Mel's W Rework in Patch 26.22: Reflect Is Gone, a Shield Arrives, and the Data the Source Never Showed

One line changed in a patch note. In the Vietnamese-language localisation, the ability is called Phản Quang Chiết Kính — Reflect Prism. The old tooltip said, plainly, that an incoming projectile would be bent back at its caster. The new tooltip does not contain that sentence at all. In its place: the projectile is destroyed, Mel gains a shield, and she and nearby allies get movement speed.

For me this is a human event before it is a patch note. Playing out of Kuala Lumpur on Southeast Asian ranked ladders, I have watched players whose entire climb was built on that one ability — the enemy ultimate, the enemy crit shot, the enemy crowd control, all sent home. For that player this is the same kind of cut as a batter losing his front-foot drive.

Here is the first asterisk, and it matters. The source names no professional player, no team, no tournament. There is no win rate, no pick rate, no ban rate. What exists is a design decision and an official rationale: reducing player frustration. I keep the asterisk because the bronze never told the whole story — and here there is not even a bronze, only a rewritten tooltip.

What Mel was, and why the reflect became her identity

Mel arrived in early 2026 as a mid-lane mage, and her W was sold from day one as a counter-projectile identity. The solo-queue effect was direct: projectile-dependent champions showed up, and their most expensive coordinated play — an ultimate or a crit — returned to land on them instead. Riot's own patch note cites two examples: Seraphine's ultimate charming its caster, and Jhin's fourth shot critting its way back. Those are precisely the classes that now get breathing room.

Riot calls the reflect her signature characteristic, which is the most interesting sentence in the entire item. A publisher is stating, in its own words, that the thing which defined the champion is being removed — and that it tried to preserve it, ran multiple adjustment rounds, and then gave up. That detail separates this from a reactive hotfix. This is a considered removal.

The patch is numbered 26.22 as the source states. Read against Riot's season-based scheme, that places it in 2026, launching roughly three weeks after the announcement. I am not calling the numbering verified; what the source says is what I am working with.

The shape of the source is also information. This is a Vietnamese localisation of global patch notes — LMHT, Reflect Prism — not a team or regional report. A VCT Champions 2026 link sits alongside it, which only tells us the outlet covers multiple titles. On what this does to mid-lane priority in professional play, the source says nothing. Everything past this point is my reading, not the outlet's claim.

Ability by ability: one mechanic gone, four dials turned

W is where the earthquake is. Reflect meant the opponent's own investment came back at them. The new version stops and destroys the projectile and returns a shield plus movement speed — movement speed that extends to nearby allies. Notice that the direction has flipped twice: the attacker is no longer punished, and Mel gains something she never had, which is team utility.

This is an old Riot instinct. Mechanics that make the person on the other side feel cheated get pruned — the old Poppy and old Yorick reworks came from the same philosophy. Reflect sat squarely in that category. The interesting question is not whether Riot did something unusual; it is what Riot handed back while cutting.

The passive loses early damage and duration but gains health and armour scaling. That is not cosmetic. When a mage's scaling tilts toward health and armour, the build tilts with it: the full-AP burst path narrows and a tankier off-meta path widens. I suspect that path was not fully stress-tested.

Q is the cleverest change and the easiest to skim past. Damage moves out of the first explosion and into subsequent hits, and the projectile slows down. In plain terms: guaranteed burst down, the price of tracking and aiming up. That is not a nerf, it is a skill-expression tax — the cheap route to damage has been made expensive.

E carries a two-sided lever as well. Shorter root at close range, longer root at maximum range. That rewards spacing and punishes the panic close-range engage. For any champion-specialist pool this is a skill filter, and skill filters are usually the least-discussed line in a rework note.

The ultimate takes a numerical cut. It is the change the source says the least about, and possibly the one that will matter most — because Mel's value as a counter-pick was built on kill pressure, and kill pressure is built on the ultimate.

Stack all four together. Raw damage down across passive, Q and ultimate. Utility and durability up through shield, movement-speed aura, and health-armour scaling. The result is that Mel moves out of the burst counter-pick slot and toward sustain or enable. That is not number tuning. That is a role-identity shift, which is why this has to be called a rework rather than a balance change.

The draft economy: who gets to breathe

When a champion loses a counter-pick tool, the effect does not stay with that champion. It lands on the price of everything the tool used to scare. Seraphine, Jhin, and behind them Ezreal and Caitlyn style poke-crit marksmen all sit in the class whose most expensive coordinated play had a lock on it. With the lock gone, their draft priority ticks up. The change is not large enough to measure cleanly, but it is real.

On the other side, teams that kept Mel in mind only as a situational counter now have an empty slot. One name has dropped off the counter-pick board — which, in scouting language, is champion-pool devaluation, the same shape as a player's reflex curve falling off a cliff.

This is where the data is missing

There is not a single figure in the source. No win rate, no pick rate, no ban rate. I do not find that strange; it is the normal condition of localised patch news. But it has a hard consequence: it is impossible to say how much less frustrating the game actually becomes when a reflect becomes a shield. Direction can be inferred. Magnitude cannot. Anyone concluding today that Mel is now balanced is holding an expectation with no number under it.

From nine years of watching this game, one pattern holds: multi-ability reworks almost never land correctly on the first pass. They cut too much or too little. And when a signature mechanic is removed, the first patch usually spends its time waiting for the second. Treating 26.22 as final is premature.

The contrarian angle: a skill check cuts both ways

The weakest claim in the coverage around this patch is that Mel's identity has been preserved. Removing the ability Riot itself calls the signature characteristic does not preserve identity; it replaces it. A throwing mage is now a saving mage.

The second argument is more uncomfortable. Reflect is remembered as Mel's ability, but functionally it was a knowledge check on the player facing her. Against Mel, you had to think about when to fire the ultimate, when to send Jhin's fourth shot, when to throw Seraphine's charm. That thinking was part of the game, and it was a measure of the opponent's skill. Removing reflect does not make Mel easier; it makes playing against Mel easier. The skill floor drops on one side of the lane. Some portion of what Riot calls frustration was actually difficulty, and removing difficulty is not automatically an improvement.

Third, the direction risk. A shield plus a movement-speed aura pulls a mid mage toward support. If Mel's power migrates into enabling and does not survive in burst, the worst outcome is not that she is too strong. The worst outcome is that she is nothing — neither fish nor fowl. That is the most common accident of a rework: the anti-fun mechanic goes, and the champion's gravity goes with it.

Fourth, timing. The source says the patch arrives roughly three weeks after the announcement. If that window lands just before or during a major event, a practice-patch versus tournament-patch divergence opens, and a rework-level change is most expensive inside that gap. No event is named in the source, so I am flagging this as a warning rather than a claim.

What I will be watching

Scouting is archaeology with a stopwatch: you date the talent by the traces. The next three patches will leave Mel's traces, and I will be logging three things.

First, whether the health-armour build shows up. If it does, her carry ceiling falls and her enabling floor rises, and she stops being a solo-queue specialist and becomes a coordinated-team asset. Second, whether she returns to the draft board or slides off it into a side option. Third, and most interesting, whether Riot ends up restoring a partial version of the reflect. Publishers are comfortable removing identity and equally comfortable folding when the numbers turn bad.

Page 96 still has a margin wide enough for the truth. The question was never whether Riot reduced frustration. The question is whose identity is being saved when the least-liked ability on a champion was the identity itself — the player's, or the experience of playing against her?