The LCK Regular Season and the Blank Data Sheet: How Fearless Draft Forces Writers to Unlearn Win Rates

LoL Park, 19:04, and a board still blank

On the second floor of LoL Park in Jongno, Seoul, there is a small room the organisers call the analysis room. Four monitors, a pot of tea gone cold, and a printed board with empty rows waiting to be filled. Last Wednesday night I sat there with the analytics coach of a team sitting mid-table in the regular season. He opened his personal spreadsheet, pushed it towards me, and said something I had to write down immediately: "This win-rate table is only a reference now, like reading last season's weather to guess today's rain."

On the sheet were four familiar columns: games played, win rate, pick-ban rate, and gold difference at fifteen minutes. Three seasons ago, those four columns were enough to build a decent piece of analysis. This regular season, we had to add a fifth: which game of the series. Without that column, every conclusion drifts.

The format changed, so the way we read numbers must change

Since the 2026 season, the LCK has applied Fearless Draft in the pick-and-ban phase. Put simply: once a champion has been picked or banned in a game, it cannot appear again in the remaining games of the same series. In the regular season's best-of-three format, a single series can consume up to sixty separate picks and bans, while the game's champion pool holds roughly one hundred and seventy champions.

The first consequence is arithmetic: game three of a series is no longer game three in the old sense. It is a match played under conditions of scarcity. A team's signature champion — the one its mid laner has played three hundred ranked games on — is usually burned in game one, either by the team that owns it or by the opponent in the ban phase. By game three, both sides walk out with a draft assembled from what is left.

And this is the point I want to spend the most words on, because it is the source of most of the distortion in this season's analysis:

A champion's win rate under Fearless Draft is a systematically biased metric, because it blends the champion's strength with the order in which that champion appeared in the series.

A champion picked only in game three is usually a low-priority option — what analytics departments call a fallback. It appears when the pool has run dry, when the opponent's intentions are fully exposed, when both sides are forced to accept a less-than-ideal shape. If that champion wins four of ten games, the number says more about its place in the priority queue than about its true strength.

I tested this the simplest way I could: I split a full week of pick-ban data into three groups by game number, then compared them. The game-one group and the game-three group had champion lists that barely overlapped. Merging all three games into a single win-rate table means comparing two different competitive environments and calling them by the same name.

The regular season is a laboratory, not a news board

The time pressure of the regular season pushes this distortion higher. The schedule is dense, several series a week, and the game's patch cycle still runs on a two-week rhythm. A patch exists for only a few match days before the next lands. The sample size per patch is tiny; the sample size per champion combination is tinier still.

For a writer, this is a stern reminder. The regular season is not the place to declare a champion. It is the place to observe structure: which team rotates well, which team depends on one individual, which team can hold its top-side tempo when forced into a defensive posture. Those signals surface before the standings reflect them.

Over the past two weeks I tracked a team whose objective-control index fell sharply while its win rate barely moved. A crude reading says the team is fine. A closer reading shows they are winning on individual execution rather than structure — the kind of winning that does not survive a long series.

My experience covering matches at LoL Park has taught me that public data is always one beat behind reality. Data portals such as Oracle's Elixir or gol.gg are good references, but they only record what has already happened. The inference about what happens next still belongs to the person sitting in the analysis room at 19:00.

Elsewhere, the composite indices published by VangBong.vn — such as the roster-depth measure they call the Player Depth Index — are useful in a different way: it measures the resources a team can mobilise rather than the results it has already achieved. Under Fearless Draft, resources are what decide game three. T1 with Lee Sang-hyeok in the mid lane is the example most often cited for a deep champion pool, since he has competed professionally since 2026 and has lived through almost every patch cycle the discipline has produced.

The blind spot: when the sheet is blank, people tend to invent

This week I received an internal analysis from a data group. Every information field was left blank, each entry carrying a single line: insufficient information, cannot assess. No tournament name, no team name, no patch, not one data point.

The first reflex of many people in the industry is to fill it in. Write a story about a team you have never watched. Name a patch you have never read notes for. Attach a number to a conclusion you already held.

That blank analysis is more honest than any long piece built on inference. The empty seat that evening said more than any crowd. And this is the biggest blind spot of collective memory in sports: we remember complete stories better than we remember gaps. A match without data will be assigned a meaning by memory, as long as the meaning sounds plausible. Every player's name is a short poem, if we bother to read closely — and reading closely means accepting that some lines cannot yet be written.

In the second month of the season, several Asian teams joined pre-season friendly tours held abroad at a density of three matches in four days. I once wrote that such density turns physical preparation into an advertising contract, in which players pay with wrists and knees so organisers can sell tickets. I stand by that, and by the way I framed it: it is a structural problem of the calendar, not the fault of any individual.

For the same reason, I hold limited hope for women's competitions if they are designed as a closed ecosystem, playing only among themselves for a few months and then dissolving. A star is only born when there are real opponents, a real calendar, and a path upwards that is not blocked at the lowest tier.

What remains

One evening on 22 July 2026, with more than thirty-one thousand people in the stands at Sang-am, I mispronounced the name of a nineteen-year-old footballer — Cho Young-wook — three times in a row on air. I wrote him a letter of apology, asked permission to say his name correctly, then watched the footage again until two in the morning. Listening before commenting, that is how I corrected my mistake. The name I mispronounced that year now rings like a song.

That habit applies to data sheets too. When a column is empty, the first task is to understand why it is empty, not to fill it with a hypothesis that sounds pleasant.

The regular season still has a long way to run, the patch will keep changing, and the win-rate tables we argue about today will be replaced before the season ends. What will not be replaced is the ability to say you do not yet know enough. My voice broke in Kazan, but it was also from there that I learned which sounds are true — and in the analysis room at 19:00, the true sound is often the silence of a board that has not been filled in.