Mel and Patch 26.22: Riot Deletes the Very Thing That Made a Champion
**Câu trả lời cốt lõi**: Riot Games làm lại kỹ năng W của Mel trong LMHT bản 26.22: bỏ cơ chế phản chiếu đòn đánh, thay bằng phá hủy đường đạn, tạo lá chắn và tốc độ di chuyển cho đồng minh gần đó. Thay đổi nhằm giảm cảm giác ức chế khi đối đầu và tinh chỉnh toàn bộ bộ kỹ năng của tướng. **Dữ kiện chính**: - Bản 26.22 dự kiến lên máy chủ ngày 4 tháng 11 năm 2026, khoảng ba tuần sau công bố ngày 14 tháng 10 năm 2026. - W mới vẫn phá hủy đường đạn, nên cảm giác ức chế cấu trúc khi đối đầu vẫn còn nguyên. - E giảm thời gian trói chân ở cự ly gần, tăng ở tầm xa tối đa — thưởng cho giữ khoảng cách. - Nội tại tăng hồi máu và giáp theo cấp; Q chia lại sát thương; R tăng hồi chiêu, giảm hệ số theo cấp. - Bài công bố không nêu tỷ lệ thắng, tỷ lệ chọn, tỷ lệ cấm hay tần suất pha phản chiếu. **Nguồn**: Báo cáo phân tích chuyên sâu Stage-2 về bản cập nhật LMHT 26.22, công bố ngày 14 tháng 10 năm 2026. **Hỏi đáp liên quan**: - Hỏi: Mel có còn chơi được đường giữa sau bản 26.22? Đáp: Có thể, nhưng giá trị của cô ấy chuyển dần sang tiện ích giao tranh cuối trận và khả năng xoay vai trò hỗ trợ. - Hỏi: Vì sao Riot xóa cơ chế phản chiếu thay vì giảm sức mạnh? Đáp: Cơ chế có tính nhị phân, không có tham số trung gian để tinh chỉnh sau nhiều vòng điều chỉnh. - Hỏi: Ai hưởng lợi nhiều nhất từ thay đổi này? Đáp: Các tướng phụ thuộc đường đạn như Jhin và Seraphine, cùng nhóm đấu sĩ lao vào.
Jhin's fourth shot leaves the barrel, and in that exact instant Mel presses W. The bullet turns back on its owner, detonates, and the Jhin player dies to the strongest hit in his own kit. In an ordinary ranked game, that sequence is nothing more than an angry status line typed before queueing again. But once it is cut into a clip and posted, it becomes a different kind of evidence entirely: emotional evidence.
Six seconds. That is the full runtime of the clip that kept the League of Legends community agitated for months. No miraculous jungle path, no Baron steal, no individual performance worth cutting into a highlight reel. Just one player pressing one button and another player dying to his own damage. The frustration does not come from losing. It comes from the system turning around and biting the player. In seven years of watching matches through spreadsheets, I have learned one thing about this category of clip: it travels faster than any chart, and it misleads decision-makers faster than any chart too.

On 14 October 2026, Riot Games announced a rework of Mel's W in patch 26.22. The projectile-reflect mechanic — which the design team itself described as the champion's signature trait — is being deleted. The new W destroys incoming projectiles and grants a shield plus movement speed to Mel and nearby allies. This is a rework-level change, not a numbers pass.
But read the description closely and one detail stops you. The new W still destroys projectiles. Which means the feeling of "I pressed my ability and nothing happened" survives intact. The only thing removed is the feeling of "my ability killed my own teammate". Those two feelings do not share a root, and that is where this analysis begins.
Context: an ability that cannot be balanced, not an ability that is too strong
Mel was designed as a mid-lane mage with an unusual defensive philosophy: the best way to protect yourself is to turn the opponent's attack into their own weapon. In a game where every ranged champion depends on projectiles, that creates a very specific interaction layer — it does not block damage, it returns it.
The problem with mechanics like this is not that they are strong. The problem is that they are binary. A reflect either catches the hit or it does not. There is no middle ground to tune into. Riot stated the team went through multiple rounds of adjustments trying to preserve the mechanic before finally abandoning it. That is the most important detail in the entire announcement, and it is the one most often skimmed past.
Design teams do not discard a signature mechanic after several failed tuning passes because they like change. They discard it when the mechanic is structurally untunable. Reduce the reflect window and the ability becomes useless. Increase it and it becomes oppressive. No parameter sits between those two states, because the payoff of the ability does not come from a number — it comes from reading the timing correctly.
Patch 26.22 does not stop at W. It is a full-kit tuning pass. The passive loses early-game damage and duration but gains health and armor scaling per level. Q front-loads less burst and redistributes it into subsequent hits. E reduces root duration at close range and increases it at maximum range. R gains cooldown and loses per-level scaling.
Read those four changes side by side and a shape appears that is invisible when you read them one at a time. E is the sharpest signal: less root up close, more root at maximum range. That is a tax on self-peel when dived, and a reward for holding correct distance. The kit is teaching Mel players a new posture — sit further back, hold spacing, play on your teammates' rhythm.
The R nerf runs in the same direction. More cooldown and less scaling means Mel's value no longer sits in pressing ultimate to start a fight. She loses the "wait for R then go" play pattern and is pushed toward generating continuous value through smaller actions.
On timing, patch 26.22 is expected on live servers roughly three weeks after the announcement, in early November 2026. That is the late-season window. If the 2026 international calendar follows the usual patch-lock practice, the change cannot appear at a major event already in progress. The competitive shock is therefore deferred, and teams get a full offseason to adapt. I hold this at low-to-medium confidence, because it depends on two unverified variables: the exact release date and the 2026 event calendar.
And here is the part that must be said plainly: the announcement contains no performance data at all. No win rate, no pick rate, no ban rate, no presence figures. No PBE data. No frequency count for reflect-swing moments. The sample size for the central question in this story is zero. Every conclusion below must therefore be read as conditional inference, not statistical finding.
W changes jobs: from punishment tool to peel tool
This is the core, and it is not a power reduction. Removing reflect and adding an ally-facing shield plus movement speed is an identity migration: W moves from a punishment tool to a peel tool, and Mel changes champion category from a defensive anti-burst mage to a utility mage with flex potential.
A reflect ability operates on the defender's logic: you wait for the opponent to act, then you answer with their own action. A shield-and-movement-speed ability operates on the opposite logic: you must actively stand next to someone else, anticipate when the fight breaks, and choose the right person to cover. Same button, two fundamentally different decision models.
I read the support pivot as the intended outcome rather than a side effect. Ally-facing shields and movement speed are the signature of the enchanter class. Add E's max-range reward and the kit is describing the posture of a support, not a solo-lane duelist. Confidence: medium.
A cohort loses a skill, not a stat line
A damage nerf is survivable: good players recalculate thresholds, change all-in windows, change push timings. Deleting a mechanic is different. Mel's competitive value rested on a reactive read — deciding when to hold W to catch a Seraphine ultimate or a Jhin fourth shot. When the mechanic disappears, that skill is not converted. It is invalidated.
This is the harshest class of balance change: not a numbers nerf but a skill-obsolescence event. Dedicated Mel players lose their accumulated capital and receive no commensurate compensation, because no statistic measures the skill of holding W.
Football has an analogue I have watched long enough to trust. When the offside interpretation changed, strikers who made their living running in behind did not lose pace or finishing — they lost the exact thing that made them valuable. The spreadsheet still records 10.5 km per match. Those runs simply stopped producing goals. The eye watches one match, the data watches a completely different one — and both are right.
Half the frustration is removed
Riot states the goal is to limit situations that feel too frustrating when playing against the champion. That goal is only half achieved. The new W still destroys projectiles, so the structural frustration survives. Only the spectacular frustration is gone.
The distinction matters more than it appears. Structural frustration is frequent, steady, and a direct consequence of the opponent owning a projectile-block ability. Spectacular frustration is rarer but leaves a much deeper mark, because it breaks the expectation of damage ownership.
The rework targets the second kind. That is the friendlier choice for communications, though not necessarily the right one for design. A Jhin player who gets his fourth shot reflected will remember it for a long time, but he meets it a handful of times per season. A Seraphine player who fires an ultimate into Mel and watches it evaporate meets that far more often.
Examples selected for salience, not for frequency
Riot's two illustrations are reflecting Seraphine's ultimate to charm the entire enemy team, and reflecting Jhin's fourth shot back at Jhin. Both are the most cinematic edge cases in the game. Neither comes with any frequency figure.
This is a bias I meet constantly in my work: sampling by memorability instead of by frequency. It is not dishonest, but it produces a picture weighted toward what is vivid. Balance decisions driven by viral clips rather than frequency data always carry over-correction risk.
In 2026, as a high-school student in Munich writing a data blog, I argued against a well-known commentator who said Croatia were simply lucky in the World Cup semi-final. I rewatched all seven Croatia matches, minute by minute, and showed their shot quality was overwhelmingly superior. The piece was mocked — a child lecturing experts. What I learned was not that I was right, but that to refute an impression you must count how often that impression occurs. Nobody counted for Mel. Including me.
The flex paradox: lose lane power, gain draft value
Here is the door I want to leave open, because without it this piece is just a well-sourced complaint.

Purely as draft value in professional play, the change is not clearly negative. An ally-facing shield plus movement speed opens a genuine support flex. Flex picks are disproportionately valuable in pro drafts because they delay the role reveal until the opponent has committed. Losing mid-lane power while gaining role ambiguity can be net-positive for pro viability. This is counter-intuitive and I hold it at medium confidence, because it depends on whether the new kit is actually strong enough to support — something only post-patch data can settle.
One small detail runs the same way: the passive now scales armor. Armor scaling on a mage passive is unusual, and it points toward surviving dives rather than winning duels. That is a defensive signal, not a damage signal.
Who wins and who pays
The clearest beneficiaries are projectile-dependent champions who were punished by their own damage — Jhin, Seraphine and the wider artillery-mage family. Second are melee divers and assassins who previously could not safely dive a W-holder. Third, and least discussed, are Mel's own teammates, who now receive shield and movement speed from an ability that used to serve only her.
The losers are Mel specialists and the teams that used her as a counter-pick into projectile-heavy compositions. At pro level, a rare counter-pick like that carries distinct tactical value, and that value has just been deleted.
On the solo-queue side, fairness requires noting the rework genuinely solves the opponent's discomfort. Dying to your own ability is one of the worst experiences a competitive game can produce, because it violates the most basic rule of damage ownership. Removing it is a sound player-experience decision.
The Vietnam–Germany lens: one number, two readings
I grew up in Vietnam and work in Germany, and the difference in how the two places read the same data still catches my attention. A high frustration rate on Vietnamese servers is usually read as a champion being too strong in solo queue. The same figure on German servers is usually read as a champion being hard to coordinate in scrims.
Two readings produce two solutions. The first demands a power reduction. The second demands a structural kit change. Patch 26.22 chooses the second, and I think that is correct — but it also shows the decision was made from one specific reading, not from globally published data.
At 17, during the 2026 empty-stadium season, I built my own dataset because no clean source existed. It showed home teams losing a clear share of their average points versus the previous five seasons. A German football outlet published it. The lesson was not "empty stadiums favour away teams" but this: when clean data is missing, people substitute collective impression. Riot is in exactly that position with Mel.
The contrarian angle: what exactly are we measuring?
The entire decision rests on an unverified premise: that Mel is frustrating enough to require intervention. No win rate, no ban rate, no satisfaction index was published. There is a very strong correlation between clip volume and balance decisions, but correlation is not causation. Clips go viral because they shock, and they shock because they are rare.
There is also an internal contradiction worth flagging. Riot says the change will make Mel more balanced while preserving her unique identity. But the unique identity in question is the reflect mechanic that has just been deleted. Both statements cannot hold in the strict sense.
I also have to argue against myself. If the reflect mechanic truly was binary and untunable, deleting it is correct, and my scepticism is just nostalgia for an elegant mechanic. A design team that tried multiple adjustment rounds before giving up has probably already thought what I am thinking. My only counter-evidence is the absence of data, and absence of evidence is not evidence of absence.
What will tell us whether this worked
Four signals over the first two weeks of 26.22. Mel's support pick rate, because that is where the new kit wants her. Her mid-lane presence, because a collapse there means the compensation was insufficient. Scrim leaks, because the pro scene always finds flex value before the crowd. And patch 26.23, because if Riot has to hand power back to W, the initial compensation was priced too cautiously.
A line I keep returning to: curses do not exist, only data we have not finished reading. Mel was never cursed by an overpowered ability. She carried an ability the game's own measurement system had no scale to read. The number is the only thing on the pitch that speaks without needing to be cheered, and here the number has not yet spoken.
In three weeks we will have what the announcement lacked: real data. If the support pick rate rises as the design intends, then Riot did not delete a mechanic — they moved a champion into a different profession, and did so without asking the people who spent thousands of hours learning the old one.
The question I want to leave behind is not whether Mel is still strong. It is this: if a mechanic gets deleted because it is too hard to measure, what happens to the other mechanics that are equally hard to measure — and who decides that hard to measure means it must go?
