Riot Phroxzon: "Mage bot through roaming supports is something we're not happy with"

So the tweet you made recently — it might change again, you might revisit your opinion of it?

There are some things we want to do that are core to the fundamental satisfaction of the role. For example, Berserker's Greaves being weak for quite a while relative to options like defensive boots or, in some cases, Gluttonous Greaves — that's something we'd want to address. Same with Fleet Footwork, which doesn't really have any ranged users other than Gnar, situationally, and Jhin in some cases. So there are pieces of the AD carry system that are underpowered regardless of how we feel about mage bot, and this is a great opportunity to increase the satisfaction of the role even if it doesn't "solve" the meta problem. That said, in regular play, mage bots have elevated win rates and performances overall, and part of that is the shallow mastery curves I mentioned. If you play Ziggs bot for the first time, your performance is usually pretty high relative to the effort — similar to playing Malphite top or jungle — because mage bots are "easy to play" and a lot more fault-tolerant. Whereas if you play ADC and don't know what you're doing, you just get completely clowned on. So we're trying to hit two sides of the problem: in regular play, players have complained about mage bot for a long time

Though we don't want to move too hard, because mage bots and bot-lane diversity are quite healthy for the game.

And separately, the pro-play situation, where teams lean on their support. Two sides of the same coin.

Doesn't this point at something structural about marksmen — that they have the least individual agency right now? A mage bot brings range, wave clear and CC on its own, while the ADC brings damage that depends on four other people keeping them alive. Can you patch that, or is it baked into the class?

There are certainly aspects we can change, but I'd challenge the premise a bit that ADCs have low agency. There's a lot of focus on botlane right now, and if you're the better ADC, you can navigate a jungle gank better, which prevents your top or mid from getting ganked. You can draw pressure to your lane and potentially set up a counter-gank. So I'd push back on the idea that ADCs have low agency, especially in the current environment. If they get to their spike, particularly two or three items unobstructed, they can be very powerful.

On the meta as a whole — so much feels permissive right now: AP bots, ranged tops, assassin mids, roaming supports, bruisers or tanks in the jungle. Was that openness the actual goal this year?

We've been making a lot of changes over the past several years to add certain champions into the jungle — Darius jungle, Quinn jungle, Teemo jungle, things like this. Part of that is because jungle is a heavily auto-filled role, and letting players play things they're comfortable with makes the autofill experience a lot more tolerable — especially with changes we've made to Aegis of Valor as incentives to play those autofill roles. We're applying that same philosophy across all the roles now. That said, we don't want toplane completely overrun by ranged tops, or botlane by mage bots, or very unconventional strategies — I think I saw Aatrox bot and Mundo bot. We don't want that to be the norm. But we want players to feel comfortable when they're filled into those roles, and for innovative, "cooking"-type players to be able to demonstrate their creativity.

Pro play still revolves around mid, jungle and support — the roles that dictate the pace. Role quests were meant to give top and ADC more power. You already challenged the idea that ADCs lack agency, but would you say those two roles are where you want them right now?

The balance between the roles is still a little off. ADCs broadly are still more powerful than top lane, both in agency and pure power level. Agency is how much your decisions impact the game; power level is — imagine an ADC as a walking turret, very slow with few decisions to make. How powerful is that if you just right-click a walking turret down mid? Pretty powerful, even though it's not making many decisions. Both of those are in the ADC's favour versus the top laner's, especially at higher levels of play. Part of that is that ADCs are considered so powerful that pathing to bot is a very strong strategy, so the top laner is on a "top lane island," which limits their influence. You can get the enemy top laner down 0-2, but if botlane has exploded repeatedly by 10 minutes, it can be hard for a top laner to impact the game. That's a dynamic we're trying to balance better, because we want toplaners to feel some sense of agency — but they also don't like being ganked by the jungler, and that's one of the main ways we'd give a class more agency.

With the mage meta, it's been a bit like that too — botlane on an island, with jungle-mid-support and the fights on the river being very important in pro play. Are you happy with both bot and top being an island right now, or is that something you'd change?

Of the set of things we'd want to change for the bot meta overall, roaming supports is the thing we feel isn't great for gameplay, for a variety of reasons. One is that ADCs, at the champion-kit level, aren't really designed to be 1v1 champions — a lot of them have supportive elements. Think of something like Jinx's trap, which isn't really meant to be landable without an ally. A lot of kits just stop functioning properly when they're played 1v1. And roaming supports create a lot of negative follow-on effects: midlaners getting ganked frequently, or supports walking into the enemy jungle and messing up their clear for pretty minimal consequences.

So regardless of how mage bots play out, mage bot through the avenue of roaming supports is something we're not super happy with and would want to take action on.

If you fix the support roaming, aren't you risking diving straight back into the old marksman meta of enchanters and hyper carries only?

It's certainly a risk. We took a pretty significant set of actions on 26.11 to nerf enchanters overall — we want tanks — and we do think that caused some amount of a meta shift, even on top of the roaming-support stuff. So we're a little worried about it. That said, things like Ziggs bot were valid before he even started doing this roaming stuff, and Mel bot and so on. Even double ranged botlane can be a pretty powerful strategy — a bunch of teams were playing Seraphine plus another ranged champion. We think that's more the way we want mage bot to manifest, rather than Ziggs holding down the fort 1v2 weak side, just wave-clearing while Pyke runs around the map making everyone sad. That's more the type of gameplay we'd be excited to see.

Is there an aspect of the pro-meta, outside of mage bots and roaming supports, that Riot has been particularly looking at — something you're not satisfied with and might want to fix before Worlds?

Two come to mind. One is top-lane carry champions. There's a pretty strong focus on top lane being a safe weak side or safe blind pick, so the jungle can walk around with the roaming support. But even outside that case, toplaners are typically picking Ambessa, Renekton, all the usual tanks. So we want to encourage, and have more structural things supporting, those carry champions — Jax, or moving Camille from support back to top lane, that kind of thing. As I said, toplaners, especially at higher levels, just don't provide as much payoff in pro as ADCs, and that's a dynamic we'd want to resolve. The second thing is getting more AP junglers into the mix — partly because we think there aren't a tonne of viable AD mids. Which is interesting, because even when teams play mage bot, they don't play it with an AD mid; they play it with an AP mid and put double AD on jungle and top. So it indicates AD mids as a whole are pretty weak, which is another reason buffing Berserker's Greaves or Fleet Footwork would be doubly viable. One of the things we were scared about with AD mids was too many ADCs in the midlane — Corki mid, Tristana mid — but the role quests haven't been super kind to ADC mid, and opportunity-cost-wise it's much better to run the ADC in botlane. So we think we've got a bit more room there. Those are the two: AP junglers and carry top.

On the state of the game — we've seen serious LP inflation in solo queue across servers, players smashing rank records. Is that a concern, or was it intended?

Broadly, it was intended. League is a game where we want to strike the best balance of matchmaking being fair and being perceived as fair. When we looked at the ladder most recently, something frustrating for players was wide skill gaps in certain divisions, especially Diamond. So we wanted to make those divisions have narrower bands of skill. We knew that would result in Masters having too many people, or LP values getting pretty high overall — but we wanted to make those divisions, especially Diamond, which had such a broad range of skill before, more narrow. In the future we can make the apex tiers more granular, add more divisions and breakpoints, and take a swing at the inflation. LP values are getting pretty high, and while that was a known side effect, we want to find ways to resolve it.

Why only one new champion this year — what's the thinking behind that pace?

There's a lot of strategic planning around what we release, and we're always experimenting with different shapes of seasonal packages. It just so happened that this year, the way Locke came out timed really well with the seasonal narrative we were spinning for seasons one and two — the progression of Demacia from a bastion of hope into a darker flavour with the demons for season two. So when we planned out our tapestry for the seasons, we felt a champion release in season two was the best way to go, with Shyvana's update in season one. Broadly, we're always trying to do different things, and champions are sometimes part of the way to make a season epic, but there are other ways too.

Season two didn't bring that many changes. Why, and should we expect more from season three?

For season three, with Worlds coming up very quickly, at least on the Summoner's Rift side

We're not looking to make super major changes — we're looking to stabilise the meta.

One thing we heard strongly from 2025 was that there was just too much change coming through the game, especially on SR — Feats of Strength, Atakhan, then changing Atakhan in season two, then changing stuff again, plus a lot of items. So going into 2026 we're re-evaluating how much change is enough, and how much is a good amount for players to be happy. Season three is a lot more of a stabilisation window: resolving some of those pro-meta concerns, getting balance into a good spot, resolving the agency of various lanes, improving matchmaking quality. We just shipped a set of matchmaking changes in season three, and we'll keep balancing the impact of queue times and that kind of thing.

The game feels extremely swingy right now — early leads convert far less often than they used to, and we've seen T1 lose games off big leads across every region. Supa told me yesterday that teamfighting is close to the only thing that matters. How do you explain how hard it's become to snowball, and is that volatility something you're happy with?

There were a bunch of intentional movements toward making games more uncertain, and it's something we need to keep an eye on, because as games go on in their life cycle they get better and better optimised. One thing we don't like to see in the pro game is a game that ends 7-2 — a 29-minute game where they basically played laning phase, slowly eked out small advantages, won one teamfight and ended it. Not a great viewing experience, not a lot of suspense. At the same time, we need to balance how that plays out for pro versus regular play — you could argue lane lethality and the ability to kill people early and set the snowball rolling isn't high enough in pro, but in the regular game it's almost too high. If you're in Bronze and Garen runs at you pressing Q-E, you're probably dying. A Bronze player isn't going to say "yeah, make me die faster." So there's a fine balancing point there. Broadly, for pro play

We do like games being pretty swingy — it results in quick decision-making, teams can make more mistakes, it makes games more exciting and volatile, and it's just a much better viewing experience.

But we balance that against regular play feeling like the right amount of lethality, comeback systems not being too strong, and so on. So it's in a pretty reasonable spot overall.

Pro play right now looks like skirmish after skirmish. Last year a lot of people worried about the "five minutes of doom" and concepts like that. Are you happy with where the pro game sits now?

We were explicitly trying to make adjustments to resolve things like the five minutes of doom, because it's a pretty boring experience. Even though side laning is very strong strategically and pro teams want to do it a lot, if the game skews too heavily into side laning and "don't fight anybody" map pressuring, games get boring. So with things like the crystalline overgrowth and some minion behaviours, we wanted to open the door for teams to demonstrate side laning and play the map more strategically — but without going overboard. We still think one of the most exciting things about League is skirmishing and teamfighting, and continuing to do that creates the most dynamic behaviour and lets players show their skill. So in that respect, I wouldn't say the meta needs a significant overall swing.

Last topic — consumption and player base. How do you see the way people consume games evolving right now?

It's a very interesting question, especially now, because a lot of games are trying to hyper-specialise for their audiences, and League is no different. We just released League Classic, and we've had good success with ARAM Mayhem and with core Summoner's Rift. One of our learnings is how to specialise different experiences for different subsets of the audience. Classic is aimed at old-school players who've played a long time but are short on time and want something more relaxed. Mayhem is targeted at a brain-off, high-octane environment and pace.

In terms of actual stress, game pace, understandability and clarity, it's [League Classic] really targeted at that audience. So the way we're moving is toward hyper-specialising experiences for various audiences, because the player base isn't one type of person — it's a collection of many. And especially for Summoner's Rift, a lot of players are watchers now; they don't play as much as they used to, but they still love watching. So we need to be very conscious of all those audiences and curate our experiences to meet their needs.

League's player base is, in majority, ageing. How closely do you track that, and how much of a priority is bringing in new players — versus bringing old ones back with Classic and such?

We track it closely, and new players are one of the places to focus. New-player onboarding has been pretty painful in recent years, and we're looking at efforts across a lot of avenues to make it a lot better — taking stances on disruptive behaviour, better matchmaking, new onboarding flows. There's a moderate amount of effort going into meeting new players' needs, the same way we're putting effort into those other three categories, and finding a mechanism that's well-curated for new players as well as existing ones. If we don't have new players, it's hard for the game to have longevity, so it's still an area of focus.

If someone had never tried League, give them one good reason to be excited about where the gameplay is heading.

League is an environment where the impossible can become possible, and it's one of the best games at doing that.

So that's what I'd say.

RAPID FIRE QUESTIONS

A rapid-fire round to close out — yes/no answers, with optional context.

(RFQ Youtube link)

Fight-heavy pace here to stay? → Yes.

Can the game get faster or are we at the limit? → Yes.

Will marksmen still be the default botlane class in two years? → Yes.

Are rolequests here to stay? → Maybe.

Is the current pro diversity something you're deliberately aiming for? → Yes.

Are pro teams consulted before you ship a patch before a major international event? → Yes.

Does the ageing player base worry you? → Yes.

Should a 5K gold lead at 15 minutes mean more than it does today? → Maybe, maybe or no.

Will WASD be an edge for pros who master it? → Yes.

Could the map still evolve in the coming years? → Yes.

Can a single patch redefine a pro meta? → Yes.

Are you broadly happy with the state of the game? → Yes.