The A-League Data Revolution: A Spreadsheet That Never Watched Kick-Off

On an evening in 2026, sitting in my Brisbane flat, I opened a spreadsheet I had named "Roar_underlying_v3". The goal was simple: figure out how much of Brisbane Roar's fourth-place finish was luck and how much was craft. Forty-two points against 36.8 expected points. Jamie Maclaren's 19 goals against 14.7 xG. That gap between the two numbers became the centre of my career. I realised that a large part of what Australia calls a "data revolution" in football is a tactical sketch drawn in PowerPoint that never actually walks onto the pitch. So let me ask the question directly: when smaller leagues import analytics, what context do they import it into, and who pays the bill?

For the past three weeks I have been sifting through match-by-match A-League data, because late in a regular season, teams that slide down the table almost always leave a process signal first. In the 2026-26 regular season, Melbourne Victory's PPDA (passes per defensive action) rose from 9.4 to 13.1 over their last five matches. At the same time their high-intensity sprints per 90 dropped from 21.7 to 18.9. Read together, those two numbers tell you what a scoreline never will: the first line of the press is no longer synchronised, so opponents are walking through midfield into the defensive third. And yet that week's headlines were about Victory's "grit and fight".

This is the real trap. A-League broadcast language and club analytics departments speak two different dialects. On one side sit possession, passes completed and distance covered — the three metrics that get treated as proof of heroism in commentary. On the other sit progressive passes per possession, shot-ending sequences and post-shot xG, which show exactly where a match was lost. In my spreadsheet, I looked at eight A-League clubs across three seasons. The sides finishing outside the top six averaged 51.2% possession but produced only 5.8 shot-ending sequences per match. Half the ball, fewer than six entries into the opposition box. Possession percentage here is not a certificate of quality; it is a lacquer on sideway passing.

Take distance covered, the other comforting number. A team runs 113 kilometres in a match, and 38% of it is backward tracking or futile chasing after a lost ball. I once matched timestamps for one midfielder who covered 12.4 kilometres, of which only 2.9 kilometres was connected to on-ball decisions. The rest was shadow running. Shadow running produces beautiful bar charts that circulate on social media as "relentless work rate".

The real problem is not the number but the question behind it. A-League xG models often borrow their weighting from the Premier League or the Bundesliga, yet shot quality, goalkeeper positioning and defensive line height are entirely different here. A situation the EPL model rates at 0.08 xG will be rated 0.14 in the A-League, because the box is more open, defensive compactness at smaller clubs is lower and goalkeeping is more reactive. A club that runs recruitment, training and tactical adjustment off that imported signal is simply making fast decisions in the wrong place. Training-ground discipline improves, but match-day pressing triggers still fail.

I wanted Germany to prove me wrong. Their 2026 group-stage exit proved me right instead. The same logic applies to the A-League: a club that sees itself only in the mirror of the points table and possession percentage never measures its own throat. When I covered an A-League match in 2026, a man in the stands told me afterwards, "This team runs around, but it never goes forward." That sentence said more than my spreadsheet ever could.

But here I want to turn back on myself a second time. The more I mock the data, the more I have to admit that neither lens survives alone. The eye test will tell a club to change the manager, but it cannot identify the pattern of why the press broke. The model will say xG is falling, but it cannot tell you the dressing room has lost trust. The truth is that the space between number and eye is the most honest place in analysis. Those who understand that gap make tactical decisions on process; those who do not settle accounts on results.

Last week I went to watch a small A-League club's training session. An analyst sitting outside the pitch told me he codes 34 sequences per match, and only four of them leave a footprint in the coaching staff meeting. The rest become files. It is a terrible waste, but the waste is not in the data — it is in the bridge. Data is built, models are built, but nobody builds the language that turns them into match-day decisions. That is where the smaller league's data revolution stalls, and that is where my next piece lives.

My biggest doubt about the A-League's data revolution is this: I want to stake one specific prediction. When the first ten rounds of the 2026-27 regular season finish, of the four clubs that launch new positional-data models, at least two will sit in the bottom six. Imported models take time to establish their own truth, and the league table does not give time.