Mel loses Reflect in Patch 26.22: Riot deletes the most frustrating ability in League of Legends but leaves another frustration behind

Jhin's fourth shot flew back and killed Jhin. In a game I had to rewatch four times just to be sure I was not seeing things, the blue-side marksman fired his last critical round into Mel, and W — Reflect — returned the damage, unchanged, to its exact source. Jhin died to the strongest bullet in his own kit. A different version of the same story: Seraphine casts her ultimate, Mel reflects it, and the entire enemy team turns and attacks their own teammates. Those clips live a long time online — cut, slowed down, scored with music. This week, Riot Games announced it is removing that mechanic from League of Legends, replacing it with an ability that destroys projectiles and grants a shield plus movement speed to nearby allies. The ability Riot itself called "the most frustrating in the game" will no longer exist. What was deleted is the loudest part of the frustration. The quiet part is untouched.

Context: one ability, several tuning passes, one abandonment

Mel entered League of Legends with a kit built around the idea of retaliation. Her W did not block projectiles — it caught them, held them, and fired them back at the caster with effects intact. Against champions who live on projectiles such as Jhin, Seraphine, Xerath or Zoe, Mel turned their own damage into a liability for them. Riot described this as the champion's "signature trait," and the design team stated explicitly that it went through multiple rounds of adjustments to preserve the mechanic before abandoning it entirely.

The rework in patch 26.22 does not stop at the W slot. It is a full-kit tuning pass. The new W destroys projectiles instead of reflecting them, and additionally grants a shield and movement speed to nearby allies. The passive loses early-game damage and duration but gains health and armor scaling. Q's front-loaded burst is redistributed into subsequent hits. E's root duration is reduced at close range and increased at maximum range. R's cooldown goes up and its per-level scaling goes down. The entire package hits the live servers roughly three weeks after the announcement.

One note on scope: this is a live-server balance report, not a tournament report. The announcement names no team, no player, no tournament, and supplies no win-rate, pick-rate or ban-rate figures whatsoever. Every conclusion below therefore has to be downgraded one notch in confidence — including the ones I feel fairly good about.

Riot did not cut numbers. Riot changed the champion's profession.

The old W was a punish tool: hold it, wait for the exact moment, and turn the opponent's ultimate into your own weapon. The new W is a peel tool: destroy projectiles, hand out a shield, hand out movement speed to allies standing nearby. Same key, two functions that belong to different categories. Mel leaves the proactive damage-denial mage class and enters the utility-mage class, with a genuine flex into the support role. This is an identity migration, not a nerf. A nerf leaves the old playerbase in possession of the champion, needing only to recalculate damage. An identity migration takes away the reason to pick the champion at all.

The telling detail is that Riot kept the projectile destruction. Enemy spells still vanish when they fly into Mel. The feeling of "I pressed my button and nothing happened" — the single largest source of frustration against any projectile-denial champion — survives intact. What was removed is only the feeling of "my button killed my own teammate." Riot targeted the spectacular frustration and left the structural one in place. Those are not the same thing, and only one of them was addressed.

Riot's own examples deserve scrutiny too. Reflecting Seraphine's ultimate to charm an entire enemy team, reflecting Jhin's fourth shot back into Jhin — these are beautiful, rare, and extremely shareable moments. The announcement offers no figure for how often such moments occur per game, or what share of games contain at least one reflect swing that flips a fight. When balance decisions are driven by how viral a clip is rather than how frequent the event is, the systematic error always leans toward over-correction. Based on what I have logged watching games on the North American server, most reflects I have seen changed a laning trade, not a game. But the ones that changed a game are the ones that get posted.

What was deleted is not a number. It is a skill.

Whether Mel reflects successfully depends on reading the opponent: knowing when Seraphine will cast R, knowing Jhin has exactly one bullet left in the chamber, knowing the enemy is waiting for you to mistime it. That is a reactive skill that can be practiced, measured and taught. Removing the reflect mechanic does not reduce the value of that skill — it makes the skill meaningless. Dedicated Mel players lose their entire accumulated capital, not a slice of their power. Against a numeric nerf, a strong player keeps a relative edge. Against a deleted mechanic, a strong player is reset to the same starting line as a beginner.

The E change is the clearest directional signal. Root duration drops when Mel stands close to her target and rises when she stands at maximum range. That is a tax on close-range self-peel and a bonus for holding maximum distance. A mechanic that forces you to stand further back in a fight is the posture of a support, not the posture of a solo-lane mage. Add W now granting shield and movement speed to allies, and the direction stops being ambiguous.

The passive says the same thing. Armor scaling on a mage is an unusual design choice, and it does not help win duels — it helps survive dives. That is the design language of a champion who stands behind, not in front. Combined with R's cooldown increase and reduced per-level scaling, and with the passive losing early-game damage, Mel loses precisely her strongest window: the early and mid game, when one well-timed reflect was enough to flip a lane. If she is to remain relevant in competitive play, her value must migrate to the late game, where a shield and movement speed for the whole team actually matter. Whether that compensation package is sufficient cannot be verified before post-patch data exists.

There is a counter-reading worth taking seriously. For competitive play, losing solo-lane power while gaining flex ambiguity is not automatically bad news. A champion who can appear in two roles forces the opponent to hold their reveal, and that value in a professional draft is often larger than the value of one strong button. Mel loses her role as a counter-pick into projectile-heavy compositions but gains the chance to become a hard-to-read utility piece. Structurally, that is a reasonable bet.

The patch calendar softens the shock as well. Riot generally ships on a roughly biweekly cadence, and 26.22 lands in the late-season window — most likely after the World Championship patch has been locked and before the following spring split begins. If so, professional teams get an entire preseason to adapt, and a rework-level change produces no mid-split disruption. I flag this as conditional inference, because neither the exact release date nor the 2026 event calendar is verified in the source material.

One internal contradiction is worth naming. Riot states the changes will make Mel "more balanced while still preserving her unique identity." But the unique identity under discussion is the reflect mechanic — the thing being deleted. Those two sentences cannot both be true. The new identity may be better, may be healthier, but it is a different identity.

Where I could be wrong

I could be wrong in reading too much into a patch that arrives with no data attached. No win rate, no pick rate, no ban rate — which means my entire argument rests on kit structure rather than on outcomes. If post-patch data shows Mel retains an acceptable laning power level, my claim that she loses her strongest window collapses.

I could also be wrong in treating the support pivot as an inevitable outcome. Shield plus movement speed sounds very much like a support kit, but if the ratios are too low it is just a worse self-peel button. And if the ratios are too high, Mel becomes the new frustration in the support role — a champion who destroys projectiles, grants a shield and hands out movement speed to the whole team. Riot has traded one frustration for another before.

Finally, there is a real chance Riot is right and I am mourning something that did not deserve mourning. The reflect mechanic may be structurally untunable: its power is binary, either it reflects or it does not, with no dial in between. A design team does not discard a signature mechanic after multiple failed tuning passes unless it has concluded the mechanic cannot be tuned. If that is the case, deleting it was correct — and the correct decision still gets paid for by a specific group of players.

What I am willing to bet

Within the first two patch cycles after 26.22, in professional scrims and high-ranked play, Mel's support pick rate will exceed her mid-lane pick rate. If that happens, Riot did not rework a mid laner — they released a new support and put an old name on it. Abilities are remembered through clips, but the bill is paid in win rate. And in League of Legends, no hot take is ever too early; only the analysis gets published too late.