The White Space on the Spreadsheet: How Sports Analytics Learned to Say “Insufficient Information”

The White Space on the Spreadsheet: How Sports Analytics Learned to Say “Insufficient Information”

I want to open with a night I still cannot fully let go of. It was 2:47 a.m. on 14 March 2026, and I was sitting in front of a spreadsheet with a 14.2 in the xG column and an eight in the goals column. Jamie Maclaren, then twenty-three, had just finished twenty-three rounds of an A-League season. My model said he should have scored six more. I wrote an aggressive piece about it. My editor struck out nearly every number and told me nobody would understand xG, and that if they did, they would not believe me.

So I spent a month re-watching nineteen match tapes and re-classifying every shot by hand. What I found was not a cleaner number. It was a bigger blank space. Several chances simply could not be decided either way, and I had to write “unclear” in the notes column. That was the beginning of a professional habit I still keep: the three characters N/A.

Context: more data, more blanks

In 2026 I am thirty-nine, based in Brisbane, working as a sports data analyst and covering esports for the Australian market. Compared with 2026 the volume of data has exploded. Football has positional tracking. Basketball has tracking data going back more than a decade. Esports has public APIs for schedules, results, picks and bans, gold and damage. There has never been more raw material.

Yet more data does not erase blank spaces. It makes them sharper. With ten metrics on a single play, you discover that all ten are silent on the most important question: what the player was thinking in that moment.

So I built a nine-part check for every deep analysis: patch and meta, tournament format, roster, regional landscape, club finance, rules and governance, risk profile, public narrative, and industry transmission. The point is not the list. The point is that for each slot I must ask whether I actually have enough evidence to conclude anything. Most of the time, the honest answer is no.

The nine doors and the nine walls

Patch and meta. A publisher like Riot Games ships League of Legends updates roughly every two weeks. The patch notes describe intent. The live server describes results. Those are not the same thing. A champion’s win rate is the most misunderstood metric in the industry because without pick-and-ban context it is just a number hanging in the air. To judge how a patch fits a specific team I need at least three things: how many games that team has played on the new patch, who they played, and how they won or lost. For most teams in less-broadcast regions, I have none of the three.

Tournament format. Format decides which data exists. A best-of-three knockout produces a maximum of five games. A Swiss stage, adopted by several major international events from 2026, pairs teams by record, which means strong teams meet earlier. A team can advance 3-2 with wins against opponents who had nothing left to play for. If I average everything into one figure, I have invented a story.

Roster. Paper strength is easy to add up and easy to get wrong. In esports the problem is worse because scrim data is not public. The single most useful indicator of chemistry — performance under controlled high pressure — sits outside public view. Official match data tells me what a team did, not what they practised.

Regional landscape. A region may receive only one or two international slots a year. That means an entire professional ecosystem of hundreds of players gets judged on two series. I compare four indicators: international results, player-pool depth, academy output, and domestic ecosystem health. Only the first is public. For markets like Vietnam, the most defensible claim I can make is that the supply of talent is better than the international results suggest. Everything else I leave blank.

Finance. Transfer figures are the most distorted numbers in the sport. An esports club’s finances run on four lines: sponsorship, publisher or league distributions, salaries, and owner capital. Usually only sponsorship is ever disclosed. Most big deals are announced as undisclosed fees, and the number that circulates comes from the agent’s side. I treat agents as the largest hidden cost in the transfer system. They are not wrong to defend their clients, but the noise they generate distorts reference prices for the whole market.

Governance. Match integrity, transfer and registration rules, contract enforcement, minor protection, publisher disputes. Internal investigations are not published, and disciplinary rulings usually state conclusions without process. What I can track are behavioural patterns: how often similar incidents occur in one region, how consistently they are punished, how long resolution takes.

Risk. A full risk profile needs six categories: competitive, financial, personnel, rules, public opinion, systemic. When data is thin, insiders underrate risk and outsiders overrate it. I label every line with a confidence level, and where confidence is low but impact is high, I flag it as an assumption rather than a conclusion.

Narrative. Public stories run in heat cycles. Three straight wins and the story gets hot. My job is not to kill the story but to test its foundation: is there real substance, is the sample large enough, how long will it last. When Italy went thirty-four matches unbeaten under Roberto Mancini, their average PPDA was around 9.8 — aggressively high pressing, since lower is more aggressive. But match by match the variance was huge. The average hid two different teams inside one season.

Industry transmission. The chain runs from publishers upstream, through clubs and broadcasters midstream, to sponsorship and mainstreaming downstream. Each stage moves at a different speed. A rules change hits teams in weeks and sponsorship markets in quarters.

The contrarian angle: the industry pays for confidence, not accuracy

People pay for conclusions, not for uncertainty. A piece that says “this team will win it all” travels. A piece that says “I need two more rounds before I can call this a trend” does not. So writers learn to fill blanks with confidence, and an empty N/A becomes a sentence that sounds certain. I call this the ritual of the spreadsheet: a table with headers and formatting creates an impression of rigour regardless of what is inside it.

This connects to a problem I have followed for years: correlation is not causation. Inverted wingers have homogenised attacking football. Every club wants a left-footed player on the right to shoot inside. That has wrongly erased the traditional winger who holds width and crosses from the touchline. The mistake is choosing players on statistical correlation rather than tactical mechanism. A team defending in a low block and strong at second balls gains more from a winger who stretches the pitch than from one who drifts inside and crowds the central midfielder.

I borrowed a comparison from another sport to describe it. Watching sport climbing at the Tokyo Olympics, I was obsessed with Janja Garnbret and the way she would stop on a wall with no visible hold, waiting for the exact moment to shift her weight. That is exactly how a central midfielder receives under pressure and turns inside one square metre. I call it an anchorage point. A good anchorage buys you half a second, and half a second is enough to open a line-breaking pass. No passing statistic contains that half second.

Every number carries a story, and my job is not to ruin it. When I broke down France against Argentina in the 2026 World Cup round of sixteen, I measured Kylian Mbappe’s top speed on the decisive assist at 37.6 km/h. Mbappe’s feet always tell the truth, but I still need the number to translate them. After two sleepless nights of frame-by-frame work, I had to admit no pressing metric and no xG value explained the raw beauty of that acceleration past three defenders. Data measures what. It does not measure why we love the game.

In 2026, when the pandemic froze every league, I lost contracts with two broadcasters and had no new data to process. One night I reopened Liverpool 4-0 Barcelona and built a table of Andrew Robertson’s running: 12.4 km total, 2.1 km of it sprinting. I wrote a long blog about missing the noise of the stadium. By morning it had been shared more than four thousand times. The empty summer taught me that with no matches at all, memory still shoots from distance. And the long shot in memory always finds the top corner, while on the spreadsheet it flies straight at the keeper.

At thirty-nine I have learned that data hurts when it is twisted. Every time someone attaches a number to a conclusion that the number cannot support, the gap between analysis and performance becomes visible.

Takeaway: the signal in the next cycle

Why keep working through nine categories when most of them end in N/A? Because the value of a framework is not that it answers everything. It is that it shows exactly which questions have no answer yet. When I say I cannot assess a patch because I have no roster data, I have delivered real information: that information does not exist, and anyone claiming otherwise is guessing.

In an industry where everyone is trying to say more, the ability to say less may become a competitive advantage. The earliest signal will appear at the data layer, not the opinion layer. When a club publishes an internal report that labels confidence levels for every metric, that is a signal. When a region starts recording academy numbers instead of counting international slots, that is a signal.

When the spreadsheet speaks, the stadium must learn to be quiet. But before the spreadsheet can speak, the person building it must learn when to stay silent — and to document exactly why.