LoLMelMel got weaker. She got played less. And players kept banning her at exactly the same rate anyway. That combination is why Riot stopped trying to fix her with numbers and went after the kit instead.
Rebuttal keeps the name and loses the reflect in Patch 26.22 (Image: Riot Games)
The Mel rework arrives in Patch 26.22, and it takes away the one thing she was designed around. Her W will no longer send abilities back at the people who cast them.
Riot published the reasoning on October 20, and the dev post is unusually blunt about what went wrong. The short version: new mechanics are bets, and this one did not pay off.
Rebuttal keeps its name and loses its party trick.
The new W fires a nova that destroys enemy projectiles, shields Mel and her allies, grants them move speed, and then deals a small amount of damage back to attackers. Riot says it will protect her team better than the live version does.
What it will not do is send the ability itself back. No more reflected Seraphine Ultimate charming an entire team. No more killing Jhin with his own Fourth Shot.
Look at the three numbers Riot published, because they tell the whole story.
For most of her first year, Mel sat at a 48-49% win rate with a ban rate between 15% and 30% and a combined pick rate above 10%. Slightly under par, frustrating to some, but a champion people clearly wanted to play.
By the middle of this year, she was at 45-47% win rate. Her pick rate had more than halved. And her ban rate had not moved at all.
That last part is the problem.
Normally a champion who gets weaker and less popular also slides down people’s ban lists, because the annoyance stops justifying the slot. Mel did not. Players kept spending a ban on her while she lost more games and saw her pick rate halve. She had arrived at the worst possible position: not strong enough for her own players, still irritating enough that everyone else wanted her gone.
No number on a spreadsheet fixes that. Riot could buff her and make the frustration worse, or nerf her and make her unplayable. The kit had to change.
Riot tested versions that kept the reflect, and the post describes two of the failures.
The first shortened her range, so that a powerful defensive W felt fairer to play against. Mel players rejected it. Giving up range is giving up the champion.
The second is the interesting one. Instead of bouncing an ability back, Mel would steal a cast of the first missile that hit her. Riot calls it “very exciting” and then explains why it died: it would have required Mel players to know how to aim every projectile in the game, and in a messy teamfight nobody could tell which spell they had just picked up.
That is a genuinely fun idea killed by a practical problem, and it is the kind of thing dev posts usually leave out.
The obvious fix is to move the reflect onto her R and push Golden Eclipse down to a basic ability. Riot walked through that and rejected it, with specifics.
So the reflect had nowhere to live. It was too annoying as a basic ability and too strong as an Ultimate.
Removing the reflect frees up power, and Riot is moving it from her early game into her scaling.
The Q refund and the Ultimate scaling are the real changes here. Riot’s complaint about live Mel is that she usually manages one rotation of Searing Brilliance stacks, kills one person, and then stops mattering. Those two adjustments exist so she can stack a second target and use her Ultimate as the payoff.
The E change is the one laners should read twice. A shorter root against nearby enemies makes her noticeably more vulnerable in close-range matchups she currently handles well.
The Mel rework ships in Patch 26.22, which on the current two-week cadence lands in early November. Patch 26.21 went live on October 21, so November 4 is the expected date.
None of it touches Worlds. Riot locked the competitive patch before any of this, so the teams in our Worlds 2026 Swiss Stage coverage are playing an older build.
The Mel rework cuts both ways, and it depends on why you picked her.
If the reflect was the appeal, this is a loss and no amount of shield value replaces it. Riot says as much in the post, which is to their credit. You lose the highlight-reel moments, permanently.
If you picked Mel to play a mage, this is probably an upgrade. She has been a champion who wins lane against a specific list of opponents and then fades, and the new kit aims at doing something in the late game instead.
Riot has also been unusually willing to take things away this year, from retiring League Classic in 2027 to this. Not every removal lands badly, but players notice the pattern.
The wider question is what it says about her design. Riot shipped a champion whose identity was a counter-mechanic, and once good players worked out how to use it, there was no version of it that both sides could live with. Riot’s own post does not really dispute that reading. It describes a bet, and then it describes the bet failing.
Patch 26.22 will tell us whether a Mel without the reflect is still Mel. Our Patch 26.21 breakdown covers what is live right now, and we track every champion change in our League of Legends section. Riot’s full post is on the official dev blog.
LOLThirty-five new augments is the headline number, but the Headliner system is the actual change. Every game now runs under a band’s rule set, and two maps are missing while the stages get built.
LOLEveryone will read the Caitlyn and Jinx nerfs and move on. The changes that actually reshape bot lane are two item lines that quietly split melee and ranged users into separate rules.
LOLLeague Classic ends its run in January 2027. Let’s explore the key dates, refund details, queue changes, and what Riot says about a future return.
