Riot Games Reworks Mel's W: Reflect Becomes Shield, and a Cohort of Players Loses Its Skill

The fourth shot of Jhin came back and killed Jhin. Riot Games picked that exact moment as its illustration when explaining why it is deleting Mel's W. Another example: Mel reflected Seraphine's ultimate, and the entire enemy team was charmed in a single beat. Those two clips were enough to make Mel the most frustrating champion in League of Legends for most players who had to face her.

The telling detail sits in one short sentence of the announcement: the design team tried to preserve the reflect mechanic across multiple rounds of adjustments before giving up. An ability abandoned after repeated tuning passes is not an ability that is too strong. It is an ability that cannot be scaled — either it reflects, or it does not.

What the patch changes

Patch 26.22, following Riot's season-based numbering, is expected on live servers roughly three weeks after the announcement. The scope does not stop at W.

The W no longer reflects. It still destroys incoming projectiles. The replacement payoff is a shield and movement speed, applied to Mel and to nearby allies.

The passive loses early-game damage and effect duration, compensated by health and armor scaling.

The Q redistributes damage: less front-loaded burst, more damage on subsequent hits.

The E reduces root duration at close range and increases it at maximum range.

The R gains cooldown and loses per-level scaling.

The report offers no figure whatsoever on Mel's win rate, pick rate or ban rate. That gap has to be logged before analysis begins, because every inference below must be downgraded one level of confidence.

Three weeks from announcement to live is enough runway for players to practise, but for competitive play the timing matters more than the length. If the patch lands inside a version lock for an international event, a rework-level change only reaches teams during the preseason — meaning the competitive shock is deferred and substantially reduced. The report gives no specific date, so this is inference, not fact.

The W changes job, not power

Read line by line, it looks like a nerf. Read as a kit, the story is different: the W changes function. The old version was a punishment tool — hold it, wait for the opponent's key spell, send it back doubled. The new version is a peel tool — block the hit and open an escape lane for the person beside you.

In Summoner Rift terms, that is the difference between a jungler waiting to counter-gank and a jungler warding ahead of the play. Same position, entirely different way of accounting for value.

The E change states the intent most clearly. Less root up close, more root at maximum range. This is a mechanic that pays for distance: reward for standing far, a tax for standing close. Mel is pushed behind the team. That posture is a support posture, not the posture of a mid-lane mage trading spell for spell.

Health and armor scaling on the passive points the same way. Armor on a mage passive is an unusual detail, and it serves surviving a dive rather than winning a trade. Combined with the shield and movement speed on W, the new kit reads as a genuine support option.

The R cooldown increase and per-level scaling cut mean more than they appear to. Mel's ultimate is a long-range engage tool. With a longer cooldown, the value of using R to squeeze an enemy formation drops, and Mel is forced to hold R for the real fight. She loses the ability to dictate map pressure with her ultimate. That is the kind of loss that never shows in a stat sheet but shows in how the rest of the team has to rotate.

The Q redistribution works the same way. Mel can no longer clear a wave quickly and pin her opponent under tower in the opening minutes. She has to play longer fights, more trades, and wait for items. All four changes together push her power curve toward the late game.

A cohort loses a skill, not a stat line

This is where I want to stop the longest.

Based on my experience tracking matches, I keep a habit of logging ward counts, gold differentials and objective timings before writing anything. With Mel, what I logged was how often the W was held and how often it was burned early. Good Mel players do not win by casting fast. They win by reading the opponent's cast rhythm — knowing Jhin is about to fire his fourth shot, knowing Seraphine is about to open her ultimate — and holding W for exactly one beat.

That skill now has nowhere to go. It was not weakened; it was removed from the game. A stat nerf leaves the old player with reflexes intact and a new rhythm to learn. A mechanic deletion turns all of that capital into nothing in a single evening.

In an empty theatre, people hear the breathing of pain more clearly.

And this is the easiest place to get the writing wrong. The storytelling instinct wants to turn this cohort into victims. But dignity is not excuse-making. The reflect mechanic produced situations where the opponent had no reasonable response — their spell was used to kill them, and no amount of skill could fix it. An ability whose outcome does not depend on the skill of the person being hit is an ability with a design problem, not a difficult ability.

Both things are true. The mechanic was flawed, and the players who lost it genuinely lost something they built.

Riot's stated goal is only half achieved

Riot says the aim is to limit situations that are too frustrating when playing against Mel. But the new W still destroys projectiles. The frustration of "my spell did nothing" remains intact. What was removed is the frustration of "my spell killed my teammate".

Those are different in kind. One comes from losing value, the other from being punished. Riot addressed the second — the louder one, the easier one to clip — and left the first alone.

The choice of examples is worth noting too. Jhin's fourth shot, Seraphine's ultimate: dramatic moments, easy to clip, easy to argue about. The report offers no frequency data on how often such moments occur in a game. Balancing on the salience of a clip rather than on actual frequency carries a risk of over-correction, and that risk cannot be verified until post-patch data exists.

There is also an internal contradiction. The announcement says the changes will keep Mel balanced while preserving her unique identity. But the unique identity named in that same announcement is the reflect mechanic. Preserving an identity by deleting it is a sentence that needs re-reading.

The case that it helps

There is a counter-reading, and I think it is the stronger one.

Losing mid-lane power while gaining support viability is not automatically a loss. A champion who can appear in two roles prevents the opponent from locking a composition early. In professional drafting, positional ambiguity is worth more than a few points of damage. Mel loses her strongest window — early and mid game — but if the shield and movement speed are good enough, she becomes the flexible support option teams want for game three and game four of a best-of-five.

There is no way to confirm this before the patch goes live. That is the limit of any analysis written from an announcement.

A gank reviewed in slow motion is worth more than a match praised in a hurry.

Some people see the future before it arrives; the future only nods in silence.

Takeaway

Mel will return to professional play — the only questions are where, and how soon. If she returns as a support, that sets a precedent: Riot can migrate a champion's identity rather than tune it. If she disappears, the precedent still forms, just in the other direction. For the players who spent hundreds of games learning to hold W for the right beat, what they need is not an apology from the developer. They need a reason to believe that the skill of reading a game is not something balanced like a stat line.