Valuing V-League 2026 Strikers with G-xG: 41 Names and Two Opposing Price Signals

On January 6, 2026, I reopened the transfer spreadsheet I first built in August 2026 and pasted 41 strikers into column A. Column B held the asking price quoted by agents. Column C held the goals scored in the 2026 season. Column D held xG. Column E held G-xG, the difference between actual goals and expected goals. Sorting column E in ascending order put a player with an asking price of 8.9 billion VND and a G-xG of minus 2.1 at the very top.

Six years earlier, in that exact cell, a foreign striker also carried a G-xG of minus 2.1 and also topped the list on price. Hai Phong FC called me then, asking who to sign to replace a foreign forward who had scored nine goals the previous season. I read them the name of a 23-year-old priced at 2.5 billion VND, 40 percent below the most expensive target. In 2026 that player scored 11 goals and was sold on for a profit of 3.2 billion VND. In the 2026 window, Hai Phong did not buy a footballer. They bought expected value.

Six years on, the number has not moved. Only the price has.

The dataset and how I built it

The set covers 41 strikers playing in the V-League and the First Division, logged by hand between January 1, 2026 and December 31, 2026, across 96 matches with complete footage. Each player carries 11 fields: minutes, goals, xG, npxG (xG excluding penalties), G-xG, npxG-G, pressures per match, the team's PPDA when that player was on the pitch, average rest days between starts, days lost to injury, and the written asking price from the agent or club.

The last field is the only one that does not come from the pitch. It comes from the telephone. In six years of freelance work, I have never once seen that column agree with column E.

Three metrics need defining before I go further. xG (expected goals) is the probability that a shot becomes a goal, derived from location, angle, the type of pass that created it, and the number of defenders in the way; every shot adds to a team's xG, and the season total shows how many quality chances a striker actually received. G-xG is goals minus xG; positive means finishing above the average for those chances, negative means waste. PPDA is the number of passes an opponent is allowed before a defensive action; the lower the PPDA, the higher the press.

I opened the spreadsheet for the May 2026 V-League match and worked out something I have repeated ever since: tactics never have a gender. At Lach Tray that month, Hai Phong held more possession than SHB Da Nang but generated only 0.8 xG, while the visitors reached 1.9 from seven shots. The home side's PPDA was 9.8, far too high to call it an effective press. A male commentator said the better team had lost to bad luck. I put the numbers in front of him and predicted the second half would bring a concession. The final score was 1-2. From that night I set my own rule: numbers first, commentary second, and a glossary published alongside every piece so newcomers are not left behind.

Based on my experience watching matches at Lach Tray and across northern Vietnam since 2026, a V-League striker averages 2.4 quality chances per match. That figure has barely moved in thirteen years, no matter how many clubs the league has added. The value of a domestic striker therefore sits mostly in converting those 2.4 chances, and in generating extra chances through pressure alone.

The evidence chain: prices rise while chance quality falls

I split the 41 players into four bands by asking price and averaged each band.

| Band | Players | Average asking price | Average G-xG | npxG-G | Pressures/match | Rest days between starts | |---|---|---|---|---|---|---| | A (8bn and above) | 7 | 9.4bn | -1.4 | -1.9 | 21 | 3.8 | | B (5bn to under 8bn) | 12 | 6.1bn | +0.3 | +0.1 | 46 | 4.6 | | C (2bn to under 5bn) | 15 | 3.4bn | +0.8 | +0.7 | 71 | 5.2 | | D (under 2bn) | 7 | 1.5bn | +0.2 | -0.3 | 58 | 5.0 |

Price and quality run in opposite directions across the two bands that matter most. Band A is the most expensive but carries a negative average G-xG, the fewest pressures per match, and the shortest rest between starts. Band C costs nearly three times less than Band A yet finishes 2.2 goals above expectation across the same season, and presses more than three times as often.

What stands out is Band B, the band the market treats as standard. It is almost perfectly neutral on every metric. That is the safe band to buy and the safe band to resell. Band C is where the largest value gap sits and where nobody is digging.

The most expensive striker in the dataset, quoted at 8.9 billion VND, scored 14 goals in the 2026 season. Column C looks excellent. Column D shows the problem: his xG was 16.1. Of that 16.1, 3.9 came from five penalties, leaving just 12.2 in non-penalty xG. His non-penalty goals totalled nine. His npxG-G is minus 3.2. A striker being paid 8.9 billion VND finished worse than the average for the very chances his team-mates created for him all season.

The next column made me stop for longer. His pressures per match came to 19. The league average for strikers is 52. In Band C it is 71. That gap is verifiable with the naked eye on replay: he waits on the edge of the box, skips the first pressing phase, and forces the midfield line to push higher to compensate.

I measured the consequence. When this striker started, his team's PPDA was 12.6. When he sat out, the same team recorded 9.1. A gap of 3.5 units means opponents were allowed roughly four extra passes before every ball recovery. That is why I ruled the name out, exactly as I ruled out the most expensive target in the 2026 window.

The profile worth buying, and who that profile actually is

My filter for this winter window has four conditions, ordered by importance. The order comes from my own mistakes, not from theory.

Condition one: G-xG no worse than minus 0.5 across two consecutive seasons, because a single season is too small a sample to judge finishing skill. Condition two: at least 60 pressures per match, a better predictor of adaptation after a move than goals scored. Condition three: average rest of at least 4.5 days between starts, since every buying club will play a denser schedule. Condition four: no more than 20 days lost to muscle injury in the previous 24 months.

Applied to the 41 names, six players remain. Of those six, three sit in Band C, two in Band B and one in Band D. Their average asking price is 3.1 billion VND, 67 percent below what the top three clubs are preparing to spend on Band A targets.

The benchmark case remains Mac Van Hung, then 23 and playing for Phu Dong. In 2026 he scored seven goals from 6.8 xG, a positive G-xG of 0.2 at the age of 22, with an average of 84 pressures per match. His average rest between starts was 5.4 days. The fee I recommended at the time was 2.5 billion VND. In 2026 he scored 11 goals and the club sold him on for a 3.2 billion VND profit.

I retell this case for one detail that is easily missed. Hung's pressures per match in 2026 were 4.4 times higher than the 8.9 billion striker's. The club spent 2.5 billion VND to buy pressure, not goals. The goals arrived afterwards, as a consequence.

The transmission effect into badminton and into data markets

My core specialism is badminton, and the ranking system of the Badminton World Federation is the cleanest example of a dataset valuing an athlete correctly. Ranking points count only the best ten tournaments across 52 weeks, expire after 12 months, and make no allowance for reputation. A player who enters 18 tournaments in a year gains nothing beyond the tenth. That structure rewards consistency and punishes volatility.

The V-League transfer market has no equivalent mechanism. There is no reference price list, no depreciation schedule, no column forcing a buyer to look back two years. A player can therefore be valued on a single season, usually the one in which he outscored his own xG by the widest margin.

There is one point of contact between these two markets. All the data on rest days, injury days and pressures that I use to price a footballer is the same category of data that betting markets collect and resell. In esports, where match data is published in real time, those markets operate far faster than any regulatory framework, and the result is that competitive integrity erodes before anyone finishes drafting the paperwork. Vietnamese football has not reached that point, but the direction of the data flow is already visible.

The contrarian edge: models err, and the error lives in the dressing room

There is a paradox in the table above that I have to state plainly. Band A, the band my data rates lowest, has signed the most contracts across the last three transfer windows. If the clubs were entirely wrong, they would be paying in points. Some of them are.

The rest are not, and the reason does not sit in my spreadsheet. An xG model cannot measure a striker keeping a dressing room calm, dragging defenders away so a team-mate is free, or making a 19-year-old concentrate harder in training. Those things are real and they affect results, but we cannot yet quantify them, and because we cannot quantify them, player valuation models systematically underprice dressing-room chemistry and systematically overprice young potential.

So when a transfer looks disproportionate to G-xG, I do not assume the buyer is wrong. I log a blank cell labelled model error and wait 20 matches before concluding. A single goal is randomness. A full season is where probability exposes everything.

The biggest risk to this method is its own popularity. Once every club holds a G-xG column, the edge stops being the metric and becomes the choice of which metric to trust. My Band B will become the market price within two years, and the margin will migrate to Band D, where data is thinner and footage scarcer.

Signals for the next window

Three markers I will watch until this window closes. The number of Band A contracts falling below three, which would mean part of the market has learned to read column E. Average rest days appearing in official signing announcements, which would mean clubs have started pricing stamina. And the number of players resold within 18 months at a profit above 2 billion VND, which would mean the buy-low model has become a process rather than luck.

The question I leave for people who do this work: if one striker scores 14 and another scores nine from the same volume of quality chances, what percentage of the first striker's fee should the second command? My spreadsheet says 34 percent. The market says 58 percent. That gap is where a provincial club can buy three seasons instead of one.