique. A decent report must state which system the team plays, whether the block sits high or deep, and, more importantly, must offer at least one process metric. Expected goals per match is a process metric. Passes allowed per defensive action is a process metric. Pass completion in the final third is a process metric. If a report contains only goals and assists, it is a results sheet, not an analysis. Viewers watch the move; I watch twenty-two numbers moving — and wait patiently for them to tell a different story. In 2026, when Phan Van Duc was twenty years old and had scored only five V.League goals for Song Lam Nghe An, his expected goals per match reached 0.48, above the average for foreign strikers in the same league. That figure was not mentioned in a single bulletin all season. I wrote that he would become a fixture of the national team within three years. Many said I believed in spreadsheets to the point of delusion. In December 2026 he scored the decisive goal at the AFF Cup. Tactics requires one further layer: comparison with the broader trend. The world looked at Croatia and saw an underdog; I looked at them and saw a series of coefficients nobody had dared to mine. At the 2026 World Cup, Croatia under Zlatko Dalic pressed with a PPDA of 7.9 against Argentina, lower than even Spain, the side famed for ball control, in the same period. A metric only means something when placed beside the league-wide benchmark. A report without a comparison baseline is an unfinished report. Here I must be explicit about sample size and confidence intervals, because this is the point Vietnamese football writing skips most often. Three matches is the sample size of one working week. Ten matches is the threshold at which process-metric trends begin to mean something. One season is the minimum bar for discussing style. Any conclusion that exceeds its own sample size is inference, and inference in football always finds supporters because it sounds reasonable. The second checkpoint is club finance and the transfer market. This is the weakest area in domestic reporting, because most clubs do not publish their revenue structure. But a lack of disclosure does not mean a lack of traces. A wage bill leaves traces in squad depth. A transfer fee leaves traces in the payment terms. An instalment deal spread over three years says more about cash flow than the headline figure printed in the papers. The transfer market is a game for those who look far, not those who look often — value always arrives after patience. Small V.League clubs are being pulled into a familiar pattern: taking young players from bigger clubs on loan, with an obligation to buy triggered once the player reaches a certain number of appearances. At first glance it looks like an opportunity. On closer inspection it is a future liability recorded in the legs of a footballer. Small clubs pay the wages, provide the playing time, cover the medical and recovery costs, and then must buy back the very product they developed, at a price the selling side set in advance. A three-year financial plan is locked by a clause that cannot be renegotiated. Commercial benefit flows to the big club, risk flows to the small one, and the loop repeats every transfer window. The third checkpoint is results and the public-opinion cycle. A three-match winning run has never been evidence of long-term form; it is evidence of three wins. When a newspaper writes that a team has found itself again after two rounds, the newspaper is describing a feeling, not data. Public pressure, by contrast, is entirely measurable. It leaves traces in the number of personnel questions at press conferences, in a coaching staff suddenly changing the language of its answers, in a club closing a training session without notice. I once tracked a V.League head coach across eleven rounds and found that after every defeat, the number of personnel questions doubled. That is data; it is simply data nobody has bothered to count. The fourth checkpoint is the league landscape and the team's positioning. An analysis is only valid when you know which tier the club occupies: title race, continental qualification, mid-table, or survival. The same pressing metric carries opposite meanings at two different tiers. A relegation battler pressing high is committing physical suicide; a title contender pressing low is capping its own ceiling. The fifth checkpoint is rules and compliance. Here I must say plainly what the commentary trade rarely admits: VAR does not make controversy disappear, it moves controversy off the pitch and into the review room and the grey areas of the law. A disallowed goal still causes an argument; the difference is that the argument now happens in front of a monitor with the rulebook in hand. A rise in the number of overturned decisions does not mean the number of errors has fallen. It means the intervention threshold has been lowered, and every lowering opens a new zone of dispute. In Vietnamese competitions, the cost of running a review room and the cost of training referees have never been placed in the same ledger as the broadcasting benefit. The sixth checkpoint is management and the dressing room. In 2026, when major leagues were suspended by the pandemic, I spent six months digging back through V.League data from 2026 to 2026. In 2026 the stands were empty, but every pass still fell into a cell of the model, and I understood that data never keeps company with a pandemic. What I found was a simple rule: clubs that changed chairman mid-season saw their win rate fall by 23 per cent over the following five matches. The cause was not technical. The cause was that every mid-level personnel decision was frozen awaiting a new signature: contract extensions, assistant hires, recovery medication, flights for away fixtures. Six months after the retrospective series was published, a club executive called to thank me, saying the table had helped him postpone sacking his head coach at the most sensitive point of the season. My model does not cry and does not celebrate, but after every match it owes me a lesson. The seventh checkpoint is the risk register. An honest risk table must distinguish two very different states: a risk that does not exist, and a risk that has not been assessed. In analysis documents these two states are usually written identically, with the same symbol. That is the most serious error a reader can make, because it converts ignorance into reassurance. Here I place on the table a technical issue Vietnamese football will meet repeatedly in the coming seasons: anterior cruciate ligament injury. Rushing a player back after ACL surgery is damaging the second phase of many careers. Fear in the head is harder to repair than a ligament in the knee. A medical report that says only “fit to play” is an incomplete report. It needs the number of contact-training days, the maximum number of direction changes in a session, and a plan for gradually increasing minutes week by week. The eighth checkpoint is media and expectation. The transfer window is the season of noise. The skill required there is not reading more rumours, but ranking sources by weight of evidence. Three simple tiers: tier one is information from the club itself or an agent with a signed mandate; tier two is information from an intermediary with a direct interest in the deal; tier three is information that cannot be traced to any source at all. Most of the transfer rumours Vietnamese fans read daily sit in tier three. The strongest signal-jammer in a transfer window is a fee quoted without its structure. One hundred billion dong paid outright in a single season is entirely different from one hundred billion dong spread over four years with appearance bonuses and a sell-on percentage. Small clubs bear the reverse impact under every such structure. The ninth checkpoint, and the most overlooked, is transmission through the industry. A decision in an academy reaches the first team three to five years later. A decision in the first team reaches the broadcasting contract one to two seasons later. A decision on broadcasting reaches the transfer value of the whole league a few years after that. This chain is long, slow, and because it is slow nobody wants to draw it. Yet this is precisely where a data person creates the greatest difference, because the edge lies not in the speed of reading news but in the time spent holding a model. Correlation is not causation, and this is the trap that kills more football analysis than anything else. The nine checkpoints above are not a formula for reaching conclusions. They are a filter for discarding cheap conclusions, the ones built from three matches, three bulletins and three rumours. The most common mistake is treating empty data as neutral data. When a table has nothing in it, readers tend to fill it with their own prior judgement. That is the moment prejudice is legitimised by the form of a report. In the worst case, people read the line “no risks flagged” and understand it as “no risks exist”. Those two sentences are worlds apart, and the distance between them is an entire player's career, or an entire small club's budget. I have also set myself one principle to counter my own memory of success: new data always has the right to defeat old data. Having once been right about Croatia in 2026 does not grant me the right to be right again. Every season is a new sample, and a new sample has the right to overrule an old model. I do not trust coaches; I trust the model. But I listen to coaches in order to fix the model. A spreadsheet plus half an hour of conversation with an assistant analyst is often worth more than ten days of running algorithms, because the model cannot measure what happens on a training morning when the captain has just lost a family member. The next round of fixtures will not answer who wins the title. It will return only a few small signals: which club starts publishing recovery days, which team dares to give a nineteen-year-old a full ninety minutes, which coaching staff cancels a press conference without explanation. Those traces, added together, form a more honest picture than any prediction. What I will do in the next cycle is print the sample size on the first line of every analysis, so readers know whether they are reading a season, a month, or three matches. A model is only trustworthy when people know how much data it stands on. And if an analytics sheet has all nine sections and every section says “insufficient information”, read it as a warning, not as a safe conclusion.