Lessons from an Empty Spreadsheet: The Discipline of Saying 'Insufficient Data' in Cricket Analysis

In a small conference room at a Kuala Lumpur cricket academy, the monsoon rain was falling outside. On the table lay an open spreadsheet and a cup of tea going cold beside it. The column headings were weighty — 'average', 'strike rate', 'economy rate', 'recent trend' — but almost every cell was empty. One read 'N/A', another 'insufficient data'. The coach sitting opposite looked up and asked, 'So what is your view on the boy?'

That moment is the hardest test of my professional life. The easiest answer is to invent a story — one viral clip, two handsome cover drives, and a catchy prediction. But I genuinely had little. I had watched one match in full, gathered scorecards from three innings, and seen a single net session. Nothing more. The oldest temptation in analysis is to fill empty cells with guesses, guesses, and more guesses.

I took the cup in hand and said, 'With what we have now, I cannot tell you his future. And whatever I said would be a betrayal of his career.' The coach was surprised at first, then he smiled. That small incident is the heart of this piece — the risk of building analysis on empty data, and the discipline that stands against it.

Cricket is now a flood of numbers. A hit-map for every ball, a wagon wheel for every innings, a release point for every bowler, line-length zones, and a metric whose name begins with 'expected'. From professional leagues down to under-16 regional tournaments, cameras are being installed and data trackers are being set up. Academies in Kuala Lumpur, Dhaka, Colombo and Lahore now place a spreadsheet on the first page of their scouting reports.

From my years of watching matches, I can say this spread of data is genuinely a blessing. But when that data is thin, incomplete, or belongs to a single format, a trap opens before analysts. Some see an empty cell and write a guess; some predict a whole career from a two-match sample; some mix a Test average with a T20 strike rate and build a neat story.

My own working step is simple — first break raw information into small 'information points', then stand the analysis on those points. An information point means a fact with evidence behind it: a date, a score, a match context. If there are no information points, the analysis has no foundation. Build a house on zero and the first storm brings it down.

On my desk wall in Kuala Lumpur hangs an old note: 'Evidence first, story later.' It is not a newspaper slogan but a warning I wrote to myself. Since watching the Malaysia under-22 final at Bukit Jalil in 2026, I have tried to live by it. That day Safawi Rasid took five shots and completed seven dribbles — I put those numbers on Twitter, but before them I described the noise of the stadium, the disappointment of the stands, and the smell of wet grass. That sequence is still my method: scene first, numbers later.

The shortage of data first runs into the question of format. In cricket, Test, ODI and T20 are three different games with three different logics. In Tests, the arithmetic of patience and bowling load differs; in T20, the freedom to take risk differs. If you read a young batter's Test average and then write his T20 future, that is not analysis but the misuse of numbers. Any conclusion that mixes formats is eventually proven wrong, because the benchmark does not transfer from one format to another.

The second trap is the small sample. From three innings' scorecards you can write a batter's 'recent trend', but that is not a trend, it is coincidence. Below twenty innings any average wobbles — a six in one innings, a duck the next. I have seen it myself: a young leg-spinner took nine wickets in his first four matches and earned the tag 'the next big star'; over the following ten matches his economy jumped into the nines. What did the analysts do? They deleted the old prediction and wrote a new story. A small sample can never be the foundation of a prediction; it can only be the thread of an enquiry.

The third trap is home-ground effect. In home conditions, averages, strike rates and spinners' success all inflate. On away tours those numbers collapse. In 2026 I analysed nine Bundesliga matches played behind closed doors — a football context, but the lesson holds in cricket. In those nine matches, the home-win percentage fell from 43.3 to 21.2 percent. That is, the thing we call 'home advantage', which we treat as a constant, itself changes when the environment changes. In cricket, when crowds return, when the pitch changes, when dew falls — the same player's numbers are not the same. Home data often hides weakness, so no assessment is complete without away performance.

The fourth trap is luck. The toss, dew, Duckworth-Lewis, and DRS. Without separating how much of a result is skill and how much is luck, analysis becomes misleading. In T20, the advantage of a team batting second after winning the toss often comes from dew, not from bowling skill. A DRS decision can also turn a match. I never jump straight to a conclusion from a result; first I ask who won the toss, when the dew fell, and which review changed the match. Without separating luck, analysis becomes mere storytelling.

The fifth trap is the abuse of metrics beginning with 'expected'. Just as xG is dragged in football to explain every decision, cricket now has metrics like 'expected runs' and 'expected wickets'. These metrics are the output of a particular model's assumptions, not a player's decision. Why a batter played a shot, why a bowler dropped the yorker in the final over — a model cannot say. Expected metrics give only context, not explanation; reading a player's decisions, form and umpiring standards needs eyes, not formulas.

The sixth trap is injury history and workload. If a scouting report lacks a player's minutes, travel, recovery and injury history, it is half a report. In 2026 I tracked Pedri's 66 matches — Barcelona, Spain, the Euros, the Tokyo Olympics. I built a workload model and recommended rest, but the club ignored it. In cricket this arithmetic matters even more, because the same bowler runs all year between franchise leagues, domestic tournaments and the national side. Without workload data, any assessment of a young talent is incomplete, because the body is ultimately his biggest limit.

A good analysis should cover eight dimensions — format and match nature, player technique and data, team landscape and ranking, league and commercial ecosystem, rules and governance, risk, public narrative and expectation, and industry transmission. The foundation of each of these eight is the information point. When information points are empty, the honest answer in every dimension is the same: 'insufficient information, assessment not possible.' That is not failure, that is honesty.

Let us descend into a real context. Say a young Malaysian all-rounder plays a few good innings in T20. Suddenly he becomes 'the next big name' on social feeds. Someone shares video of two of his sixes, someone writes that he will take the country to a World Cup. What is an analyst's job at that moment? I recall the lesson of the Safawi thread — following it from the feed all the way to the stadium. I brush off the dust and look: how many matches, on what pitches, against which bowlers, how many overs has he bowled, how much has he travelled. Most often it turns out the sample is so small that to predict is to gamble.

The reality of associate cricket is harder still. Malaysia, Singapore, Hong Kong, Nepal — full data from these teams' matches is not always available. Some scorecards are online, some are not; video streams exist, or they do not; ball-by-ball data is often simply absent. It is within this emptiness that I must write analysis. And here lies the biggest temptation — to pull together local-language reports, the coach's spoken words and stadium gossip into one story. The story turns out beautiful, but has no foundation.

This is where I work jointly with local journalists. A parachuted analyst often errs — he treats Malaysia as a colourful backdrop and assumes UK cricket norms are universal. I was born in the UK but work in Kuala Lumpur; that distance has taught me that without local voices and local-language sources, analysis is incomplete. Seeing from the margin does not mean seeing wrongly, but seeing from the margin without evidence means imagining.

Silence is data too. Which player did not celebrate a half-century, which young talent withdrew from a tournament, which stand stayed empty — these voids often tell the biggest story. But be careful: there is a difference between 'what has not yet been said' and 'what will not be said'. One player may be hiding an injury, another may have just returned to fitness. Reading silence requires asking, not assuming. Do not step into the easy trap of romanticising silence; the cause behind the silence is the real story.

And in speaking about workload, one must respect a limit. In the name of protecting a player, he must not be belittled. Speak on the foundation of data and the player's own testimony, not by imposing decisions from above. The thin line between protection and paternalism is drawn by listening to the player's voice, not by speaking on his behalf.

Similarly, 'I spotted him first' — this hipster vanity is also to be avoided. Finding young talent does not mean making him one's property. The credit really belongs to the local coach, the family and the academy. The analyst's job is only to document.

Today's reader is drowning in a sea of rumour. The transfer window, league auctions, retirement whispers, coaching changes — together a din. Within that din, the analyst has one task — to provide a reliable filter. To separate which news has evidence behind it and which is mere air. Every piece should carry at least one new insight — something the reader did not know before.

A bigger question is also rising about the reliability of data. Match information is now so scattered that it is hard to verify who published what and when, and what was later changed. It would be good to have an 'immutable ledger' of cricket information — where each number's source, date and context are recorded, and no one can alter it at will. This idea is like a blockchain ledger — once written, it cannot be erased. An analyst's integrity is similar: once a mistake is written, it stays in history.

Now to the reverse side. We usually think an analyst's value lies in his confidence — how firmly he can predict. But my experience says the opposite. The most valuable analytical answer is often unexciting: 'With this data I cannot say anything certain.'

Yes, this is boring. The reader wants clear language — 'this boy will be a star.' The feed wants a sharp prediction. The sponsor wants excitement. But firmness standing on empty cells is not courage, it is arrogance. An analyst who every time predicts 'either this is the truth or nothing is' actually erases his own chance to admit error.

Note that there is a tempting trap here — imposing hype prematurely. Declaring a teenager 'the next big name' after one clip is the easiest task. The real labour lies elsewhere: his footwork, his ball release, his body language, his sleep, his diet, the pressure from his family — digging all of this out slowly. Hype is made in a day, talent is made over years; an analyst who runs with hype cannot walk with talent.

And one more reverse point. Even without data, we often write anyway, because not writing makes us feel we are not working. But if writing 'insufficient data' in a scouting report helps reach the right decision, then that is the most valuable line in the report. Because a wrong report can ruin a young player's career — false expectation, unequal pressure, the wrong role in the wrong team. The cost of an analyst's error lands on the player's body.

At the end, the coach in that rain-soaked Kuala Lumpur room said, 'All right, then let us watch him for three more months.' I agreed. Three months later we had data from seven matches, and only then could I make an honest prediction for the first time — and even that was not certain, only probable.

The next decade of cricket will bring an even bigger flood of data. But standing in that flood, the question remains the same — which number is true, and which is arranged. Those who can keep an empty cell empty will be the ones who can see a player as a person, not only as a number.

So the next time someone shows you a clip and says 'this boy is the next star', ask them — data from how many matches, in which format, against whom, and who verified it?