Mel's W Rework: Reflect Out, Shield In — And the Column Patch 26.22 Leaves Blank

Two Forty-Seven in the Morning

The item reached my feed in Vietnamese localization roughly fifty minutes before the English version. The headline had three words — "Phản Quang Chiết Kính", Reflect Prism. Two forty-seven at night by my clock. I took a screenshot, because nine years of habit teaches you one thing: the language of patch notes changes, the line-item count does not.

By morning the picture was clean. Mel's W — the ability that let her catch an enemy projectile and turn it back on its owner — was gone in Patch 26.22. In its place: a shield, and movement speed for herself and nearby allies. Alongside it, hands on passive, Q, E and R in a single pass.

The champion Riot itself called "signature" had her signature cut in the same patch. And what stung most wasn't the W. It was the column that appeared nowhere in the story. No win rate. No pick rate. No ban rate.

The Rules of Patch-Note Journalism

Patch notes are the one discipline where the primary document sits in your own hand. In the transfer market I need two documents — a contract, a release letter, or a signed termination. Here the document was published by Riot itself. For "what changed" I need no second source. For "why it changed" and "by how much" I need one, and those are precisely the two places where Patch 26.22 stands empty.

The numbering follows Riot's season-based scheme, so 26.22 is a 2026-season patch, landing on live servers roughly three weeks after the announcement. Three weeks means three weeks of noise — PBE clips, "new Mel is OP" thumbnails, and every large creator's first-impressions video, all before anyone owns a single real match of data.

The source is a Vietnamese-language localization, and that detail matters in a small way. It means only that the global patch notes were translated for the Vietnamese market — this is not a regional meta story. Regional meta requires regions. A champion-level balance patch ships on the same server at the same time everywhere. There is no team here, no pro player, no tournament. Which means the team-fit analysis I normally do is unavailable to me this time, and I won't fill it with something else.

What Mel Actually Was

Hold one scene in your head. Seraphine is winding up her ultimate — the wave that charms an entire team. Mel presses W. The wave comes back, and now the charm lands on Seraphine's own side. Second scene: Jhin's fourth shot, the crit one that deletes a squishy in a single bullet. Mel presses W. The bullet returns, crit included.

That was her entire business. She wasn't picked for damage; she was picked for denial. You took her when the enemy draft was projectile-heavy — Seraphine, Jhin, and the same class of poke marksman like Ezreal and Caitlyn. Mel was a counter-pick whose value was to hand the enemy's most expensive asset, their coordinated burst, back to them for free. In draft terms the price was direct: the opponent had to abandon their most reliable plan.

Riot has admitted, in writing, that the reflect was her signature. That sentence is a studio concession, and it is the most honest line in this whole story. Everything else comes after it.

Reflect belongs to the ability class designers call "anti-fun" — the mechanic that kills not through its own strength but by turning an opponent's best moment back on them. Riot has a long habit of pruning such mechanics even at the cost of champion identity. The old Poppy and old Yorick reworks sit on that list. That is context for the decision, not a defence of it.

The Ledger: Thirteen Line Items, Five Slots

I went back to the ledger. Thirteen line items, five slots. Not one of them carried a win rate beside it.

W lost the reflect. In its place: projectile destruction, a shield, and movement speed for herself and nearby allies. Passive: early damage down, duration down, health scaling up, armour scaling up. Q: damage shifted off the first explosion onto subsequent hits, projectile speed down. E: root shorter at close range, longer at max range. R: nerfed.

Four readings follow, and all four point the same direction.

One: damage down, utility up. Passive, Q and R all lost damage. W lost the reflect. In exchange came a shield, movement speed, and health-armour scaling on the passive. Raw power points down. Role points sideways. Mel is sliding from burst counter-pick toward sustain/enable mid. Calling that number-tuning is easy; it is an identity transfer.

Two: Q is a skill tax. Moving damage off the first explosion onto later hits means guaranteed burst falls and the value of tracking and aiming rises. A slower projectile means more room to dodge. Both do the same job — raise the skill floor. A player who treated Q as press-and-forget has just lost damage out of his hands.

Three: E is a two-sided lever, not a straight nerf. Shortening the root at close range punishes panic engages. Lengthening it at max range rewards precise spacing. This is reward-precision-punish-panic design. Calling it a nerf is easy, but what it actually does is install a filter inside the Mel player base — those with spacing climb, those without fall.

Four, the least visible: the passive's health-armour scaling is a build-direction signal. Cutting early damage and duration while raising health and armour scaling means Riot is pushing Mel toward a tankier item path. That is structural, not cosmetic. If the path fits in testing, it needs retuning next patch; if it doesn't, Mel has no reliable build left at all.

Then the line Riot wrote itself and nobody counted: the studio tried to preserve the reflect, and failed. It's a short sentence in the notes, and it carries the most weight. It means this was not a reactive hotfix. It was a considered design decision preceded by several continuous adjustment rounds that failed, after which the mechanic was cut. Mechanic removal is always the last resort. When a studio cuts its own creation, there is a metric behind it that never gets shown.

The Column That's Blank

No win rate. No pick rate. No ban rate. The Patch 26.22 notes in front of me contain not one figure on how strong Mel was in the previous patch. That is not a small omission.

Consider what I need in a transfer story: the signing date, the contract status, the buyout, and the single fact that would kill the whole tale. That fourth fact is missing here. If Mel's ban rate was forty percent, removing the reflect was justified. If it was twelve percent and the rest was forum noise, Riot dropped a champion's identity in front of noise. Both versions of reality look identical in this note. The reader has no instrument to separate them.

I know why the figure isn't published. Riot doesn't publish champion-level data. Fine. Then it should also be written that "less frustration" is a claim, not a measurement. The only evidence in the story is the sentence that many players felt frustrated facing Mel. That is a feeling, not a figure. A feeling can justify a patch; a feeling cannot establish that the patch is correct.

And here I have to rein in my own habit. I don't report a transfer I haven't watched live — a scrim block, a LAN, a trial. The consequence of that rule is that I will not write "Mel is weak now" or "Mel is strong now" until someone has played her on live servers for three days. Anyone forming a verdict off the notes alone is standing outside the room.

One Line from Shanghai, Then Forward

I was in that room in Shanghai. The stream saw a scoreline; I saw a room go quiet. Since that night I've kept one rule — don't put into a claim what the screen didn't show. Translated to patch notes: don't write that a champion is "balanced" until someone has played her live for three days.

The Contrarian Angle: "Her Identity Survived"

The weakest line in this patch isn't the one Riot wrote. It's the one the surrounding coverage added — that after the changes, Mel is still Mel.

That sentence doesn't hold. Riot wrote itself that the reflect was her signature. Removing an ability the studio calls a signature isn't preservation of identity; it's reconstruction of it. Shields and movement speed are good tools, but they were never Mel's tools. They could belong to any enabling mage. A champion's identity isn't built from how many tools she has. It's built from which tool no one else can give her.

The second contradiction is larger, and it sits in the dead centre of the official story.

The official story is that removing the reflect makes the game less frustrating. That claim is about solo queue. But what the new kit adds — a shield, and movement speed for nearby allies — is exactly the part of a kit whose value rises with coordination. In solo queue nobody reliable stands beside you; in a five-stack or a pro game, granting movement speed to three allies at once is a different commodity entirely.

So the patch that promised to reduce solo-queue frustration may be the same patch that makes Mel more valuable in coordinated play. That isn't a prediction, it's kit logic: aura-based utility always returns more in team settings. And if it does, the same problem Riot is solving today returns in six months — not in solo queue this time, but on a Spring Split draft board.

That is the blind spot in the frame that says "anti-fun mechanic removed, meta got healthier." The frame assumes solo-queue play and pro play are the same place. They are not. The same kit is worth different amounts in each, and Riot has shown its arithmetic for only one of them.

The Next Domino

My eyes are on the first seventy-two hours of data, and the order matters: ban rate first, win rate second. Ban rate is the first metric that will tell us whether players feared Mel or avoided her. If ban rate doesn't fall after the reflect is gone, then the fear was never about the reflect.

The second thing I'll watch: whether Patch 26.23 trims the shield or the movement-speed aura. If it does, the direct reading is that 26.22's testing missed that part. Third: a pick-rate bump for projectile champions like Seraphine and Jhin. If the fear of Mel is gone, that bump should arrive; if it doesn't, we have to accept that the reflect's share of draft planning against Mel was never as large as everyone assumed.

Fourth, and most interesting: whether some form of the reflect returns. Riot has a habit — after an identity removal, if the data collapses, half the road gets walked back. I'm not writing that as a prophecy. I'm writing it as a question whose answer arrives in three weeks, and which I don't know either.

And when all of this settles, what remains is one unnamed player. The Mel one-trick who spent two seasons on that single pick, who built an entire ladder strategy on that reflect, and who now has no route through the next twelve months except learning a new champion from zero. Her name appears nowhere in the patch notes. No name appears anywhere in the patch notes — no team, no pro player, no tournament. Which is why I didn't fill one in. Inventing a name to fill a blank column in a ledger is the easiest job available, and the worst mistake.