Revaluing the LCK 2026 Transfer Window: Three Undervalued Names and a Misread Model

In the opening match of the LCK Summer 2026 group stage, a mid laner ended his game at minute 34 with 312 minions, a 78% kill participation rate, and 612 damage per minute. The 612 figure ranked in the top three across the entire league at that point in the season. The salary that two independent transfer sources confirmed to me placed him eleventh among all active LCK mid laners. Eleventh. A top-3 damage player on an eleventh-ranked salary for his position. That is the gap the transfer market leaves behind, and it is the starting point for this entire analysis. I spent four weeks tracking the mid-season transfer market between Vietnam and South Korea, gathering data from Oracle's Elixir, gol.gg, internal salary sheets from three organizations, and direct conversations with two agents. My goal was never to report who moves where. I track the transfer market not to catch the news, but to catch the patterns. And the pattern here, once the noise is filtered out, is fairly clear. In this article, I will present a valuation framework built on four variables that I use to re-read player value in the current transfer window, apply it to three specific names the market is mispricing, and then point out the blind spot that my own model shares. This is the part I call public self-correction, except this time I am correcting myself before the market has the chance to do it for me. The first thing to say is about context. The mid-season 2026 transfer window is happening in a market whose structure has changed compared with three years ago. After the LCK moved to a franchise model alongside this summer's addition of academy slots to the main stage, the supply of high-tier contracts compressed. Big organizations no longer buy at scale. They buy selectively, and precisely because of that, mispricing has become more expensive. When a team only buys one player, choosing the wrong one is no longer a small mistake that two backup signings can paper over. It is an architectural decision. The second layer of context is the flow of Vietnamese talent into South Korea. Over the past two years, the number of VCS players invited to try out with LCK and LCK Challengers teams has risen sharply. I spoke with an analytical coach from a mid-tier Seoul team in March, and he offered a remark I recorded verbatim: Vietnamese talent is highly rated for micro mechanics but undervalued for macro indicators. In other words, people pay for their ability to win lane, but they do not pay for their ability to read the map. This asymmetry is one of the market's largest blind spots. Before moving into the individual cases, I have to set my own error threshold. My valuation model, which I call VPM (Value Per Minute), is calibrated on 428 LCK and VCS games over the past 12 months. If a player's VPM deviates more than 22% from market valuation for two consecutive months, I treat it as an undervalue or overvalue signal. If my model fails to predict at least 65% of renewals over the next three transfer windows, I will publish a correction and rebuild VPM from scratch. That is my commitment to readers, and it is how I bind myself to the falsifiability of the data. What the data does not see. Before going further, I will say plainly what many inside the industry know but few write down: the four variables I am about to present cannot measure the locker room. A player with a high VPM can still destroy a roster if he refuses to communicate. I have watched an LCK Challengers team fall apart for reasons that were entirely outside the spreadsheet. So read what follows as a talent valuation model, not a verdict on a person. My VPM framework rests on four variables. The first is damage per minute per unit of gold received (DPM/Gold). I use this ratio rather than absolute damage, because a player fed resources will always show high damage. In fact, in my 428-game sample, absolute damage explained only 41% of the variance in win rate, while the gold-to-damage conversion rate explained 58%. That is a large enough gap to change how the market reads. The second variable is conditional objective control, coded OCC. I define OCC as the share of major objectives (Dragons, Herald, Baron) a team secured in games where that player held a lane advantage at minute 14. The reason I add the lane-advantage condition is to separate a player's ability to convert personal advantage into team resources from the ability to benefit from an already-strong team. A mid laner who wins lane but cannot pull objectives has not created macro value. The third variable is the pressure resilience index, coded PSI. I measure PSI as the rate of being targeted in the first 60 seconds of each rotation to a side lane, multiplied by the survival frequency after that attack. Put simply, PSI is the capacity to endure being focused. I borrowed this variable from an old observation in the team sport of football, where a low PPDA on a champion side reflects control rather than aggression. In esports, a player who draws heavy attention and still survives is a frontline defender. The fourth variable is the space-creation score, coded SCS. This is the hardest to measure and the most contested. I measure SCS by counting how many times a teammate can reach an objective or attack a map angle within five seconds of that player forcing an opponent out of position. SCS attempts to quantify what casters often call gravity — the ability to make opponents react, creating space for teammates. It is the variable closest to the unmeasurable, and I include it with a full reliability caveat. With that four-variable framework, I scored the three names below. I deliberately avoid real names in the first part of the analysis because I want readers to absorb the mechanism before they read the label. Names appear later, alongside concrete evidence. The first case is a mid laner on a team ranked 6th to 8th in the LCK. His DPM/Gold sits in the top 15% of the league, and his OCC is 71% — meaning in games where he held a lane advantage, his team secured 71% of major objectives. That figure is nearly 20 percentage points above the average for mid laners in the same rank band (52%). This is what I call a textbook undervalue: he converts personal advantage into team resources at a top-3 team's rate, but his team ranks 6th to 8th, so his price is anchored to the team's standing. In other words, the market prices him by team results, not personal contribution. This is a systemic error common in team sports. In football, this phenomenon is called outcome attribution bias. A playmaker who records 12 assists in a season while his team finishes ninth is often valued below one who records 7 assists while finishing second. The market does not read process; the market reads the table. The second case is a jungler on a rebuilding team. His standout metric is a PSI of 84% — that is, in 84% of times he was attacked during transitions, he survived. The league average is 61%. His space-creation score SCS is also high: 4.2 per game against an average of 2.7. But his DPM/Gold sits only in the middle band, and that is why the market files him as mid-tier. What the market overlooks is that he is carrying the topside for a team whose two solo lanes frequently lose lane. This is where I want to pause and stress something about how I watch games. Based on my experience watching matches over many years, I have come to see that most analysis looks only at scores and damage, while ignoring how much a jungler has to move to patch holes. A jungler covering two losing lanes will show low attack numbers, but that is a consequence of the assignment, not of ability. The scoreline is a liar; data is the only witness I trust — but even data must be read in the context of its assignment. The third case is a young marksman from the VCS currently in a tryout phase in South Korea. His DPM/Gold sits in the top 20% of the VCS sample, his OCC is 68%, but his PSI is only 49% — meaning he is attacked and dies significantly more than average. This is precisely the Vietnamese asymmetry I mentioned. The Korean market looks at the low PSI and concludes he is not ready. But when I dug deeper, his low PSI was unevenly distributed: in the first 10 minutes his PSI was only 38%, while from minute 20 onward it rose to 71%. What does this mean? It means this young marksman dies a lot early — a phase where he lacks adequate protection at VCS level — but once the game reaches teamfighting phase, his survival rate jumps above average. He is not mentally weak. He is someone who dies because he is abandoned early, and survives because of good positional instinct late. But the market only sees the aggregate PSI of 49% and applies a label. I wrote a note to an LCK team about this name in mid-June. In that note, I wrote that if he were placed in a system with proper support and jungle cover, his DPM/Gold had a 64% probability of landing in the top 10% of the LCK within one season, based on the distribution of comparable samples. I marked confidence as medium-high, and I still stand by this prediction. At this point I need to shift to the most important part of the article: the blind spot of this very model. Over the past four weeks, I re-tested the 428 games with a variant of VPM in which I removed the SCS variable. The result made me think hard. When SCS is removed, the model's predictive accuracy in identifying successful signings drops from 67% to 52% — roughly equivalent to a coin flip. This confirms that SCS, the hardest variable to measure, is the most important one. The paradox is that the market barely prices SCS, because it never appears on the live broadcast stat sheet. An agent told me in early July something I treat as the key to this entire transfer window. He said teams buy players based on what they can present to ownership, not what they can present to the head coach. And what they can present to ownership is KDA, damage, standings. SCS is not in the PowerPoint. That is why the market keeps paying premiums for players with flashy stats while lowballing players who create space. There is one more layer I want to add, and this is the most counterintuitive part of the analysis. For years I assumed that distance covered and burst frequency were signs of effort. But after cross-checking the data, I had to admit something hard to accept: running without purpose also produces pretty numbers. A player who moves a lot may be losing position rather than applying pressure. In my sample, the correlation between distance covered per minute and win rate is only 0.09 — statistically almost meaningless. That is why I removed the effort variable from VPM last October. What I just said does not mean effort is worthless. It means effort must be priced by purpose. A rotation has value only if it opens space for a teammate or creates a threat to the opponent. This is why SCS includes purposeful movement, while raw distance is excluded. The market does the opposite: it counts distance and ignores purpose. Now to the part where I have to speak plainly about a subject few in the industry want to address. In the mid-season window, sponsor pressure is acting more strongly than ever. A low-ranked team needs a headline signing to retain sponsors, and headlines tend to come from players with flashy stats, not players with high SCS. As a result, the market is distorted toward preferring marketability over winning ability. In the short term this may be acceptable as a business matter. In the long term it creates a self-destroying loop: a team buys a flashy player, does not win, loses sponsors, then buys another flashy player to keep sponsors. I want to return to the first case, the mid laner ranked 6th to 8th, and give a concrete valuation figure as I do in every analysis. His current salary, per two sources, sits in the middle band for his position. Read through VPM, his fair value sits in the top 25%. The gap between market price and fair price is roughly 40% to 55%, depending on how performance bonuses are counted. That is a clear undervalue, past the 22% error threshold I set. For the jungler on the rebuilding team, the undervalue is smaller but still meaningful, roughly 25% to 30%, mainly because his PSI and SCS are superior. Notably, at least three teams hunting for a jungler this window did not place him on their shortlist. I verified this with an assistant coach, and he confirmed the reason for exclusion was that the candidate's attack metrics were unremarkable. That is a decision made from the stat sheet, not from the film. For the young Vietnamese marksman, this is the case where I attach the highest confidence in potential, but also the highest risk. His probability of succeeding in the LCK within one season, per my model, is 64% if placed in the right system and 31% if placed on a team without jungle cover. This is why I told that LCK team that signing him must come with a technical condition, not just a purchase decision. A contract without a usage plan is a neglected investment. Now to the counterargument. There is one claim teams often make to defend their choices, and I want to treat it seriously rather than dismiss it. The claim is that past data cannot predict the future in a new environment, because the meta changes, teammates change, and stage pressure is greater. I concede this point to a degree. But when I tested it, I found that macro variables such as OCC and SCS are far more stable over time than micro variables such as KDA or damage. The correlation of OCC between two consecutive seasons is 0.71, while KDA is only 0.48. In other words, map-reading is a durable skill, while teamfight performance is a volatile one. This leads me to a conclusion that may be contentious in Korean analytical circles. I believe the transfer market is betting wrong on stability. It assumes that flashy stats from minor leagues do not translate to major leagues, and therefore buys familiar names. But my data shows that the quieter macro indicators are precisely what translate best across leagues and metas. Buying a familiar name with a formed reputation is politically safe internally, but not necessarily technically safe. There is another way to put it that I prefer: the market pays for the past and ignores structure. A team buys a player with 200 peak games behind him who is now in decline, instead of a player whose skill structure fits the system. In my sample, players over 27 show an average DPM/Gold decline of 14% from peak, while average SCS declines only 6%. Space-creation skill decays more slowly than individual finishing skill. This is an argument for buying high-SCS players, even when they are older. Of course, I must be careful with the correlation-versus-causation fallacy. That high SCS correlates with high win rate does not mean SCS causes high win rate. Both could be consequences of a third variable I have not measured, such as in-team communication. A good communicator may be trusted and supported by teammates, and thereby have both high SCS and many wins. I flag this limit clearly to avoid repeating the mistake I criticize: reading correlation as causation. This is also why I add a research-methods section to my analyses. I want readers to be able to check the data, not just trust the conclusion. In the appendix of this piece, I list the four variables, their computation, and the error thresholds I use. I do not claim to be right. I claim to be checkable. Back to the bigger picture of the transfer window. If I had to sum up the current market in one line, I would say it is a market paying for spotlight and ignoring structure. The most talked-about signings are the ones with flashy stats. The most valuable signings, per the data, are often the ones barely mentioned. This information gap is where competitive advantage lives. For Vietnamese teams preparing for their domestic transfer phase, my message is clear. Do not imitate the LCK market's habit of paying premiums for flashy stats. Instead, build an internal evaluation system that measures space creation and pressure resilience. Even a rough version of SCS and PSI beats looking only at KDA. Because while the big market keeps mispricing, the small market holds the advantage of being a smart buyer if it knows what it is looking for. For Vietnamese players seeking opportunities in South Korea, I have advice grounded in data rather than inspiration. Your early-game PSI is dragging your value down, not your late-game. That means train the early phase and communicate with your jungler, because that is where the data scores you low. Improving early-game PSI by 10 percentage points could lift your valuation by a salary tier. That is a concrete number, not a pep talk. I return to the principle I live by in this profession: before the ball rolls, the number has already whispered the result. In the transfer window, before the contract is signed, the number has already whispered who is being mispriced. My task is not to state the future with certainty, but to point out where the market is misreading the signal, and to place a public bet that can be checked. I will place the public bet now. Of the three names I analyzed, I predict at least two will post a next-season VPM at least 30% above the current season, and at least one will sign a contract in a higher salary tier than his current one. I set confidence in this prediction at 70%. If by the end of next season I am wrong beyond this threshold, I will publish a public correction and adjust the VPM framework. The transfer stage is always a place where noise exceeds signal. Rumors, sponsor pressure, herd psychology — all of it pushes prices up and down on emotion. Amid that noise, the data practitioner has one advantage: the calm to look at structure rather than spotlight. I do not trust feeling. I trust chances created, space opened, pressure endured. The question I leave readers, and myself, during this transfer window is not who will win the title. The question is: among these four variables, which one is the Vietnamese market overlooking the most, and do we have the courage to buy the players the PowerPoint does not speak for?