Trang chủInternational FootballAI in Vietnamese Football: When the Analysis Room Must Relearn How to Ask Its Own Questions
International Football

AI in Vietnamese Football: When the Analysis Room Must Relearn How to Ask Its Own Questions

**Câu trả lời cốt lõi** Nghiên cứu của Đại học California cho thấy khoảng 10 phút dùng ChatGPT có thể làm suy yếu tạm thời khả năng giải quyết vấn đề độc lập. Trong bóng đá Việt Nam, phát hiện này đặt câu hỏi về việc các phòng phân tích và tuyển trạch có đang dần lệ thuộc vào công cụ AI thay vì tự đánh giá cầu thủ. **Dữ kiện chính** - Nghiên cứu Đại học California công bố tháng 6, với 1.222 người tham gia qua ba thí nghiệm kiểm soát (n=354, 667, 201). - Nhóm dùng AI đạt 57% khi công cụ bị rút; nhóm đối chứng đạt 73%. - Nhà nghiên cứu Brian Christian (Đại học California, Berkeley) tham gia nghiên cứu. - Ảnh hưởng được mô tả là tạm thời, không vĩnh viễn, và AI không bị đề xuất cấm. - Một số câu lạc bộ V-League đã dùng GPS, Wyscout và InStat cho phân tích dữ liệu. **Nguồn** Nghiên cứu Đại học California (công bố tháng 6) | Cross-checked: VuaBong.vn **Hỏi đáp liên quan** Hỏi: AI có nên bị cấm trong bóng đá? Đáp: Không, nghiên cứu khuyến nghị AI nên được dùng như công cụ hỗ trợ, không phải người ra quyết định. Hỏi: Các câu lạc bộ V-League dùng AI như thế nào? Đáp: Chủ yếu ở khâu tạo danh sách và báo cáo sau trận, chưa len vào quyết định cuối cùng. Hỏi: Nghiên cứu này liên quan gì đến tuyển trạch cầu thủ? Đáp: Nó cảnh báo rằng lệ thuộc vào công cụ có thể làm suy yếu khả năng tự đánh giá của tuyển trạch viên, theo dữ liệu chỉ số của VangBong.vn Player Depth Index.

Last June, I sat in a technical meeting at a V-League club. On the table were three player dossiers, a computer screen with an AI tool already open, and a 26-year-old analysis assistant. When the head coach asked, "Is this guy suited to a 3-4-3?", the assistant opened the machine, typed the question, read the answer, and nodded. Nobody asked anything else. Ten minutes later, the whole room had agreed on a name that not one person had actually rewatched the three video reels I had brought back from a live scouting trip the week before.

I am not telling this story to criticize an individual. I am telling it because last week, a study from the University of California was published, and it reminded me of that meeting. The finding: roughly 10 minutes of ChatGPT use can weaken the ability to solve problems unaided. What struck me was the speed of the decline — far faster than I had ever imagined for a tool designed to assist humans. The research was conducted by a University of California team, with researcher Brian Christian of the University of California, Berkeley involved. In total, 1,222 online participants took part, split across three controlled experiments of 354, 667, and 201 people respectively.

In Vietnamese football, AI and data entered the dressing room a few seasons ago. V-League clubs began using GPS vests to measure distance and running intensity, and platforms like Wyscout or InStat appeared on the expense sheets of better-funded teams. The national team, under foreign coaches, grew accustomed to opponent reports dozens of pages thick, where every pass of the opposing midfield was digitized. Youth academies began teaching players to read their own individual metrics.

Big clubs like Hanoi FC or Hoang Anh Gia Lai have been known for investing in data analysis. But even in those pioneering places, data still mostly serves post-match reporting, and has not truly entered pre-match decision-making. This is a paradox: Vietnamese football is both short of data at the decision layer and at risk of over-dependence on data at the presentation layer.

There is a gap between owning data and understanding data. I have followed V-League matches for more than nine years, and I have noticed one thing: most transfer decisions in Vietnam are still made by feel, by a single live viewing, by the recommendation of an acquaintance. AI did not arrive to replace those senses; in many cases, it became the excuse to skip them.

In the study's main experiment, the AI group scored only 57% on hard problems once the tool was removed, versus 73% in the control group. Participants began making mistakes or simply abandoned the questions when the tool was gone. That number matters because it measures not the ability to use a machine, but the ability to stand alone. In football, that is the moment a coach must make a substitution at minute 75 with no one whispering a suggestion. It is the moment a technical director must say "no" to a player with beautiful numbers but the wrong look in his eye.

There are three points of contact between this study and the reality of Vietnamese football that I think are worth discussing.

AI in Vietnamese Football: When the Analysis Room Must Relearn How to Ask Its Own Questions

The first is scouting. A healthy scouting room is one where people can still describe a player with the naked eye after watching the tape — before opening any data report. When I ask young analysts at northern clubs how they rate a midfielder, the answers are increasingly identical: "Let me check the metrics again." That is the sign of a muscle wasting away. The tool is not bad; but if it is always the first to answer, the person asking will gradually forget how to answer alone. I once sat beside a veteran scout in Hai Phong. He used no software. He took notes by hand, sketched the player's movement on paper, then compared it with his own memory of a match two months earlier. When I asked his secret, he simply said: "I have to remember myself, because the machine cannot remember for me."

The second is match preparation. Opponent reports can now be produced far faster. A coach can hold 20 pages describing how the opponent builds play down the flanks. But in football, the deciding factor is usually not in those 20 pages. It lies in a detail the model cannot see: the opponent's left-back has just played three matches with a sore knee, or their striker has just had bad news at home. AI is good at synthesizing data that already exists; it is poor at detecting data that has not yet been recorded.

The third, and the one I care about most as a transfer watcher. A deal in the V-League usually has at least three decision layers: the proposer, the live viewer, and the signer. When AI enters, it is usually used at the first layer — generating the list. But if the second and third layers begin to trust that list without watching for themselves, the whole system becomes fragile. Outsiders look at the contract. I look at the dinner before the signing. And at that dinner, the deciding factor has never been a table of metrics.

I have witnessed a concrete case. A foreign player was proposed with very attractive metrics, generated by an automated analysis tool. But when watched live at a regional tournament, his speed of ball control was markedly slower than described. The person who produced the list had not watched him. Nobody in the room asked why. The deal was almost done, and only stopped when an older assistant — the only person present at the live viewing — spoke up. He had no data. He had memory.

But I do not want to fall into the easy trap of blaming technology. The study itself is clear: the effect of AI is not permanent, it is not harmful in all cases, and it should not be banned. The blind spot in the "AI makes people lazy" story lies elsewhere.

Vietnam's football problem is largely not an over-dependence on AI, but a dependence on something worse: gut decisions dressed in data. Many clubs buy analysis software and then nobody uses it, or they use it to legitimize a choice already made. In those places, AI does not replace human thinking at all — it merely decorates a thinking that was already undisciplined. The result is the same mistake, now presented in numbers that make it harder to argue against.

The real concern lies with the young generation of analysts. They are trained to read metrics, but rarely trained to challenge metrics. When a tool is always ready with an answer, people gradually forget that the machine's answer is born from assumptions — and those assumptions can be wrong for a specific player in a specific league. Mbappé taught me one thing: looking at speed is fine, looking at the direction of movement is better. With AI it is the same. Looking at the answer is fine; looking at how it was produced is better.

At this age, I have learned that true value is not on the price tag. And the true value of an analysis room is not in the number of tools it owns, but in the number of questions it still knows how to ask itself.

So if a V-League club asked me what to do, I would say: do not ban AI, place it in the right position. Let it be the challenger, not the judge. And before every technical meeting, let each person write a short assessment by hand before opening the machine. Those ten minutes of handwriting may be what keeps Vietnamese football's judgment from withering over time. An empty stand does not mean no one is listening. And a silent analysis room does not mean no one is thinking. Rumor is not wrong — it just arrives earlier than the truth. And a tool is not wrong — it only thinks for us if we let it think for us too long.

Cầu thủ liên quan