/Dev TFT: Into the Arcane Learnings

Welcome back to our once-a-set conversation where we reflect on our favorites, our not-so-favorites, our successes, hiccups, and everything in between. These articles are always deep, and since we’re making three sets a year, we’re now using them to discuss just how we’re thinking about sets and TFT on a longer timeline. TFT is pretty big now too, so we’re going to be calling some friends from across the team to help us make this piece a bit more holistic. 

Okay, let’s turn on that rainy lo-fi and get reading! 

TL;DR: We’re going to dive deep into our Into the Arcane learnings on the following topics:

• How did Arcane land as a set: Arcane is a high-stakes, gritty, and character-driven epic… TFT has a penguin with a wooden sword. So how did we merge the two, and how did it go? Well, it depends on how much you like Smeech or Sevika, but it also depends on how much you enjoyed Loris and Steb… Shipping alongside the show also meant a shortened PBE cycle leading to balance not being as tight as we wanted, and bugs that slipped through the cracks. 

• Evolving the set with 6-costs: Shipping our release version of the set without 6-costs felt incomplete, but our staggered approach just made the most sense.

• Anomaly and Champion Elevation mechanics: The Anomaly was a pretty successful mechanic, especially after we solved some of its initial issues. We had a ton of fun pushing champion fantasies further in novel ways, and this is a space we intend to explore in the future, but not with our next set. 

• Augment Design: We had several smaller learnings in the Augment space from new ways to make High-stakes Augments to learnings around Champion Augments. 

• Opening Encounters: There are clear tradeoffs with having these versus Portals, but the pros continue to be worth their use. That being said, we have learned about what makes a good/bad Opening Encounter.  

• Chem-Baron: Chem-Baron was our first high-stakes trait that directly empowered the units of the trait after the cashout, and that was an important exploration for us. We learned a lot here (some good, some less so), but we’re going to shift to a different style of cashout trait for our next set. 

• Multiple-Choice Options: Form Swapper was our most successful multiple choice trait, but it taught us a valuable lesson about perception, options, and the rounding-up issue. 

• Melee Champion Power Scaling: Because of how much HP Melee carries gain when they star up, the difference between a two-star and three-star melee champ is quite large, as they benefit from the HP and the AD growth. We have a solution for this, but it’s going to be a while before we can bring it to live. 

• Tanks and Durability: Tanks have four different types of abilities, and ones that grant Durability perform the worst. We’re going to have to add more levers to Durability tanks in the future, or find alternative hooks that keep them viable throughout the fight. 

• Festival of Beasts Revival: Modernizing Festival of Beasts had its challenges—and we’ll dive into some of them here.

How did Arcane land as a set?

Arcane is a high-stakes, gritty, and character-driven epic… TFT has a penguin with a wooden sword. Creating a TFT set that bridged our favorite game with our favorite show meant crafting a set darker, bolder, and more intense than anything we’ve done before. And while we paid homage to Piltover and Zaun with our sixth set, Gizmos & Gadgets, Into the Arcane was an entirely different beast. It was designed with the lofty goal of letting players relive and rewrite Arcane in their own way… in the Convergence. So, how did we pull it off, and how do we think it landed?

First, we built a roster. Into the Arcane’s lineup was split three ways—one-third Arcane characters, one-third Piltover and Zaun, and one-third Runeterra’s finest—to balance trait webs and gameplay variety. This let Arcane fans who were new to TFT instantly build their Enforcer squad, experiment with the perils of Shimmer through Chem-Baron, or chase those ultra-powerful 6-costs. Meanwhile, longtime TFT fans could still run it back with the Conquerors of Noxus or their favorite Emissaries from across Runeterra. But while this structure catered to both Arcane fans and seasoned TFT players, committing to 10 brand-new champions was no small feat.

Making 10 new characters was an exciting undertaking, but also a major lift for our team—and some of our new units may have fallen short for our less-Arcane driven players. For every Loris and Steb we added, we missed the chance to include a Yasuo, a Syndra, or another beloved League champion. New faces like Loris and Steb simply didn’t have the depth to inspire the same level of fandom fervor. That said, we did see some big wins—units like Sevika and Smeech were instant favorites. 

Beyond units, we had an Arcane-driven mechanic—which has its own section—but for now, we’ll just hit on its theme. In the current state of TFT, we see set mechanics as smaller scope pieces that drive the theme and make each set feel distinctly set-y—if that doesn’t make sense, let’s apply this logic to Into the Arcane real quick. The Anomaly wasn’t meant to be a colossal game changer; it showed up on Stage 4-6 to empower a single champion in a unique way. It was just one piece of the set’s identity, alongside an Augment refresh and an entirely new roster. The Anomaly wasn’t about flipping the meta—it was about injecting Arcane’s defining moments into TFT. What happens if you send Ekko into the Anomaly, and he comes back not just with a deeper understanding of Jinx… but also laser eyes? Now that’s peak TFT wackiness—but it’s also Arcane.

The release of Arcane was a huge moment across all of Riot, not just TFT, so our plans had to be synchronized to the hilt. This created complexities of course, in the way we were able to ship the set and talk about it ahead of its release, as well as a shortened PBE cycle, which made day one balance even tougher and had some bugs slipping through the cracks. But this release timing also led to some really cool moments, including releasing our 6-costs as they were revealed in Arcane. As episodes dropped, the set continued to evolve, which was our way of bringing Arcane to TFT players in the most authentic way possible.

Evolved Set / 6-costs

Well, we couldn’t just reveal Mel's super cool magical powers before Act three even aired, could we? 

We always planned to include high-impact 6-costs in Into the Arcane—their characters were just too impactful not to—but timing was everything. Launching an evolved version of the set would bring our vision to completion, but it also meant shipping an initial version of the set that, let’s be real, felt incomplete. I can’t even count how many times we heard, “Where’s Viktor? Where’s Warwick?” and we loved the player theories—like “I bet they’ll swap out Malzahar for Viktor” (which, honestly, made a lot of sense). But in the end, our staggered approach just made the most sense.

We also made the choice to let anyone access 6-costs, regardless of player level, once the Anomaly appeared (or in Viktor’s case, a bit earlier). This meant that every player—whether they preferred power-leveling or reroll strategies—had a fair shot at these game-defining units. It also ensured reroll comps didn’t get left in the dust just because 6-costs broke the power ceiling (board-wide stuns, elimination prevention, very scary dog).

But this approach had tradeoffs. Aside from hitting Level 10 or hoarding gold, there was no real agency in finding a 6-cost—it was just luck. And while that rush felt amazing when you hit, it also meant feels-bad moments when your opponent rolled a game-changing 6-cost right after the Anomaly appeared. Ultimately, we landed on this approach as the least problematic—it kept 6-costs accessible without forcing every player into a power-leveling arms race.

Now, let’s talk about power balance. Our goal was simple: When a 6-cost showed up in your shop, you’d want to slam it into your board—even at the cost of trait synergy. They needed to be strong, flexible, and viable with or without items. Lofty goals, sure. Did we nail it? Not quite.

On release, 6-costs had too much utility. We overcompensated, thinking they needed extra versatility since they didn’t contribute to traits. Instead, we saw players picking them purely for their effects—Viktor for his shred/sunder and stun, or Mel just to trigger her Banished Mage passive before selling her. Not ideal. We’ve since fine-tuned them (Patch 13.6 is about to drop with what we hope is the final tweak), but it took longer than we’d like.

We learned a ton from these champions. Before launch, we worried a board-wide stun might be unbalanceable, but once Viktor released, we realized that wasn’t even the scariest part of his kit. In fact, after removing his shred/sunder, we had to buff him multiple times. Warwick, meanwhile, taught us some hard lessons about stat-loaded melee carries—but that’s a story for another section.

Anomaly and Champion elevations

Our goal with set mechanics has changed in the past few sets (let’s call this time period the post-Augments era)—TFT’s already a complicated game, so since Gizmos & Gadgets, we’ve been keeping mechanics smaller in scope and more tied to the thematic. Within this paradigm, the Anomaly was a pretty successful set mechanic, especially after the first few patches, where we fine-tuned the distribution rules and added a few more interesting options. 

On release, you were able to hand pick the Anomaly you wanted, which had a few pros in enabling compositions that could only exist with a specific Anomaly (Urgot eating Sett was undeniably exciting), but the cons were significant, as players could force specific Anomalies, and thus comps that were believed strong, over and over again, reducing game variance. It’s worth noting that some players really liked being able to hand select a specific Anomaly—there will always be a subset of players that want less variance—but for most of our players, most of the time, the opposite is true.

So, as we do when we see a way to make TFT more fun, we acted fast. Our first change for the Anomaly created a cap of 12 rolls before you could see repeats, after which rolls would be truly random, which created a new problem. When outcomes are truly random, there's an equal chance for each option to appear at any given time, including the option you just rolled past. This is a tough reality of statistics, but effectively, each Anomaly was a number on a 66-sided die, so when you rolled a 3, which was Kill Streak, then you had the same probability to roll that again as you would any other specific Anomaly. So, while the math here was correct, we needed to add an invisible rule to make sure it wasn't truly random; that rule came another patch later, which prevented you from seeing the same Anomaly more than once within 12 rolls. 

Probability and math are foundational to game design, but another vital piece we took away from the Anomaly was which ones resonated most with players—aka, which Anomalies were the most fun. Even when at a balanced state, our most-picked Anomalies offered broad, universally useful power rather than niche applications. These were things like Infectious Anomaly, The Finisher, and Giant Sized. 

These Anomalies were popular because of their general use cases but also because of their simplistic nature. That’s an important reminder for us to make sure we have more straightforward outputs that prevent an increase in dizziness and are more broadly applicable to a variety of comps. But we, and you all, also learned another thing that starkly contrasted the above point, and that’s the way information spreads. 

Alright, one last thing about the Anomaly, then I promise we’ll move on. The Anomaly was a fresh way to push a champion’s power fantasy beyond just being a three-item carry—whether that be pulling enemies towards them with Deep Roots, or firing an endless laser at their target with Laser Eyes. Even the Center of the Universe Twitch example above turned a champion’s power fantasy into something even more fantastic. This is a fascinating space, and we fully intend to revisit it in the future—but not in our very next set.

Augment Design

There are several smaller learnings we’ve taken from our latest batch of Augments. They’re all pretty niche, so we’re going to move  through them quickly: 

Expected Unexpectedness v. Call to Chaos: When we diluted the exciting and random power of Call to Chaos into three smaller events, none of those felt very good and the Augment felt less satisfying. We still like the Augment, but this example reinforces that random highs are better as sharp, singular moments than drip releases. 

Silver Champion Augments: During Magic n’ Mayhem we had difficulty balancing our 1-cost Champion Augments at the Gold Augment power level, as they were either strong enough to scale into the late game (not entirely what we want from a 1-cost), or a complete bait. By placing them into the Silver tier we were better able to balance them, allowing them to have the appropriate power level more often while also creating some very fun experiences with Trolling, Mad Chemist, and Blade Dance all creating exciting and unique comps.

Golemify and Dummify: These are just fresh takes on high-stakes Augments that are worth calling out as great continuations in the space. High-stakes Augments may have been a big part of Magic n’ Mayhem, but many of them were fun enough to inspire further iterations. We’ll continue exploring the space, but don’t worry econ trait enjoyers, there will certainly be a trait (or two) for you…

Team-Up Augments: These successfully allowed relationships from Arcane that weren’t viable boards due to our trait webs to exist—and that was cool cause Into the Arcane was partially about doing just that. But while the fantasy these provided was great, the power and gameplay excitement for these Augments was diluted. A lot of their power had to be in the unit being accessed early, not leaving much room for the effect to contribute. Augments like Unlikely Duo (Sevika and Jinx) and Marshal Law (Caitlyn and Ambessa) all provided very powerful units, often before you could access those 5-costs reliably, so much of the combat power was instead a sort of economic power of just getting the Champions early. 

6-cost Augments: These all ended up under-tuned, as they had to be balanced around the ceiling they could provide, that ceiling being a 6-cost that you can put into the Anomaly to create a monstrous unit. So if you ever got to play Warwick + Power Consumption to make giant Warwick or another powerful carry Anomaly, then you know just how devastating these can be. But for most players, most of the time, they didn’t quite hit the mark. 

Opening Encounters

Opening Encounters are still something we have active discussions about because there are clear tradeoffs with them and because they’re something we see come up often within the community. 

Having Opening Encounters instead of Portals where you vote on an Opening Encounter option continues to hit our goal of lowering the complexity barriers (and avoiding the occasional anti-social consequence of the voting process), so we’re going to continue forward with Opening Encounters. But, we have learned a lot about Opening Encounters this set. Early on, we had no Opening Encounter occur far too frequently under the idea that players liked a more vanilla TFT experience, but that quickly got old, so we ended up going from 40 percent No Opening Encounter in our early playtest events to 20 percent No Opening Encounter in PBE, all the way to 10 percent by the time we hit live. We also learned that Opening Encounters that dramatically shift the core way you play TFT should probably not exist in most cases. Here, we’re referring to Warwick’s Hunger, which shifted the tempo of the lobby too significantly. To an extent, the Ambessa, Wandering Golem encounter does this too, but it’s something that players (and ourselves) overwhelmingly love, so we’ll make an exception and revisit the Opening Encounter if it falls out of favor with you all. 

Chem-Baron

Chem-Baron was our first high-stakes trait that directly empowered the units of the trait after the cashout, and that was a cool space to explore, especially given how many exciting Chem-Baron carries there were to play. 

But what plagued Chem-Baron during Into the Arcane was the threshholdy balance issues it had. On release, the trait was especially weak, with its items not giving enough power upon cashout and the recipe for accessing powerful loot being too hard (Emblem and the three-cost early). We ended up overcorrecting for our competitive players during the middle of the set, and it wasn’t until patch 13.5 where we made structural changes to the trait to eliminate the power ceiling, while also making it more consistent to hop into (requiring hitting one of the three-costs early, instead of an Emblem and a 3-cost). This change was in the right direction, as it allowed us to better balance the trait for our most competitive players by placing new controls on just how powerful Chem-Baron players could get, but it did result in the trait landing weaker than we’d hoped post-rework. 

Finally, let’s talk about the cashout rewards. Getting one super illegal Chem-Baron item was novel, but it just wasn’t flashy enough compared to the likes of Fortune which offers a bunch of random and cool stuff. While we’re going to take a break from this style of high-stakes trait with our next set, we should continue to make interesting cashout traits to see what works and what’s worth innovating in. Sugarcraft, high-stakes Augments, Chem-Baron—they’re all explorations in the space that have been loved by some, or just a few, and next set we’re trying another style of cashout trait—this one geared towards one high moment that’ll require some big brain pivots.  

Creating Multiple Choice Traits

Form Swapper was an exciting trait on paper that required a lot of extra work since we had to make two versions of each champion—but was it successful? Kind of. Success here would be seeing each champion played in each form during clear comp-building moments (e.g. Elise as a backline carry in a vertical Bruiser comp), but even when we had balance on both forms as close as it gets, the perception was for each Form Swapper, there was a correct and incorrect placement (e.g. Elise frontline almost no matter what). This may, in part, be social, but the reality is when one form is 51% successful while the other is 49%, the perception becomes one form is always better than the other.

Now, Form Swapper was our most successful of the multiple-choice traits, looking back at things like Storyweaver from Inkborn Fables or Sona from Remix Rumble. We saw both forms of Swain and Jayce often, but Elise and Gangplank were often seen just in one form of the other. Looking forward, we’re always going to have to deal with the rounding-up issue with these types of traits, but the challenge they present is enticing enough to keep chasing! 

Melee-Champion Power Scaling

This one’s a bit mathy, but pretty simple. Melee champs mostly survive by taking damage, gaining mana from that damage, and then casting their ability. Because of how much HP they gain when they star up, the difference between a two-star and three-star melee champ is quite large, as they benefit from the HP and the AD growth. This is especially an issue because melee-carry success is gated around their ability to survive, meaning they need those stats to succeed, but when they succeed, they serve as both a carry and a tank. The key examples from this set that had this star-based threshold were Smeech, Violet three-star (and four), Urgot three-star, and even Warwick, who as a 6-cost had the stats of a two-star 4-cost. Long term, we’re going to have a solution for this problem on a systems level, but it’s going to need a bit more time. 

Tanks and Durability

Tanks have four different types of spells: self-heal, crowd control, self-shield, or gain Durability (or Armor/Magic Resistance). Of these four types, gaining Durability is consistently the weakest because, by the time you cast, you’ve lost a lot of your HP, making the newly added Durability less effective. When Leona casts her ability at 100 HP, that 100 HP just becomes 150 effective HP; but when a tank like Blitzcrank gains a 450 HP shield at 100 HP, well, that’s just 550 HP. 

While Leona/Nunu were the key offenders here, there was one exception to this rule. Singed was able to find success with Damage Reduction due to the extra HP bars his BFF, Renata Glasc was throwing out (don’t tell him those shields were for everyone). So looking ahead at our tank class, we’re going to need to pay special attention to tanks that gain Armor/MR, or Damage Reduction to make sure they have ways to activate that early or alternative hooks that keep them viable throughout the fight. 

Festival of Beasts Revival

Hey folks, Riot Xtna here, the Product Manager behind Revivals (and a bunch of other stuff). We shipped another Revival during Into the Arcane, and this one was huge as it brought a fan favorite set into the modern era of TFT. The Festival of Beasts Revival was our third revival, and its performance blew us away. The Revival peaked at just under half of total TFT hours globally, but just because there were a ton of play hours, doesn’t mean there weren’t things to improve (ahem, Teemo, sadge). 

Modernizing Festival of Beasts had its challenges—how do we preserve the spirit of the original champion design without making the experience oppressively unfun for players? Players can no longer craft items like Zephyr or Trap Claw to counter abilities/traits as they could when the original set was live (shoutout to our old-school TFT knowers). Admittedly, we struggled a bit with combat pacing in Revival: Festival of Beasts, but we shipped large patches after its initial release to somewhat course-correct here. Expect us to continue to try our best to find that balance between the fun of playing ridiculous units from past sets and the modern principles that make TFT the game we love today.

Personal bags, or personal champion pools, were another change we tried this time around—one that leaned into the ‘play what you want to have a good time’ aspect of the Revival. We’re not quite sure if we’ll be keeping these for our next revival, but that decision will be made soon as we wrap up collecting feedback from the mode. Regardless of if you loved them, hated them, or didn’t care (valid), let us know what you think. 

Okay, the last part of Revivals that I want to talk about is the Revival Ladder, which is a more progression-focused competitive system that allows players to embrace the relaxed, for-fun nature of revivals but still rewards those who wish to climb. In the Festival of Beasts Revival, we added a new Leaderboard system to celebrate Visionaries and will be keeping this for next time, as it acknowledges the best-of-the-best in a way that tiers couldn’t do on their own.

Wow, that was long. And just like making TFT, I couldn’t have written this one alone, so a big thanks to the entire team that lent a hand in Into the Arcane and our Festival of Beasts Revival, and a big thanks to you for playing, providing feedback, and reading—even right now. We’ll have plenty to talk about with our next set coming in the future (from the future?), and we hope you’ll see these learnings in action for sets to come!