Mel's W Rework, Patch 26.22: What Changed When the Reflect Came Out, and Who Carries the Cost

It is 2:40 a.m. in Sharjah. A clip is paused on the screen — Jhin's fourth shot leaving the barrel like a red line, and Mel's W opening in front of it. The shot returns, enters Jhin's own chest, and he falls. That clip lives in my "First Eight Rounds" folder with a timestamp on it, because in that one second the price of a match changed: no spell, no item, just an ability that took another player's decision and mailed it back to him.

Opening the notes for patch 26.22*, the same ability is there — same slot, same icon — but the reflect is gone. In the Vietnamese client it used to be called "Phản Quang Chiết Kính," the prism of reflection. That name is no longer in the tooltip. In its place: a shield, some movement speed, and a radius that includes teammates.

*The patch number comes from the source. Under Riot's season-based scheme it places as a late-2026 patch, but I could not independently verify a launch date — and when the date is unverified, every claim about its tournament window is unverified too.

Context: what actually changed

Mel's W was the identity of a champion who took the weapon out of your hands and used it on you. According to the source, the new W does not reflect projectiles — it destroys them, returning a shield and movement speed to Mel and nearby allies. That is not a word-level difference. Reflect meant the opponent had to think twice before firing his most expensive ability at Mel. A shield means he can fire it, knowing the damage will be reduced. The first is fear; the second is arithmetic.

The rest of the kit got a numerical pass in the same patch. The passive loses early damage and duration but gains health and armour scaling. Q moves damage off the first explosion into subsequent hits, with a slower projectile. E becomes a two-way lever — a shorter root at close range, a longer one at maximum range. The ultimate's numbers come down.

Riot itself acknowledged that the reflect was Mel's "signature characteristic," and said it had tried to preserve it before pivoting. That sentence is the heaviest part of this story, because it says the decision was deliberate rather than reactive.

And the thing that matters most: nowhere in the announcement is there a single win-rate, pick-rate or ban-rate figure. When there is no number, an estimate of magnitude becomes a mood rather than a measurement.

Core: pruning the anti-fun, and who pays for it

Across years of watching VODs, one pattern keeps surfacing: Riot cuts mechanics that make the losing player feel cheated. Old Poppy, old Yorick, and now Mel's reflect belong to the same chapter. The principle is explicit — the experience of the player facing the champion outranks the attachment of the player piloting her. That is a governance decision, not a balance decision. In balance you tune numbers; in governance you delete the ability.

The two examples Riot cited are technically elegant. Seraphine's ultimate, when reflected, charmed Mel with her own crowd control; Jhin's fourth shot came back at him with the crit attached. Notice that both belong to the projectile class — an enchanter-mage's ultimate, a marksman's fourth shot. The reflect's heaviest casualties were never tanks or assassins; they were the champions who win on range and timing. After 26.22, Seraphine and Jhin — and in the same rhythm, Ezreal and Caitlyn — get back the breathing room that was taken from them. That is not a footnote. It will show up on the draft board.

Removing a signature is a transfer of identity

The weakest sentence in the whole announcement is the claim that the changes "preserve her identity." Removing the mechanic Riot itself calls signature and then calling the identity preserved is like banning a player's signature pick and insisting his style is intact. Identity does not live in a name; it lives in a decision pattern. Mel's decision pattern has changed. She used to say: fire, and I will return it. She now says: fire, and I will absorb it, and move my team forward. The first sentence was a duel. The second is a partnership. Both are interesting. They are not the same champion's sentence.

E is a spacing tax

The smartest part of this patch is E, and it is getting the least discussion. A shorter root up close, a longer root at maximum range. That is not a nerf; it is skill-indexing. Previously a Mel player could panic-cast E at melee range and collect nearly the full root. Now panic is punished and precision is paid. A player who pilots Mel by the centimetre is not badly served by this patch; a player who used her as a panic button is. Two skill tiers have opened inside one champion, and the gap comes from mechanics, not matchups.

Q redistribution: the death of guaranteed burst

Q moves damage from the first explosion to later hits, with a slower projectile. The translation is direct: damage that used to be guaranteed is now earned. A slower projectile demands tracking from the Mel player and hands the opponent a dodge window. It is the same logic as E — skill floor up, skill ceiling up. Both point the same way, and together they argue that this patch was not merely a numbers cut.

Passive health and armour scaling: a crack in the build path

This is where my suspicion is highest. Cutting early damage while raising health and armour scaling is not cosmetic — it quietly tilts the item path. If Mel's scaling leans into health and armour, a build route may open that Riot has not fully stress-tested. That is precisely where a second hotfix is born. The people saying she can become a tankier mid laner may be right — and if they are, 26.22 is not the final patch.

The tournament window: what the source does not contain

The source names no tournament, no team, no pro player, no region. That is a fact, and it should not be hidden. When a patch of this size lands near a major event, two patches start running in parallel — the practice patch and the tournament patch. If a rework-level change does not reach the tournament server, Mel's real competitive value will be known one cycle late. If it does reach, then every team that kept her as a situational counter-pick finds a preparation note gone stale overnight. Both outcomes cost something. Because the source names no event, I will not claim one — only that this is the question every analyst should ask and nobody in this announcement asked it.

Rework honeymoon, then the hotfix

Reworked champions have a predictable life: pick rate climbs for two weeks, win rate sags, a hotfix arrives, then another. Mel has every chance of following it, because this patch touches five abilities at once. Anyone writing "Mel is balanced now" is reviewing a patch whose live data does not yet exist. The first number was never the story; the silence after it was — and here the silence is that nobody has a line about 26.23 or 26.24.

Contrarian: two romances, both false

Two romantic readings are circulating, and I want to un-write both.

The first: "Riot finally fixed the most annoying champion." That is the relief frame, and it is data-free. No win rate, no ban rate in the announcement — only the sentiment that players were frustrated. The sentiment may be real. It is not a measurement.

The second: "Mel is dead, her identity erased." That is the grief frame, and it is equally data-free. In organised play the reflect was never Mel's win condition; it was a solo-queue irritant, where one mistimed ultimate killed you with your own damage. Anyone who has watched her in scrims knows her real weight sat in positioning and zone control. "Mel is finished" is as inflated as "Mel is fixed."

The real loss has no banner. It belongs to the player with eight hundred Mel games, whose fingers had memorised the reflect timing, whose rank rested on the muscle memory of that single instant. For him this is not a rework; it is a change of season in a career. I had to un-write the profile before I could see the person — and it applies here. The most honest reporting on this patch is not the piece that calls the removal a tragedy, nor the one that calls it relief. It is the piece that counts how many people now have to rebuild an identity.

Takeaway

Do not read patch 26.22 as a final document. The real document is written in 26.23 and 26.24, where we learn whether Riot's intent was clean or its estimate was wrong. Everyone forming a verdict today is calling a score in a match whose first minute has not started. Every fandom keeps a ghost version of the game it loved — Mel mains now have theirs: the Mel with the reflect. The question is not when the ghost returns. The question is who counts the people still playing while it is gone.

About the writer: Lim Tae-yang is an esports analyst based in Sharjah. He is the author of "The Bench" (2026) and the "First Eight Rounds" file, which tracks where more than 180 Indian players went after their rosters let them go.