One N/A Row in the Transfer Window: Pricing Players When the Data Doesn't Exist

Three in the morning in Hai Phong. The mid-season registration window had been open for four days, and I was sitting in front of a tracking sheet with eighteen rows.

Seventeen rows had numbers. Games played, KDA, kill participation, gold share, average minutes per game, professional pick-ban rate. Row eighteen — the name that at least three community pages were calling "the most interesting signing of the window" — was completely empty. Value cell: nothing. Games cell: nothing. Source cell: nothing.

"Three in the morning, the market is asleep. That is when the numbers are most awake."

That night I did not write a single word. I highlighted row eighteen in red and waited.

Context: a transfer window that runs on belief

The mid-season transfer window in Vietnamese esports has a feature the standings never show: most deals are decided in far less time than it takes to verify them. A team needs a new top laner, has three weeks, and in those three weeks must choose between a player with full public data who costs a lot, and a player with a thin record who comes with a very good recommendation.

I have followed this market since 2026, when I was still competing and organising tournaments, before moving into the data side. The structure of the problem has barely changed in nearly two decades. What has changed is the supply of data.

In top-tier leagues, every game is captured through an API. People know exactly how many minions a player has at minute ten, where he stands on the map when his team loses an advantage, what percentage of his early-game movement goes into the enemy half. In lower-tier leagues, or in regions where the publisher does not open detailed data, people only have VODs — and VODs do not count. Viewers remember beautiful plays. Spreadsheets do not.

The gap between those two worlds is where pricing mistakes are born.

When a cell is empty, the model starts to invent

In every data sheet I have ever built, an empty cell is the most dangerous cell, not a bad-value cell.

A player with poor defensive stats is still a player with stats. He can be coached, placed in a suitable system, protected by a team structure. A player with an empty cell is someone nobody knows anything about — and when nobody knows anything, humans tend to fill the empty cell with narrative. That is basic psychology: the brain cannot tolerate blank space.

The most dangerous thing in a dataset is not a bad number, but an empty cell filled with expectation.

I lived through exactly this mechanism in June 2026, when I analysed the profile of Rimario Gordon for a sports news site. Fourteen matches, an expected-goals figure of 0.32 per match, the lowest among ten foreign strikers in V.League at the time. I predicted he would score five goals that season, and he scored exactly five before his contract was terminated. But the lesson I kept was not "I was right." The lesson was in that press room: nobody argued about the number. They argued about who I was.

Since then I have had one rule: every claim must come with a raw data table, and every empty cell must be labelled empty, never inferred.

Three cases that show complete data is still not enough

If there is one lesson to take from nearly two decades of transfer markets, it is this: complete data does not protect you from mistakes. It only makes your mistakes legible.

The first case is Do Duy Khanh, known as Levi, leaving GAM Esports for 100 Thieves for the 2026 season. At the time I had almost every required metric: he was a dominant jungler in the region, his kill participation was the highest I had recorded for a Southeast Asian jungler, and his impact on early skirmishes was measurable. My model said the deal was reasonable. Context said otherwise: language barriers, a different pacing of play, a different way of reading games between a region built on fighting and a league built on lane control. None of that sits in a column.

The second case is Le Quang Duy, SofM, joining Suning in 2026 and reaching the League of Legends World Championship final on 31 October 2026 in Shanghai, where his team lost 1-3 to DAMWON Gaming. This is the opposite situation. At the time of the transfer, the public data on him in the region was not detailed enough to build a decent valuation model. People relied mostly on the feeling they got from VODs. The final outcome far exceeded any quantitative forecast I had ever seen for a Vietnamese player.

"Charts do not lie, but they do not tell the whole story. I look for the part that was left out."

The third case is row eighteen in my sheet that night — the case I had no name to write about, which is precisely why I did not write.

Why I will not price a player without data

There are three questions I always ask before putting a number on a player with a thin record.

First: is the sample big enough to say anything? Three games at a friendly tournament is not a sample. It is a single observation. I have seen contracts signed on the basis of seven games in a domestic championship where the opponents were two tiers weaker. Those seven games predict nothing except that the player can beat weaker opponents.

Second: does the metric transfer to a new environment? A player with a high kill participation rate on a weak team is often not high because he is good, but because he is the only one who can generate kills. Put him on a strong team and that metric drops — and people will call it a decline, when in fact it is context adjustment.

Third: is this data gap random? If a player has no public data because he competed in a league that was not recorded, that is an infrastructure problem. If he has no data because his old team deliberately hid scrim results, that is a completely different signal. The same empty cell, two causes, two opposite conclusions.

"My numbers do not need applause. They need to be right — time is the referee."

The counter-intuitive angle: sometimes the silence is the data

There is a paradox it took me years to accept. In a market where everyone is trying to publish as many numbers as possible, a team publishing nothing is the most valuable information of all.

Teams preparing a big move tend to behave in a very specific way: they stop posting content about the player about to leave, they push scrims into time slots with fewer viewers, they change their line-up structure in non-competitive friendlies. No cell records any of that. But it is data.

"The night in Hai Phong taught me one thing: people look at the price board, I look at the movement board."

At the same time, I have to state the opposite clearly, so as not to turn myself into a data fanatic. The paradox of reading silence is that it can be misread in both directions. I once watched a team accused of "hiding something" simply because they had no staff to produce content. Not a signal. Just a shortage of people.

And this is the part where my spreadsheet regularly fails. I once sat in an empty stand during the period when every tournament had to be played without spectators, and noted something simple: with no roar behind them, some young players performed as if liberated, others as if they had lost their reference point. Their metrics before and after were nearly identical. The thing that differed had no column to hold it.

My first-hand experience of watching matches live gives me one humble conclusion: data is a map, not the territory. A map does not tell you whether the road has potholes.

What I am waiting for in the next registration round

Row eighteen is still red in my sheet. It may be filled by a successful deal, and I will have to write that my model missed a variable. It may be filled by a failed deal, and people will ask why I did not say so in advance.

The answer to both possibilities is the same: because at that moment I had no data. Without data there are no conclusions, only expectations — and expectations are not analysis.

What I want to see in the next registration round is not a blockbuster signing. I want to see one team publish properly formatted scrim data, enough for an outside analyst to verify. Then the white space on the map starts to shrink — and my job finally means something.

For tonight, in Hai Phong, I am leaving that row red.