The Last of Us Tops the 100 Greatest Games List: The Cost of a Ranking With No Published Rubric
Trả lời cốt lõi: Kinda Funny Games xếp The Last of Us ở vị trí số một trong danh sách 100 tựa game vĩ đại nhất, công bố không kèm tiêu chí hay trọng số. Tranh luận cộng đồng xoay quanh hai thước đo không tương thích: ảnh hưởng lịch sử và chiều sâu hệ thống gameplay. Danh sách không tạo ra dữ liệu cạnh tranh nào cho phân tích esports. Dữ kiện chính: - The Last of Us do Naughty Dog phát triển, ra mắt năm 2013; The Last of Us Part II ra mắt năm 2020. - Danh sách gồm 100 tựa game do Kinda Funny Games công bố; vị trí số một thuộc The Last of Us. - Bài báo mở bằng tranh cãi trước khi tiết lộ thứ hạng, đảo trình tự nguyên nhân và hệ quả. - Không nêu tiêu chí chấm điểm, trọng số, doanh số hay tỷ lệ hoàn thành nào. - Phần tin liên quan nhắc LCK, VCT Pacific 2027 và MLBB Việt Nam nhưng không kèm số liệu. Nguồn: bài phân tích giai đoạn 2 dựa trên bài báo của Kinda Funny Games về danh sách 100 tựa game vĩ đại nhất; thời điểm công bố ước tính 2025–2026 do bài viết nhắc Clair Obscur: Expedition 33. Hỏi đáp liên quan: H: Vì sao The Last of Us được xếp số một? Đ: Phe bảo vệ dựa trên ảnh hưởng lịch sử và dòng chảy sáng tạo từ Uncharted đến The Last of Us, không dựa trên chỉ số định lượng nào. H: Danh sách này có giá trị với phân tích esports không? Đ: Không, vì nó không chứa đội, tuyển thủ, giải đấu hay patch nào; chỉ phần tiêu đề phụ nhắc đến LCK, VCT và MLBB. H: Tín hiệu nào cần kiểm chứng trước khi sử dụng? Đ: Thông tin LCK tăng lượt xem sau sáu năm franchising và việc GAM Esports tham gia VCT Pacific 2027 cần xác minh độc lập vì thiếu đường cơ sở đo.
In my tracking notebook there is a column that has held nothing but zero for five years: the number of “greatest of all time” rankings that arrived with published scoring weights. I began logging it in 2026, after an all-time player vote at a major league cost our desk three days to settle one internal question — what exactly is being measured here. This week the column is still empty. Kinda Funny Games published a list of the 100 greatest games, The Last of Us took first place, and within hours the community split into three camps: defenders of the first game, advocates of the second on gameplay grounds, and a third group pulling Sekiro and a few other names into the argument. At three in the morning in Seoul I read the piece, closed it, opened my spreadsheet, and did what I do after every big match: I counted how many criteria were named. None. There was one first place and one headline.
Kinda Funny Games, a games outlet built by a group of former journalists and hosts from the United States, published a ranking of the 100 greatest games. First place went to The Last of Us, developed by Naughty Dog and released in 2026. The sequel, The Last of Us Part II, arrived in 2026 and has been the centre of nearly every argument about the series since. The article covering the list was built in a telling order: it opens with the controversy, then reveals the ranking, then closes by quoting both sides.
The defending camp chose historical influence as its ruler. They draw the lineage from Uncharted to The Last of Us, argue the game cannot be skipped when the industry’s history is told, and hold that its cinematic approach to storytelling became the benchmark for a generation of AAA production. The opposing camp chose systems. They say Part II plays better, that mechanical depth is what deserves measuring, and that a linear action title cannot sit above games with more complex systems. Between them sits a third group invoking Sekiro as an undervalued candidate.
What matters is that the two sides do not disagree on facts. Nobody disputes the release years, the developer, or the structure of the games. They disagree about the ruler. One measures by influence, one measures by play experience, and both call their result “the greatest”. In my work, an argument shaped like that has only one handling: separate the criteria, publish the weights, and only then let a ranking exist. Without weights, every list is simultaneously right and wrong.
One classification point must be stated plainly: this article does not belong to esports. It has no tournament, no roster, no patch, no organisation as its subject. Tagging it as esports is an error, and a costly one: if pieces like this enter an esports data stream, they distort entity counts, topic distributions and every sentiment baseline behind them. It is still worth reading, for two reasons. First, its argument structure maps almost exactly onto all-time player and team votes in esports. Second, the outlet’s own related-headline block shows it runs an esports vertical: LCK, VCT Pacific, VCT Champions, MLBB Vietnam.
In my spreadsheet, a model that does not publish its weights is not a model. It is an opinion with formatting. When I build an expected-goals figure for a match, I know exactly how much a shot inside the box weighs, how much a one-on-one weighs, how much a strike from distance weighs. Because I know, I can be wrong in a measurable way, and the next time I can fix the exact part that failed. A ranking of 100 games with no weighting section permits nobody, including its own authors, to correct anything.
The key point: a ranking that publishes no criteria is not a broken evaluation, it is a content format engineered so the argument never closes. The article itself records the community challenge — is this judged on gameplay quality, on popularity, or on historical influence — and supplies no answer. That ambiguity survives because resolving it would end the argument within two comments.

The evidence for the “controversy” also needs to be read at the right level. The piece quotes “one player”, “some gamers”, “many people” — no platform, no count, no timeframe. By the standards of an analytics desk, that is a sample size of one presented as a trend. I once built a community sentiment tracker for a tournament and learned an expensive lesson: anonymous quotes systematically inflate consensus, because writers select the sentence that fits the frame they already chose. When I replaced them with real sampling — comments counted by day and by platform — the deviation routinely exceeded 30 percent. The “backlash” in the piece may well be a handful of selected comments.
The article’s structure also deserves to be read as data. It opens with the controversy before disclosing the ranking, placing the effect ahead of the cause. In my language, that is a derivative that was manufactured rather than observed: the temperature of the debate was never measured, it was declared in the first sentence. When a piece announces a reaction before giving the reader anything to react to, that ordering is itself an editorial decision.
Based on my experience tracking matches and votes, I keep a small metric I call the criteria-disclosure index. The method is simple: for every “greatest of all time” list I encounter — football, basketball, esports, games — I record whether weights were published, then follow what kind of argument emerged. Across roughly forty lists, fewer than one in ten published weights, and among the few that did, the dispute almost always shifted into a resolvable form: “I disagree with your weighting” rather than “you are wrong”. This is my own log, not a published study, and I use it as a filter rather than as evidence. It is enough to establish one thing: the problem is not the first-place slot, it is that nobody knows what the slot was calculated from.
Move to esports and the structure repeats almost intact. Every time a forum opens an all-time greatest player vote, camps form along three axes: trophies, peak form, and longevity. The trophy voter wins every popular vote, because trophies are the easiest thing to count. The peak voter loses, because peak requires definition. And nobody publishes the weights between the three axes. The result is an argument that recurs every season, ends in the same place, and reliably generates the highest engagement of the year. I do not believe in inspiration — I believe in standard error. And the standard error of a vote with no weights is infinite.
One smaller detail in the piece is worth logging: the opposing camp, at the end, still concedes that The Last of Us is one of the most outstanding games in history. That is a familiar consensus-manufacturing device. By conceding a weaker claim, the article converts a split into an apparent near-consensus while pre-empting backlash. In match analysis I meet the same structure in teams that win on secondary metrics: they concede territory to protect the result, and the press later calls it character. When the numbers stop lying, my heart starts listening. Here the numbers have nothing to say, so everything left is rhetoric.
The related-headline block is the only place the outlet touches esports, and it is the place that demands the most discipline. One headline says LCK viewership doubled after six years of franchising. There is no absolute figure, no measurement window, no baseline. In my work, a claim like that goes into a verification queue, never into a model. At least two alternative explanations inflate a “doubling”: a shift in the measurement platform, and the inclusion of co-streaming channels in total viewership. Same with the item on GAM Esports becoming a VCT Pacific 2027 partner — if accurate, a notable cross-title move worth tracking, but for now only a headline. And on MLBB Vietnam I set the trigger concretely: a top-four finish at MSC or M-series in two consecutive seasons before I treat it as a regional power shift. In my world, luck is only the residual I have not explained.

One more environmental variable sits inside how the piece tells industry history: it runs from Pong to Clair Obscur: Expedition 33. Using a title released in 2026 as the endpoint of that history is a recency-bias marker, the equivalent of ranking a striker first all-time on the strength of his last three matches. It also gives me a timestamp: the piece cannot predate 2026, and the attached esports headlines carry 2026 dates, so this belongs to the current cycle. For an argument of this type the half-life is short — usually under a month, before the next ranking replaces it.
There is a reverse reading I consider more accurate. The crowd treats the heat of the debate as proof that the list matters. Causality runs the other way: the list is built to produce heat, so heat cannot validate the ranking. This is the correlation-as-causation error I meet every week — a team wins more after changing coaches, everyone concludes the change caused it, when the real variable may be a softer schedule.
If heat does not validate the ranking, then arguing about first place spends energy where there is no leverage. The operative decision inside any ranking is the weighting decision, and that decision is made in an editorial room, not in a comment section. A ranking forms a canon: it tells studios which design paradigm currently counts as top-tier quality. Over a multi-year horizon, that effect is far larger than who sits first for a week.
I have to correct one of my own assumptions. My criteria-disclosure index implicitly assumes transparency is good and opacity is sloppiness. In this case the assumption fails. Opacity here is functional: it is the condition that keeps the argument open, and therefore the condition that lets the format exist. Transparency is measurable, but transparency is not the same as correctness. I can publish my weights and still be wrong — merely wrong in a way others can point at.

The one element worth keeping as a tool is the opposition between the first game and the sequel. It exposes a paradigm split: safe refinement of a proven formula against experimental risk. In esports that split has another name — playing the structure that has already won against playing high-variance. The structural side wins more and gets called dull. The variance side produces peaks and gets called unstable. Both are right, because they are measuring two different things with the same word.
The practical answer does not lie in whether The Last of Us deserves the slot. It lies in readers building their own ruler before arguing: four axes — systems, narrative, historical influence, community durability — with weights declared in advance, and an acceptance that changing the weights changes the answer. The same frame applies to every all-time player vote in esports, and it turns an endless quarrel into an argument that can end.
Three signals I will track over the coming months, in priority order. One, any official LCK viewership disclosure with a multi-year baseline, enough to separate real growth from a shift in measurement platform. Two, a confirmed VCT Pacific 2027 partner list, to learn whether GAM Esports genuinely moves across titles or remains a headline. Three, results for Vietnamese MLBB teams at MSC and M-series, where a top-four finish in two consecutive seasons would signal a real regional shift.
I have counted every empty space on the pitch when the crowd disappeared. In this story the largest empty space is the missing criteria section, and it is missing on purpose. A ranking does not need to be right. It needs to state in advance where it is wrong.
