Patch 26.22: Riot Deleted Mel's Reflect, and the Changelog Still Has No Number in It

Seraphine's ultimate charms. Against Mel, that charm used to come back — the caster got charmed instead. Jhin's fourth shot crits; Mel's W used to send that crit home, and the damage landed on Jhin's own health bar. In League of Legends this was one of the very few mechanics where the attacker had to pay for their own damage. In Patch 26.22, Riot Games deleted exactly that mechanic.

The same W now grants a shield and movement speed to Mel and nearby allies. The reflect is gone, the shield is in. What did not arrive with the announcement is a number. Mel's win rate, pick rate, ban rate — not one figure is recorded in the patch note. A button that Riot itself called the champion's "signature characteristic" was cut, and no public baseline was offered for the cut.

I have a habit of walking into a story with a stopwatch. In 2026 I put forty PUBG players and forty university sprinters through the same reaction test, and the story changed the moment milliseconds moved the argument. This time there is no stopwatch on offer. So this piece is about what the changelog says, and about what it deliberately or carelessly leaves out.

Context: what this patch is, and which language it arrived in

Mel is a mid-lane mage in League of Legends. Her draft value over the last three years was not built on damage — it was built on her ability to stand in front of projectiles. Put a Jhin, a Seraphine, an Ezreal opposite her and picking Mel became a statement: your strength returns to you.

Patch 26.22 cancels that statement. It is set to hit live servers roughly three weeks after the announcement. The form the source arrived in matters too — in Vietnamese, with terms like "LMHT" and "Phản Quang Chiết Kính". This is a localisation of Riot's global patch notes, not a regional competitive report. No team, no tournament, no pro player, no prize pool.

One phrase keeps returning in the note: "continuous adjustment rounds." That small phrase is the biggest signal in the document. It means earlier hotfix and mini-nerf cycles already ran on Mel and failed to stabilise her. In balance history, that is the classic precondition for removing a mechanic — turn the numbers first, and if the numbers fail, cut the mechanic. Riot turned the numbers, then cut the mechanic.

Core: the real change is not inside W, it is behind it

The W diff is simple. Before: an incoming projectile was destroyed and its effect returned. Now: the projectile is not destroyed, and Mel plus nearby allies gain a shield and movement speed. This is not a nerf, it is a role shift — from burst counter-pick to sustain/enable mid.

The rest of the kit pushes in the same direction. The passive loses early damage and duration but gains health and armour scaling, quietly bending the build path toward health/armour stacking. Q moves damage off the first explosion and into subsequent hits while slowing the projectile. That is a skill tax on guaranteed burst: tracking and aim gain value, the safety net loses it. E is a two-sided lever — shorter root at close range, longer root at max range. Not a straight nerf, a reward for spacing. The ultimate was nerfed.

Add it up and the net equation is clean: raw damage down, utility and durability up.

Who gains? The champions whose entire game rests on projectiles. Seraphine and Jhin are the two examples Riot itself cited — the reflected ultimate charm, the reflected fourth-shot crit. Behind them sit the poke-crit marksman class, Ezreal and Caitlyn included. Teams that had to carry the fear of Mel in draft have one ban slot freed.

Who loses? Mel specialists and one-tricks first of all. The reflect was never just a mechanic; it was proof of an outplay — "I returned your best shot to your own face." Once that sentence is gone, a comfort pick stops being comfortable.

This is where provenance matters, and it cannot be dodged. Every claim above traces to Riot's own patch note, arriving in a localised form. The note carries no win rate, no pick rate, no ban rate. Which means the question "how strong was Mel" is unanswered, and the entire decision rests on that unanswered question. The figure is unaudited — in fact, the figure does not exist.

Contrarian: "her identity was preserved" is the weakest sentence in the whole note

The framing keeps returning to one line — Mel is still Mel, her identity is intact. I do not buy it, and Riot itself supplied the evidence. Riot wrote that the reflect was her signature characteristic, and wrote that they tried to preserve it and failed. The thing that was the signature is gone, so the identity did not survive — it changed. Calling that preservation is the most optimistic line in the rework.

A second angle sits on the argument's foundation. "Reducing player frustration" is an emotion-based justification, not a metric-based one. Ban rate tells you people did not want to play against her; win rate tells you she was winning. Those are different things, and neither was published. Without ban rate, "frustration" is a story, not a measurement. A balance patch is a changelog too, and we keep skipping its most important page.

A third angle is about power. Riot is the rule-maker, the owner and the judge of this game at once. It acknowledged a mechanic was an identity and removed it anyway. That is not a crisis story; it is an established governance style: the opponent's experience is ranked above the player's attachment. When India banned PUBG Mobile on September 2, 2026, I learned the same lesson — rules change first, and the player finds out last. In esports this is publisher governance in miniature, and it will repeat.

A fourth angle gets the least ink: rework churn. Champions that get a simultaneous operation on passive, Q, E, R and W almost never land correctly on the first pass. My estimate is that 26.22 is not the final state, and 26.23 or 26.24 brings an adjustment. That is an estimate, not a claim.

Takeaway: what to watch

Three things. First, the ban rate two to three weeks after launch — if it falls, Riot's framing holds; if the win rate collapses, a partial return of the reflect is possible, because Riot has a record of walking back identity removals. Second, the pro patch lock — tournament servers usually run one patch cycle behind, so the stage impact will lag the live-server impact. Third, and largest: whether Riot turns this cut into a template for future reworks.

I have coded eighty-three matches played without a crowd, and they taught me one thing — the noise always lives in the reaction, while the data stands quietly to one side. There is plenty of noise around this patch. The data has not arrived yet.