Low PPDA Is Not Cowardice: Re-reading V-League Defenses Through Pressing Data

In the last three matches, a mid-table V-League team allowed opponents to complete an average of 11.4 passes per defensive phase, the second-lowest figure in the league — lower even than the league leaders. But that team collected just 4 points from 9. This is the starting point of every misunderstanding about pressing in Vietnamese football.

People look at that number and immediately conclude: this team presses poorly, their defense is loose, they let opponents build play comfortably. That conclusion fails because it never asks a simple question: does that team want the opponent to hold the ball, or does it lack the ability to win it back? These two situations produce nearly identical PPDA figures but reflect completely opposite football philosophies.

I first went through this argument in 2026, when I was a young reporter for a football site in Binh Duong, hand-charting data from 182 V-League matches via video. I found a team with the league's lowest PPDA, just 7.8. That means opponents had to complete nearly eight passes before this team made an active defensive action — a duel, a tackle, an interception. The figure 7.8 on the standard PPDA scale sits in the "mid-pressing" range, but in the V-League context, where match tempo is slower and teams tend to pass shorter, 7.8 was the lowest in the league.

I wrote a piece with a provocative title: "Low Pressing Is Not Cowardice." A veteran coach called me a dealer in soulless statistics, saying football cannot be read from a few numbers on paper. Yet a young assistant coach at a Binh Duong club invited me to sit down and build a pressing map for his team. That argument did not make me abandon data — it made me believe more strongly that data only has value when you know how to ask the right question.

Numbers never lie; we simply have not asked the right question.

This article is an attempt to ask the right question: how is pressing in the V-League actually functioning, who is pressing well and who is merely pretending, and why does the PPDA table that so many people cite conceal more than it reveals.

---

Before turning to the data, the method must be made clear. PPDA — passes allowed per defensive action — measures how many passes an opponent is allowed to complete in a certain area of the pitch (usually the 60% of the pitch toward the opponent's goal) before the defending team makes an intervention. The lower the PPDA, the more aggressively the defending team presses. The higher the PPDA, the deeper the defending team sits and lets the opponent hold the ball.

But PPDA has three blind spots that very few V-League analysts address.

First, PPDA does not distinguish between organized pressing and individual pressing. A team can have a low PPDA because one or two players run like mad up front while the rest of the team stands still. On paper, that team looks like a high-pressing side. Tactically, it is a disaster — one player charges forward while ten others do not follow, opening gaps between the lines that the opponent only needs one through-ball to exploit.

Second, PPDA is heavily influenced by match context. A team that is leading tends to press less, giving a higher PPDA. A team that is trailing tends to press more, giving a lower PPDA. If you read a team's season-average PPDA without separating by scoreline, you are reading a number contaminated by its own outcome.

Third, PPDA does not measure the effectiveness of pressing. It measures intensity. A team can press ferociously all match with an extremely low PPDA, but if every ball recovery happens in a harmless area and every dangerous pass still slips through, that pressing has no defensive value. Conversely, a team that sits deep with a very high PPDA but needs only one interception before a counterattack to score — that low pressing is a weapon, not cowardice.

That was exactly the case of the team I analyzed in 2026. PPDA 7.8, letting opponents hold the ball comfortably, but conceding just 0.7 goals per match thanks to lightning counterattacks. Their defense did not defend by applying constant pressure — they defended by creating a trap. They invited the opponent forward, exposed the space behind the opponent's back line, then transitioned within two to three seconds of winning the ball.

If you only look at PPDA, you would say that team was lazy at pressing. If you look at the entire transition sequence, you see a machine programmed to the second.

---

To understand fully, a set of supplementary metrics must be built alongside PPDA. I use four main groups.

The first is pure defensive quality: expected goals conceded (xGA) per match, actual goals conceded per match, and the gap between the two. If xGA is high but actual goals conceded is low, that team is being carried by its goalkeeper or by luck. If xGA is low and actual goals conceded is low, that is a genuinely good defensive system.

The second is pressing efficiency: the rate of ball recoveries in the opponent's third (high turnovers), and the number of goals scored from those recoveries (high turnover goals). A good pressing team must convert pressure into chances. Recovering the ball without scoring is just running.

The third is transition quality: the average time from ball recovery to the first forward pass or shot, and the rate of counterattacks ending in clear chances.

The fourth is match context: PPDA while leading, while level, while trailing. This is the group of metrics that almost nobody publishes in the V-League, and it is precisely where the truth hides.

When these four metric groups are built for a full season, the V-League picture changes completely from what a simple PPDA table suggests.

---

Start with a paradox. The team with the lowest PPDA in the most recent season was not the team that conceded the fewest goals. The team that conceded the fewest goals had a mid-range PPDA — sixth or seventh lowest in the league. If high pressing is the key to defense, why did the least-pressing team defend best?

The answer lies in formation structure. The team that conceded fewest played with two holding midfielders close together, keeping the distance between the defensive line and the midfield line under fifteen meters throughout the match. They did not need to press to break the opponent — they used density. An opponent wanting to play through that block would need three to four accurate passes in succession in an extremely tight space, a compound probability far lower than facing a high-pressing team that is easy to play through.

In other words, that team chose a form of defense based on probability. They did not try to win the ball high up the pitch — they tried to make each opponent pass a little harder. Accumulated over hundreds of passes, that little extra difficulty becomes a season of few goals conceded.

This is where probability modeling comes in. Goals in football are a rare event. In an average match there are about twenty-five to thirty shots and two to three goals. The probability of any single shot becoming a goal is so low that a single match says almost nothing about a team's quality. We can only read quality by repeating the same structure across many matches.

That is why xG matters. A team that lets the opponent take fifteen shots but each with an xG of 0.04 has a total xG of 0.6 — meaning the opponent's average expected goals for the match is under one. A team that lets the opponent take eight shots but three of them have an xG of 0.3 or above has a total xG of nearly one goal. Same shot count, completely different defensive quality.

In the V-League, the problem is that official xG data barely exists. No provider tracks shot locations for every match. I had to build my own xG model for the V-League, based on shot location (distance, angle), shot type (left foot, right foot, header), and situation (open play, set piece, counterattack). This is a nightmare of "missing data" — but it is also precisely why this work has value. When nobody has data, the first person to build it sees what others cannot.

---

Apply this model to a specific team. I take a mid-table team, call it Team A, to avoid turning the article into a personal attack. Team A's season-average PPDA is 9.2, meaning fairly aggressive pressing relative to the V-League baseline. But when split by scoreline, the figure changes strangely.

When Team A is leading, their PPDA rises to 11.8. When Team A is level, PPDA is 9.0. When Team A is trailing, PPDA drops to 7.4.

Read conventionally, we would say Team A presses hard when trailing and sits deep when leading — the rational behavior of any team. But combine it with xG data.

When Team A is leading and sitting deep (PPDA 11.8), their xGA is 0.9 per match. When Team A is trailing and pressing hard (PPDA 7.4), their xGA is 1.4 per match — higher.

Meaning when this team presses hardest, it defends worst. Pressing intensity rises, but defensive quality falls. This is a classic sign of disorganized pressing: when trailing, players charge forward on instinct, breaking structure, exposing gaps between the lines, and the opponent only needs to counter to get a clear chance.

This is the biggest blind spot of reading football by PPDA alone. The number says Team A presses better when trailing. Reality says Team A is self-destructing when trailing.

By contrast, the team that conceded fewest has a PPDA of 9.6 when level, 10.2 when leading, and only 9.0 when trailing. They barely change their pressing structure with the scoreline. It is a system programmed for stability, not reactive to emotion.

That stability is the value. Over a long season, the team that maintains a stable structure will fluctuate less than the team that reacts to the scoreline. And in football, less fluctuation usually means more points.

---

Now to the weakness of my own analysis. A probability model measures tendencies, but it fails to capture several important things.

First, my xG model does not know who is shooting. A shot from a 0.08 xG position in the boots of a top foreign striker has a different conversion probability than the same shot from a defender. Individual finishing quality is a variable that basic xG models ignore. In the V-League, the quality gap between top foreign strikers and the rest of the league is large, so this variable is especially important.

Second, the model does not know the weather. A V-League match in heavy rain, with a flooded pitch, has a completely different xG from a dry match. The same shot can differ by five to ten percentage points depending on conditions. This is a variable European football can ignore, but Vietnam cannot.

Third, the model does not know the referee. A referee who allows heavy contact reduces the effectiveness of high pressing, because more duels are penalized, causing the pressing team to pick up cards and lose rhythm. Conversely, a referee who allows strong pressing benefits the high-intensity team. I once observed a team changing its PPDA noticeably according to the referee — something no model without a referee variable can see.

This brings me to a principle I have held since a coach called me an eccentric: PPDA and xG are tools, not truth. A number only has value when it helps us understand a specific detail of the match. If a number does not lead to a new question, it has no value at all.

Numbers never lie; we simply have not asked the right question.

---

By now the central paradox of V-League pressing is clear. But that paradox leads to a second, deeper question: if high pressing is not the key to defense, why do the top teams still use it?

The answer lies in volume. A high-pressing team does not defend by reducing the opponent's goal probability — they defend by reducing the time the opponent has the ball. If Team A controls the ball 65% of the time, Team B has only 35% of the time to create chances. The chances Team B creates in that 35% may have a higher xG if they counter, but the total number of dangerous approaches is smaller.

This is a probability trade-off: reduce the opponent's chance quantity, increase the opponent's chance quality. A team good at high defending makes the quality increase insufficient to offset the quantity decrease. A poor team does the opposite — conceding a few extremely high-quality chances, enough to lose the match.

In the V-League, most teams do not have the personnel to play high pressing with discipline across a season. They can do it for one match, sometimes two in a row, but when the schedule becomes dense, pressing quality drops, and the true face of the defense is exposed.

This is why I always split pressing statistics by time cycle. A team with a PPDA of 8.5 in the first ten matches and 11.2 in the last ten can look like a high-pressing team if you read the season average. But in reality they collapsed as the season wore on — a sign of poor fitness or insufficient squad depth.

I once tracked a team across two consecutive seasons. In the first, their PPDA held steady at 8.8 all season. In the second, with essentially the same squad, early-season PPDA was 8.6 but late-season 12.0. They finished the second season six places lower than the first. The change was not in player quality — it was in the ability to sustain intensity.

This is a pattern that data sees before the table sees it. Midway through the second season, while that team was still in the upper group, I already knew they would fall. The number had spoken while the table had not yet reacted.

---

Another aspect often overlooked when analyzing V-League pressing is the role of the foreign player factor. Many V-League teams depend on foreign strikers to score, and opposing defenses know it. When Team A faces Team B with a top-class foreign striker, the most effective defensive tactic is not high pressing — it is cutting the supply of the ball to that striker.

This explains a phenomenon I observed: in matches between a team with a strong foreign striker and a good defensive team, both teams' PPDA is unusually low. The good defensive team presses hard to deny the opposing midfielder time to spot the striker. The attacking team presses hard to recover the ball quickly and feed the striker before the defense organizes.

The result is a match that looks chaotic and high-tempo but is really a battle around one individual. This is a case where positional data would be valuable if available — a heat map showing how that striker was isolated. But the heat map is also a trap. A heat map shows where a player is, not what he does there. A striker with an empty heat map could be an isolated striker — or a striker doing the job of stretching the defense so teammates can exploit the space.

The heat map has become a new form of divination in football analysis, and in the V-League it is especially dangerous because positional data is too sparse for reliable interpretation. A heat map from one match says nothing. A heat map from ten matches might say something, but only when read together with tactical context — which very few people do.

---

Now the most counter-intuitive question: is the entire V-League analysis market reading PPDA wrong?

The answer is not "right" or "wrong" — it is "reading without context."

The problem is not the metric. The problem is how it is used. When a team has a low PPDA, there are at least four plausible explanations: that team is pressing high in an organized way; that team has one or two individuals pressing alone; that team is trailing and forced to push up; or that team is facing an opponent passing too quickly to close down. These four causes produce the same number but demand four different tactical responses.

When a coach reads a report with only PPDA, he does not know which cause he faces. When a TV analyst cites PPDA, the audience does not know what that number is concealing. When a fan says their team "presses poorly," they might be talking about a team that deliberately chose to sit deep to counter.

This is the biggest tactical blind spot of Vietnamese football at present. Not missing data — data not placed in context. Advanced metrics have entered the analysis, but the method of reading them has not kept up.

We think we understand the game, until the data table opens our eyes.

---

So how do you read it correctly? There is a triad of questions I apply to every team before drawing conclusions about pressing.

First: how does this PPDA change with the scoreline? If PPDA drops sharply when this team trails, that is a sign of reactive pressing, not tactical pressing. If PPDA is stable regardless of scoreline, that is a system.

Second: when this team recovers the ball in the opponent's half, how many goals does it score? If the conversion rate is low, that pressing is burning fitness without value. If the conversion rate is high, it is a genuine weapon.

Third: how does this team defend when it is not pressing? If it only knows how to press and not how to sit deep, it will collapse against an opponent that can control the ball. If it can switch between the two modes, it is a complete team.

These three questions, applied across the V-League, produce a picture quite different from the PPDA table. Some teams look like good pressers but are merely reacting to the scoreline. Some teams look like poor pressers but are controlling matches through structure. Some teams can do neither and survive only on the individual quality of a few players.

Teams in the third group are the most dangerous in analysis, because their data is noisy. They can win a big match on an individual moment, making the analyst think their system works. Then three matches later, that moment does not come, and they collapse. If you read their data through a probability lens, you see the collapse coming long before.

---

Here I must address an aspect I consider most important but least discussed: the relationship between pressing and youth development in the V-League.

A high-pressing team requires every player in the system to read the game the same way. The striker must know when to charge forward to force the opponent wide. The midfielder must know when to step up to block the backward pass. The defender must know when to hold the line and when to track the man. This is a distributed cognitive system — each player makes decisions independently but within the same framework.

Such a system requires training time. And in the V-League, training time is the scarcest resource. Teams play dense schedules, travel a lot, and often change personnel on short cycles. A coach wanting to build high pressing needs at least one season for players to understand each other. Very few get that season.

This is why teams that succeed with pressing in the V-League are usually those with a core stable over multiple seasons, or those with a long-tenured coach. Personnel stability is a precondition for pressing. You cannot press high if your teammates change every three months.

This has a counter-intuitive consequence: poor teams, unable to buy many foreign players, have better potential to build good high pressing — as long as they keep their squad. While rich teams, constantly changing personnel, often fail at pressing due to a lack of understanding. This is one of my favorite counter-intuitive patterns in the V-League: poor but stable can beat rich but shuffled.

Of course, "can" is not "will." Poor and stable still needs a coach who knows how to build a system and a coaching staff that understands data. In the V-League, those two conditions do not always appear together.

---

Let me synthesize everything into a concrete match-reading model for the next round.

Suppose Team X meets Team Y. Team X has a season-average PPDA of 8.9, Team Y of 10.1. Conventional reading says X presses harder than Y. But split the data.

If X's PPDA is 8.5 when leading and 9.3 when trailing, X is a stable pressing team. This is a good sign. If X's PPDA is 7.0 when trailing and 11.5 when leading, X is a reactive pressing team — they only press when forced to.

Now Y. If Y's PPDA is 10.1 but Y's xGA is only 0.8 per match and they recover the ball in the opponent's half an average of 6.2 times per match, then Y is not a poor pressing team — Y is a team playing a tight defensive block and counterattacking on plan.

When X meets Y, the most reasonable prediction is not that X will dominate. The reasonable prediction is that X will have more of the ball, Y will sit deep, and the match will be decided by X's ability to play through Y's block. If X is a reactive pressing team, they will struggle because Y's defensive block does not press, so there is nothing for X to react to. If X is a systematic pressing team, they can control the match but still need an individual moment to break the deadlock.

This is the kind of prediction I believe can be verified through research design. Not score prediction — score prediction is divination. Predicting match structure is science. You can check after the match which team had more of the ball, who pressed more, where the chances came from. If your model correctly predicted the structure, you have learned something. If it predicted wrong, you have also learned something.

The important thing is not to confuse the two levels of prediction. Predicting the scoreline from a probability model is one thing. Predicting match structure is another. The first is affected by variance so large that it is nearly meaningless for a single match. The second is far more stable and can be verified.

In 2026, I staked my entire career on a probability model named Croatia. Not because I believed Croatia would win — but because I believed my model read that team's structure correctly. Croatia was not a miracle, but a well-managed variance. They did not win because of destiny. They won because they created more chances than their opponents across a sample large enough that variance did not break the trend.

That is a lesson applicable to the V-League. There are no miracles in Vietnamese football. Only well-managed models and poorly-managed ones.

---

Before closing, an admission about the writer's own limitations.

I have a bad habit: when a topic stops being hot, I drop it and jump to a new one. I opened three or four research projects at once on pressing, penalties, xG, and empty-stadium data, and I have not finished any of them. The xG model I use in this article is a version I always meant to upgrade but never had the time to complete.

This is a real limitation, and it affects analytical quality. A complete model would produce firmer conclusions. A half-finished model produces provisional conclusions that need further verification.

I say this not to excuse myself, but to place every conclusion in this article in its proper context. What I have written is a way of reading, not the truth. If you have better data, use it to refute me. That is the only way the V-League analysis community progresses.

Numbers never lie; we simply have not asked the right question.

---

So what signals should be tracked in the next round?

First, watch the PPDA of mid-table teams when they concede early. If their PPDA drops sharply while xGA rises, that is a sign of ineffective reactive pressing — an illness that will surface in the second half of the season.

Second, watch teams whose PPDA is stable regardless of scoreline. Those are teams with a system. Over a long season, systems usually beat inspiration.

Third, watch teams that play a tight defensive block yet still collect points. The question is not whether they are lucky — the question is whether the quality of their structure is sustainable across thirty rounds. If it is, they will be in the upper group at season's end. If not, they will be the first to collapse.

And fourth, watch the question itself. Do not follow the PPDA table as a ranking of the best pressing teams. Follow it as a table of questions — every number is a question mark to be resolved with context, time series, and the naked eye.

Vietnamese football does not lack data. It lacks people who know how to ask the right question. And when you start asking the right question, what people call miracles — a poor team crowned champion, a deep-lying defense that never concedes, an isolated striker who still scores — will become well-managed variances, little by little, until you understand that they were never miracles at all.

The applause on an empty stand recorded a truth nobody wanted to hear: that most of what we call football instinct is simply data that has not yet been read. When we read it, the match becomes clearer — and also harder to predict, because we know exactly what we do not know.

That is the true gift of data analysis: not certainty, but grounded humility. And in a league like the V-League, where any prediction can be shattered by an individual moment, grounded humility is the only thing worth bringing into the next round.