Riot Games Replaces Mel's Reflect With a Shield and Movement Speed: Patch 26.22 and the Cost of an Untunable Ability

Jhin's fourth shot leaves the barrel. It flies straight at Mel, and it does not disappear. It turns around.

The Jhin player watches his own bullet travel back toward him, carrying the stacked damage of all four shots, and kill him before his finger reaches the retreat key. There is no shouting in the headset. Only breathing.

Months later, another game, another server. Seraphine casts her ultimate — the long charm wave that fills an entire narrow corridor, the most expensive opening weapon in her kit. Mel raises a hand. The wave turns back. The enemy team, including the woman who cast it, stands frozen for four seconds, and the game ends on that frame.

Those two moments are the reason Mel exists in League of Legends. They are also the reason she is about to stop existing in that form.

Some goals never hit the net; they go straight into memory. The same holds inside a League game: there are plays that never appear on the post-game damage chart, are never counted in any stat line, are never graded by anyone — they exist only in the head of the person who got punished and in clips shared ten thousand times. Mel's reflect mechanic lived in exactly that zone of memory. And patch 26.22 has just decided to delete it.

Context: an ability condemned by the very moments that defined it

Mel is a mid-lane mage built around an unusual idea: she does not dodge, she returns. Her W was a defensive window — an ability that opened for a few frames, during which any projectile flying at her was bounced back toward its sender.

Technically, this is a beautiful mechanic. It rewards reading the game rather than hand speed. A good Mel does not need to press W fast; she needs to press W correctly. She has to see, half a second early, that Seraphine is about to cast her ultimate, that Jhin is holding his fourth shot, that Ezreal just crossed the mana threshold for a snipe. That is an intellectual skill, not a mechanical one — and in a game where most of the gap between good and elite players sits in hand speed, a purely read-based skill is something worth preserving.

But it carries one fatal design flaw: its power cannot be subdivided.

This is the point anyone who has ever done balance work recognises immediately. A tunable ability has a scale — 5% less damage, 0.25 seconds more cooldown, a smaller radius. Reflect has only two states: on or off. It either stops Seraphine's ultimate or it does not. There is no version of the ability that stops 70% of Seraphine's ultimate.

According to Riot Games' announcement, the design team went through multiple rounds of adjustments trying to preserve the mechanic before deciding to replace it. I read that detail and found it familiar in an uncomfortable way. In fourteen years of covering this industry, I have watched that exact pattern repeat: a mechanic is judged too strong, gets a numbers cut, is still strong, gets another cut, is still strong — until someone in the room says the thing everyone already knew. The problem is not the number. The problem is the mechanic.

Patch 26.22 is the first time Riot has admitted that with action rather than words.

The analysis: four changes, one identity migration

The biggest change sits on W. The ability no longer reflects incoming projectiles. Instead it destroys them — stops them cold, removes them from space, but returns no damage to the caster. As compensation, W now grants a shield and movement speed to nearby allies while it is active.

Read that sentence slowly. An ability that was a punishment tool — the thing that made opponents afraid to fire — has become a protection tool. Same key, same slot on the ability bar, but the in-game function has changed category entirely.

The other three changes are quieter, but they all point the same way.

Mel's passive loses early-game damage and duration, and gains health and armour scaling. Armour scaling on a mid-lane mage passive is a very strange signal — armour does not help you win a duel, it helps you survive a dive. That is the signature of a champion designed to absorb pressure and stand her ground, not to dominate.

Q is redistributed: the front-loaded burst of the first hit is spread into subsequent hits. This is the classic way to cool down early mid-lane power. A mage with a heavy front-loaded Q can push waves, trade health and generate map pressure before the opponent finishes a first item. Spreading the damage means Mel needs more time to do the same thing — and in League, time is the most expensive resource on the map.

E is the change I find most interesting and the one most under-reported. Root duration now depends on distance: reduced on close targets, increased at the edge of range. On the surface that is a small technical tweak. Look closer and it is an incentive written directly into the ruleset — it pays the player who stands far away and taxes the player who stands close.

That is the design language of a support, not a mid laner.

The ultimate takes a cooldown increase and a per-level scaling reduction. This cuts directly at the thing that made Mel draftable in professional play: the ability to open a fight and lock a target with a single cast. Weakening the ultimate weakens Mel as a press-one-button-and-win option.

Put the four pieces together and the picture is clear: Mel loses her strongest window — the early and mid game, when one well-timed reflect could flip a fight — and is pushed toward the late game, where her value lies in keeping teammates alive.

Riot did not weaken Mel. Riot moved her to a different neighbourhood.

Who gains, who pays

The clearest beneficiaries are projectile-dependent champions — the ones who used to be punished by their own damage. Jhin with his fourth shot. Seraphine with her ultimate. Every ranged marksman and artillery mage who had to calculate whether firing was safe while Mel still held W.

The second group is melee divers. Diving a Mel with W available used to be a gamble: hold the ability correctly and your opening blow bounces back. That fear is gone.

The third group is the least discussed: Mel's own teammates. A multi-target shield plus movement speed is one of the strongest forms of support in League, because it cannot be dodged. You cannot sidestep a shield.

The losing side is clearer, and it loses in a harsher way.

Dedicated Mel players — the ones who spent hundreds of hours learning exactly which frame of W catches Seraphine's ultimate — lose that entire investment. This is the cruellest class of balance change a team can ship. A damage cut leaves good players good; they just recalculate. Removing a mechanic makes their skill evaporate. The industry has a phrase for this, and it is not a pleasant one: skill obsolescence.

The second loser is any team that used Mel as a counter-pick into projectile-heavy compositions. That option no longer carries equivalent value.

The contrarian angle: Riot fixed the easy frustration, not the real one

This is where I want to slow down.

Riot states its design goal clearly: limit the situations that feel too frustrating when playing against Mel. The problem is that the frustration W created has two layers, and this patch only touches one.

Layer one, the loud one: an opponent fires a big ability, it comes back and kills them, or worse, comes back and locks down their entire team. This is the kind of frustration that produces clips. It generates a sense of injustice because it breaks a basic rule of the game — what you fire can never hurt you. This layer has been removed.

Layer two, the quiet one: an opponent fires a big ability, and it vanishes. No damage, no effect, nothing. You spent mana, cooldown and a decisive moment, and received zero in return.

The new W still destroys projectiles. Layer two survives intact, wearing a different coat. The feeling of "my spell meant nothing" has not decreased. The feeling of "my spell killed me" is gone.

And this is the most debatable point in the entire announcement: Riot is fixing the most visible frustration, not the most common one.

The evidence sits inside Riot's own framing. The two examples offered — reflecting Seraphine's ultimate to charm an entire enemy team, reflecting Jhin's fourth shot back at Jhin — are extreme cases. They are gorgeous, they are clip-worthy, and they are extraordinarily rare.

The announcement offers no frequency data at all. What percentage of Mel games actually produced a reflect that flipped a fight? Unknown. What percentage of W casts simply blocked a single stray projectile? Unknown. Mel's win rate before and after? Not provided. Pick rate, ban rate, presence? Not provided.

I have spent a decade reading patch notes, and I have learned one thing: when a design decision is justified by two stories instead of a table of numbers, the decision was probably driven by community sentiment rather than operational data. That does not make it wrong. It makes it unverifiable, and that is an uncomfortable place to stand.

There is also an internal contradiction. Riot says the patch will make Mel more balanced while preserving her unique identity. But the unique identity in question was the reflect mechanic — the very thing being deleted. Both statements cannot be true at once. You cannot preserve a thing and delete it in the same patch.

A different reading is more plausible: Mel's new identity is a support identity. Shields for allies, movement speed for allies, range-rewarded roots, armour-scaling passive to survive dives. That is the signature of the support class, not the mid lane. Moving Mel there may have been the intended outcome all along — it simply was not phrased that way.

The most notable thing is what the announcement omits

No team is named. No player is named. No tournament is referenced. This is a live-server balance notice, and it is purely about champion design.

But the absence itself tells a story.

A rework-level change — not a numbers tweak, but a mechanic replacement — usually appears in one of two kinds of announcement. The first is routine balance, where the issue is power. The second is urgent, where a champion is damaging competitive integrity and must be handled before a major event.

This is the first kind. And the gap between announcement and live deployment — roughly three weeks according to the details given — is long enough to conclude that Riot does not treat this as a crisis. They treat it as a player-satisfaction problem, not a competitive-integrity problem.

I do not fully buy that reading, but it is the most reasonable one available from what has been published.

Another hypothesis with some grounding: the reflect mechanic was likely a persistent problem in professional scrims, not only in solo queue. The reason is simple — design teams do not discard a signature mechanic after multiple failed tuning passes unless the mechanic is structurally untunable. And when they do discard it, the pressure usually comes from the loudest room. Solo queue is loud because it is large. Professional play is loud because it is expensive.

Based on my experience watching matches and scrim blocks across many seasons, champions with narrow but deeply oppressive matchup profiles tend to survive longer than broadly strong champions. They are not banned often, not picked often, but every appearance forces someone to rewrite an entire draft plan. Mel sits precisely in that group. And that group is always handled last, when everything else is settled.

What I believe, and what I cannot prove

I believe this decision is correct on design grounds. A mechanic with only two states cannot live inside a game that requires hundreds of fine-grained balance levels. Riot tried, failed, and admitted it. That cracked-voice night taught me that honesty costs more than perfection — and a design team willing to say "we tried several times and could not make it work" is more credible than one claiming a perfect solution.

I also believe that for professional play this change may be better than it looks. Losing mid-lane power while gaining support flexibility is exactly the kind of ambiguity draft coaches love, because it hides a champion's role until the opponent has committed. In a best-of-five, ambiguity is worth more than raw power. A Mel who can go to two lanes forces two preparation plans.

But I cannot prove that, and I will not pretend otherwise. The announcement itself contains no win-rate, pick-rate or presence figures for Mel on any server. Every conclusion about post-patch strength is inference. I write this aware that I am analysing a design decision, not a competitive result.

One thing I am more certain of: every patch note is a promise not yet written in ink. Riot promises Mel will be more balanced, that her identity survives, that playing against her will feel better. Those promises can only be tested across millions of games, by players who do not read announcements, do not care about design philosophy, and know only one thing — today I pressed my ability and it did nothing.

Takeaway

There is a question I carry into every change like this, and it is not specific to Mel: when a mechanic is deleted because it is too hard to balance, are we protecting the game, or protecting the comfort of the majority? Both are legitimate goals. They are not the same goal, and every time we choose, something is lost on the other side.

Mel players who learned to read frames will carry the memory of their reflects far longer than any patch survives. Jhin's fourth shot will keep turning around in their heads, even after it stops turning around on screen.

And that may be the only part of this story no one can patch out.