The Pressing Index: 2,300 Matches, the 8.5 PPDA Threshold, and How to Read an Empty File

August 2026, a second-floor meeting room at Hai Phong FC's training ground. I placed an A4 sheet with the U19 index table on the table. Exactly one line was circled in red: PPDA 6.8. Seven people were in the room. Nobody asked what PPDA was. Coach Truong Viet Hoang asked one question: how tall is he. One metre seventy. He nodded, folded the sheet, and the meeting moved on. Three weeks later, Vu Minh Hieu's name appeared on the first team's match registration list.

PPDA is the number of passes an opponent completes per defensive action by a player. The lower the figure, the earlier a player intercepts the ball, the higher up the pitch he applies pressure. A reading of 6.8 sat in the best bracket I had ever recorded in the V.League youth system at that time. What mattered more was that nobody on the coaching staff had seen this index before that meeting. They saw a slight, short player who did not stand out in training. Two people on the same stand, watching the same session, and seeing two different footballers — because one had a data sheet and the other did not.

Round 17 of the 2026 V.League season, Hai Phong hosted Hanoi FC at Lach Tray. Minh Hieu came on in the 58th minute. Across the remaining thirty-two minutes he won the ball fourteen times and delivered one assist. Hai Phong won 2-1. The next day an assistant coach sent me a one-line message: show me the PPDA table for the whole squad.

From then on I put PPDA and ball-recovery counts into every youth report I wrote. I did not place the number above the eye. I placed it beside the eye, in a position where emotion in a meeting room could not shift it.

The current transfer window poses a familiar problem: more information than ever, less reliability than ever. Every day brings dozens of figures — transfer fees, wages, release clauses, minutes played, expected goals, duel counts. Most arrive without a source, without a timestamp, without a sample. A number missing all three carries zero information value. Worse, it takes the place of a number that does have a source.

Across the last four transfer windows I have logged every transfer rumour involving V.League clubs that appeared on domestic sports outlets. The share carrying a concrete source — a named club, a named agent, a document — hovered around twenty per cent. Four of every five remaining items were built by reinterpreting an earlier item. This is why I never begin a player assessment from a rumour list. I begin from the contract, from actual minutes played, and from a wage estimate built on the club's salary structure.

In Vietnamese football the problem sits a layer deeper. Youth player data barely exists in public form. Academies keep scattered records, mostly in notebooks or an assistant's personal spreadsheet, and those spreadsheets do not travel with the player when he changes clubs. When a twenty-year-old is promoted to the first team, the first question a coaching staff asks is usually what does he have, not what has he done over the past three years. Those are two different questions in kind. The first needs an eye. The second needs a file.

I once believed this gap would close quickly. Ten years ago I wrote that by 2026 every V.League club would employ at least one dedicated data officer. The number of clubs with such a role today still fits on one hand. It is one of my wrong predictions, and I record it rather than quietly dropping it.

In March 2026, football stopped. That was the period in which I learned the most about the limits of my own trade.

The day football stopped, I began counting every stride again.

Four months. No live data, no new matches, no scouting trips. I sat down and re-audited the full dataset of five V.League seasons and three major European leagues, 2,300 matches in total. The initial goal was simply tidying a database. It became a model.

The model has three variables. First, the squad's average PPDA, not an individual's, calculated across every turnover in a match. Second, the defensive line's distance from goal at the moment possession is lost — measured in metres, drawn from positional data. Third, the midfield line's closing speed in the first five seconds after the ball is lost. The three variables measure the same thing: the cost an opponent must pay to move the ball past the halfway line.

The grouping came out sharply. Teams averaging a PPDA below 8.5 took 1.8 points per match. The rest took 1.2. A gap of 0.6 points per match, multiplied across thirty matches, is eighteen points — the difference between a continental cup place and the bottom half of the table.

What I did not expect was how stable the 8.5 threshold proved across seasons. It barely moved. Across five V.League seasons and three European leagues, the league-wide average PPDA shifted with tactical fashion — teams pressed higher in later periods, lower in earlier ones — but the boundary between the effective group and the ineffective group held. That boundary was more stable than any attacking metric I have tested, expected goals included.

I published the model under the name Pressing Index on my personal blog. It did not generate a large debate. It generated a stream of emails from assistant coaches asking how to calculate it. To me that is a better sign than an article shared many times over.

But the most interesting part of those four months lay elsewhere. Running cross-checks, I found a substantial share of records were incomplete. Matches without pressing data. Players with data from only six games in an entire season. Seasons where the organiser's statistical table did not match the club's internal table — differing by a few percentage points on passes, by a few dozen on touches.

At first I planned to delete those records from the dataset. That would have produced a cleaner model. Then I realised it would have produced a model cleaner than the truth. I kept them and flagged them. The missing-data rate became a column in the table rather than a row crossed out.

I did not watch Germany lose. I watched a number that does not know how to lie.

In June 2026 I published a pre-World Cup analysis. Germany's qualifying data showed an average PPDA of 9.2 — for a reigning champion, that sits outside the safety zone. Circulation speed in the attacking third had fallen 11 per cent against four years earlier. Expected goals per pass into the box sat at the tournament average. Combine the three and the picture is not that of a team at the peak of its powers.

I wrote that Germany would exit in the group stage. The online reaction was predictable: people decided I only knew how to read numbers. On the night of 27 June 2026, Germany lost 0-2 to South Korea, firing twenty-six shots with a total expected-goals value of 1.5, and went out. Twenty-six shots, no goals. That is the entire story of the match, wrapped in two figures.

Based on my experience following matches across many seasons, a team can win with twelve shots and lose with twenty-six, and that contradicts no rule. It simply means the final result carries a large share of randomness, while the quality of chances is far more stable. That is why I never assess a season from the league table alone.

The lesson I took from that night was not that the prediction landed. It was that I had presented the data as an argument. An argument can be rebutted with other data, and that very capacity for rebuttal is its strength. A claim that cannot be rebutted is not an argument; it is a belief.

Since then, every report I write opens with a specific figure, with a source, with a date.

There is a trap data people fall into more often than fans: reading the silence of the data as the cleanliness of the subject.

When a player has no injury history on file, we lean towards calling him durable. In truth his file is empty. When a team has no pressing metric, we lean towards saying it does not press. In truth nobody measured. When a scouting report lists no risks, we lean towards calling the player risk-free. In truth the author never met him off the pitch.

All three share one logical error: treating the absence of evidence as evidence for the opposite. In my trade that error costs more than a wrong forecast. A wrong forecast costs credibility. An empty file read as a clean file can lead a club to sign a three-year contract with someone who has never played ten consecutive matches at any level.

During the four months of data auditing in 2026, I began marking the missing rate on every column. The practice makes a report look worse in form. It makes it more honest in substance.

This is why I stopped placing definitive conclusions at the end of reports. A report ending in an absolute statement usually conceals three gaps the author does not want to admit. My reports end with a list of what is still missing: fewer matches than the minimum threshold, no injury data, no data from a higher level of competition. The reader knows exactly how much confidence they are holding.

Numbers are a mirror. Most of the market looks into it and sees only itself.

A pressing midfielder needs no showmanship. He needs the right place, the right moment, and the data standing on his side.

Hai Phong taught me: the star is not on the shirt, it is in the index.

Next round, as transfer stories keep pouring in, I will read them the same way: find the source, find the date, find the sample. Any figure missing those three does not enter my table, however many pages it appears on. Players nobody mentions in any report may still be the ones running to the right place, and my job is to hold a place for them in the spreadsheet before anyone holds a place for them on the front page.

Data does not fill a gap. It only shows where the gap lies. And knowing where you do not yet know is the first step of any honest measurement.