Riot reworks Mel's W in League of Legends patch 26.22: reflect removed, ally shield added

At 2:47 in the morning on the Korean server, in a solo queue game, I counted eleven seconds.

A Mel player stood behind the outer turret, holding her W, not casting it, not cancelling it, simply waiting. Across the lane, Jhin was holding his fourth shot — the bullet he had saved through fifteen seconds of trading and did not dare fire, because he knew the reflect frame was still up. Eleven seconds of silence. Then Jhin fired. The fourth shot bounced back and killed Jhin.

I rewatched that clip four times, not because it was good, but because it was strange. Nobody won in those eleven seconds. Mel did not attack. Jhin did not attack. Both waited for the other to make a mistake, and the one who made it was the one forced to act first. That state is what the League community calls having your hands locked — and it is exactly what Riot Games has now decided to remove from the game.

This week Riot announced it will rework Mel's W in patch 26.22, expected to ship roughly three weeks after the announcement. The projectile-reflect mechanic — described by the design team itself as this champion's signature trait — is being removed entirely. In its place, W will destroy incoming projectiles and grant a shield plus movement speed to nearby allies.

I sat with that announcement for a long time. In nineteen years covering this industry I have watched Riot delete a lot of things. This one is different in one respect: they did not hide it. They said plainly that they had tried to save the reflect mechanic across multiple rounds of adjustments, and eventually gave up.

Why the balance team gave up

Riot cited two examples in the announcement. First, Mel reflecting Seraphine's ultimate and charming the entire enemy team with their own spell. Second, Mel reflecting Jhin's fourth shot and returning the critical strike into Jhin's face.

Those two examples were chosen because they are striking. They are also the most beautiful examples, not the most common ones. Not a single line of data in the announcement says how often such swing-the-game reflects happen per match, at which rank, or in which league. I have no right to guess on Riot's behalf. I do have the right to note that the absence of frequency data is the structural weakness of every balance decision driven by a viral clip.

There is a technical reason this mechanic could not survive tuning passes. A normal ability has many continuous dials: damage, cooldown, duration, radius, per-level scaling. The balance team can turn each dial by one percent and watch how players respond.

Reflect has no dials. It is a switch.

Either W swallows the whole hit and returns it, or it swallows nothing. There is no intermediate state in which designers can cool it down gradually. When a mechanic has only two states, every balancing effort becomes an attempt to move it between the useless extreme and the broken one. That is the practical reason people try several times and then walk away.

I once wrote about a similar mechanic in another sport, and the lesson repeated itself exactly: what cannot be measured cannot be tuned. A coach once told me he was not afraid of a player who plays badly, he was afraid of a player who produces moments that cannot be repeated. What cannot be repeated cannot be coached, and cannot be put into a system. Mel's reflect sits squarely in that category.

Two kinds of frustration, and only one is being removed

Riot's announcement sets a clear goal: to limit situations that are too frustrating when playing against this champion.

That goal is only half achieved.

The new W still destroys projectiles. Which means Seraphine's ultimate still vanishes mid-air, still deals no damage, still does nothing at all. The structural frustration — my spell left my hands and evaporated, I spent a major resource and got no feedback — remains intact in every detail. What is being removed is only the dramatic frustration: my spell came back and killed my own teammate.

Riot is fixing the frustration that gets clipped, not the frustration that happens every minute of the game.

This matters because it predicts the community reaction once the patch goes live. Players who were annoyed at being reflected will feel relief. Players who were annoyed at having their spells swallowed will notice no change at all, and may notice it more, because Mel now also shields her whole team while negating projectiles. The cost of the new ally protection is paid with the old annoyance.

This is the kind of compromise I see constantly in large design decisions: people address the loud part of a problem first, because the loud part creates media pressure, while the quiet part creates pressure nobody clips.

The compensation package: shield, armour, and a re-tuned E

W is not the only change. The whole kit was touched.

The passive loses early-game damage and duration but gains scaling from health and armour. A mage with armour scaling is unusual, and the direction is clear: survive being dived on, rather than win duels.

Q was redistributed. Damage that used to be front-loaded now spreads into later hits. For players who trade with one Q and walk away, that is a direct loss. For players who fight longer, it is a refund.

E says the most. Root duration is reduced at close range and increased at maximum range. Riot is not hiding the intent: they are paying for spacing discipline and taxing close-range self-peel. That is a mechanical incentive that is hard to disguise — it pushes Mel further back in fights, and the posture of standing far, holding position and protecting others is the posture of a support, not of a solo-lane duelist.

R gains cooldown and loses per-level scaling. Combined with the weaker early passive, Mel loses precisely her strongest window: the early and mid-game skirmish spike. If the champion is to keep a place on a roster, her value must migrate to the late game, where a well-timed shield and a well-placed burst of movement speed can decide more than one instance of damage.

Whether that compensation is enough, I cannot say. Nobody can say before the patch has been live for a few weeks. And I will not write a verdict based on a feeling.

The biggest losers are not Mel players

Over nineteen years I have developed a way of sorting patches. There are two kinds. The first is a numbers pass — players are annoyed for three days and then adapt, because their skill is still intact, they simply have to press harder or calculate differently. The second is a skill deletion — and that kind does not create better players, it only turns players who were already good into beginners.

Mel's W is the second kind.

This champion's competitive value rested on a read skill that is very hard to teach: deciding how long to hold W, which moment to wait for, how much health to trade away to save the frame for the genuinely dangerous hit. That is a skill that does not appear on a scoreboard, cannot be measured in damage per minute, and cannot be learned from a guide.

Dedicated Mel players lose all of that capital in one patch. No version of the new W requires that read, because the new W does not need to wait for the right hit — it only needs to be pressed when someone dives. This is the harshest class of change a balance team can ship: not a power reduction, but a devaluation of accumulated mastery.

Spectators watch the scoreline. I watch how they tie their boots before the ball rolls. I wrote that line back when I was following a football club in Incheon, but it holds for League too: what people see in a reflect is the moment, and what is being taken away is thousands of hours of reading situations that nobody ever recorded.

Who gains, and what they gain

The clearest beneficiaries are projectile-reliant champions who used to be punished by their own reflected damage. Jhin with his fourth shot is the example Riot named. Wave-based mage ultimates like Seraphine's are the second. For them, W no longer returning fire means a risk has been deleted from their calculations.

The second group is melee divers. Diving a Mel who still held W used to be a gamble that could cost your life to your own damage. Now the dive is only blocked, not returned. For an assassin, the difference between being blocked and being killed by your own hit is the difference between a failed play and a death.

The third group is mentioned least: Mel's teammates. A shield and movement speed for nearby allies is something the old W simply did not have. It is the only part of the package that creates genuinely new value rather than redistributing old value.

Beyond Mel players, the losers are the teams that used her as a counter-pick into projectile-heavy compositions. That card no longer counters the way it did.

Support Mel, and why that is not absurd

People remember the goal. I remember the substitute clapping for his teammates. In League, people remember the reflected ultimate. I remember the unnoticed W holds, where the player simply stood in the right place and pressed nothing at all.

The role shift is not a side effect. A shield plus movement speed for allies is the signature of an entire support class. E pays for keeping distance. The passive adds armour to survive dives. Those three changes add up to a complete playing posture, and that posture is not a mid-lane posture.

For competitive play this may be better news than it looks. A mid-support flex champion has unusually high draft value, because it lets a composition delay revealing its roles until the opponent has committed. Losing solo-lane power while gaining role ambiguity can be net-positive inside a draft room.

I am not saying Mel becomes a premier support. I am saying that if she reappears in competitive play, the higher probability is at support rather than mid. That is an inference from kit structure, not from data — and I am labelling it as such.

The misunderstanding from outside

The most common reading after the announcement is: Riot nerfed Mel because she was too strong.

That reading is wrong at the premise. Mel was not touched because her power numbers crossed a threshold. She was touched because one mechanic in her kit cannot be measured, cannot be tuned, and produced moments the design team could not control. The fact that Riot tried multiple adjustment rounds before giving up is evidence for this reading. People do not discard a signature mechanic after repeated failures if the mechanic is merely a little overtuned on damage.

The second common reading is equally popular: Riot is deleting skill expression to please weaker players.

That reading ignores a detail inside the announcement itself. Riot says the change will make Mel more balanced while still preserving her unique identity. But the unique identity they described a paragraph earlier was the reflect mechanic. Those two sentences sit next to each other in the same announcement and cannot both be true. Mel's unique identity has been removed, and what remains is a kit that is better for the team, not an identity.

The counter-intuitive point I want to put on the table is this: what made Mel frustrating was never the reflect. The reflect was only its most visible expression. The real frustration was that Mel imposed her own rule set on the opponent — to attack her you had to accept that your own damage might come back, and that rule set came with no opt-out. Even the best player in the world cannot play better by shooting less. A rule set where the other side has no optimal move is a broken rule set, regardless of win rate.

And if that argument holds, then keeping projectile destruction in the new W leaves the original problem in place, wearing a different coat.

What I do not have, and why I am writing anyway

I do not have Mel's win rate before and after the patch. I do not have pick rate, ban rate, or presence at any tournament. The announcement provides no such figure, and I have no intention of inventing them out of feeling.

Mispronouncing one word taught me that I had understood nothing about that football culture. That lesson followed me into this field. Once I misspelled an ability name in a bulletin sent out at midnight and had to correct it before my Korean colleague could read it. Since then I keep a private notebook for every ability, every mechanic, every timestamp — not to show off detail, but so I never have to retract a sentence.

Based on my experience watching matches across many seasons, a rework-level change like this typically needs three to five weeks for data to stabilise, and another cycle for players to find the optimal approach the design team did not anticipate. Writing a verdict before that point is guessing.

My job is to keep the beat so others can walk in step. The beat here is recording exactly what changed, what did not, and what nobody can measure yet.

The big-season calendar and the timing question

Patch 26.22 lands late in the season. If the calendar runs as it does every year, the version lock for the largest international event falls before the release date, which means this change cannot affect the tournament currently underway.

That is good news for teams. They get a full preseason to experiment, and a rework-level change absorbed during preseason is always cleaner than one dropped mid-split. Teams rebuild their mid-support flex charts using an entire offseason of scrim data, rather than reacting hurriedly between match weeks.

But there is one variable I cannot verify from this source: the exact release date. Roughly three weeks after the announcement is the only figure I have, and it is enough to discuss a trend, not enough to name a day.

Internal signals to watch

Three things I will write in my notebook over the next six weeks.

First, where Mel appears first after the patch. If she shows up mid first, the compensation package was sufficient. If she shows up as a support first, then Riot has reworked a mid-lane champion into a support without saying so.

Second, the reaction of the players who used to be reflected. If they are still frustrated, the root problem has not been touched, and we will see a second rework within a year.

Third, how professional teams handle the old counter-pick card. When a counter-pick disappears, the teams that relied on it must find another answer, and that gap usually shows before the win-rate table does.

Six months I buried a story because nobody was ready to hear it. This one does not need burying, but it does need time. I write slowly. Because I believe the ball never needs to be rushed.