Behind the Convergence: Bringing Teamfight Tactics to life
Through trial, error and lessons learned, the TFT team solves a new puzzle each set
Update: An earlier version of this story incorrectly stated the TFT design team began working on Set 5 during the start of the COVID-19 pandemic.
Around 11 p.m. one night during the development of Teamfight Tactics: Reckoning, Stephen “Mortdog” Mortimer and Matthew Wittrock were looking through every skin in League of Legends so they could round out the set’s cast. The designers needed a melee champion with a skin that matched the powerful aesthetic of the Dragonslayer origin to replace Xin Zhao.
Mordekaiser sat right there as a potential option, but his Dragon Knight skin didn’t quite fit with the other champions. That’s when they looked at Infernal Mordekaiser. The skin had nothing to do with slaying dragons, but it gelled all the same. Finding that replacement took a bit of extra work, but Mort, the lead designer for TFT, said it gave Reckoning a memorable character. That’s why he said it was his favorite moment of developing TFT Set 5.
Solving the puzzle of developing a TFT set might not be as flashy as playing a fully optimized team composition in the game, but building both requires a constant stream of optimization from beginning to end.
A game of inches
TFT turned two years old on June 25, and product lead TJ Bourus said the team has grown in both size and ability. Not only have they done more with each successive set since launching the auto-battler, but he said they’ve gotten better with each one. The team’s work was enough to convince them to stick with sets in the first place, and committing only helped them hone the development process.
“When TFT first came out, we weren’t 100% sure that the set model was actually what they were going to do, so they didn’t plan to have this team and these sets of processes to make sets continually,” Bourus said. “So with [sets] three and four, we had a lot of pain points there, but we continued to improve it.”
According to Mort, the development process for a set of TFT now starts with a theme ideation phase, where any member of the team can pitch an overall theme. The pain points Bourus mentioned during Sets 3 and 4 meant those sets only received three and two ideas to work off of, respectively. However, after improving their process, the team came together to pitch 20 to 30 ideas for Set 5, with 10 seriously considered as options.
Theme ideas are also constrained somewhat by the kinds of skins available, though Mort said the TFT team has years worth of variants before they have to worry about using up all their options. And at the same time, ideas must also now consider previous themes, since the game has a good number under its belt.
Narrowing down the list is difficult, Mort and Bourus said, as not everyone agrees on what would make a good theme. One idea that came up for TFT Set 5 was Foodfight Tactics, which would use the Culinary Masters skin line to play off the pun in the name.
Bourus said he loved the idea. Mort said he wasn’t a fan, mostly because the concept would stop being funny about a week into the set’s months long lifecycle.
“TFT is always supposed to be a light hearted game and I get that,” Mort said. “But this feels like it crosses the line from light hearted to a joke.”
Once the frontrunners emerge, Bourus said the team surveys a small number of players to get their feedback and decide from there. And after that, the set designer heads off with a small team to create a paper draft that contains everything from champions, to spells to traits. To do so, they group the champions by skin, decide on origins and then figure out whether there’s any crossover.
For example, in TFT: Galaxies, Mort said the team really wanted to use Gun Goddess Miss Fortune — but that meant no Star Guardian Miss Fortune. As a result, Star Guardians lacked a physical damage dealer, which made turning it into a class of casters an easy decision. Once that’s done, the team has to decide on the high and low cost units as well as what spells they’ll use, which is when other conflicts arise.
“As you do that, with each trait, you might be like ‘holy crap, I’ve got eight casters in one cost. That’s not right,’” Mort said. “So you go back and start iterating from there. We always describe it as a puzzle, where it’s like you’re trying to make it all perfect and fit.”
Back to the drawing board
Despite how it may sound, Mort insisted the paper draft was the easy part of designing a TFT set. The hard part is implementing champions into the actual game, making sure everything works the way it’s supposed to and then identifying problems with playtests. Delivering that bad news, Mort said, is his least favorite part of development.
“That’s where the exhaustion comes in,” Mort said. “That’s where the hard work comes in, because ideas are cheap. Ideas are easy.”
Mort said the team has had to do major reworks after the first week and a half of playtesting twice now, which can be painful for everyone involved. They call the identification of these fundamental issues and the attempt to fix them “the big rework.” For Reckoning, the iteration process meant swapping out 15-18 champions and three or four origins.
“That’s where creative problem solving kicks in, where it’s like ‘what’s the best version that doesn’t literally go ‘start over’?’” Mort said.
At the same time, a different portion of the development team is supposed to work on a set mechanic that changes the core gameplay beyond new champions. This has included Elemental Hexes, Galaxies and the Chosen system, though Mort acknowledged that the team isn’t happy with every effort.
“Frankly, on our set mechanic side, we’re not batting a great batting average,” Mort said. “A lot of our set mechanics have been started and then scrapped and then curveballed.”
For TFT’s third set, Galaxies were a last minute change from a failed day and night cycle. Even TFT Set 5’s Shadow Items were a late addition, as Wittrock took on the responsibility alongside creating the paper draft due to staffing issues. They ended up with three or four options, but the only one the team really liked fit much better with the upcoming Set 6 theme.
“We kind of had to last minute scramble on this one, again, kind of like Galaxies,” Mort said. “We came up with the idea for the Shadow Items, and that was definitely one of those times where it looked good and we’re out of time, so if it doesn’t work we’re in trouble.”
The closest any mechanic ever cut it to the release of a set were Lucky Lanterns for the TFT: Fates’ mid-set expansion, which they started working on three weeks before it hit the Public Beta Environment. Shadow Items were next, with five weeks from Mort’s first draft to going live.
Much like the Chosen, Mort said the team was nervous about the community’s reception of Shadow Items. There was talk of delaying both mechanics until they had more time to test them. However, they ultimately trusted their guts and the live team’s ability to problem solve as needed.
The entire development process for TFT Set 5 was made even harder by the COVID-19 pandemic, which forced all Riot Games employees to work from home during the development of Reckoning. To make matters worse, Bourus said the TFT team in particular worked around impromptu conversations based on their relative proximity to each other in the office prior to the pandemic. Out of any other team at Riot, they relied on Slack and internal messaging tools the least.
With everyone separated, Bourus had to help clear a new communication hurdle. He said the solution included introducing formalized processes, like the kick-off meeting that starts the brainstorming for each set.
“It was probably one of the most stressful and hardest times to communicate during this period,” Bourus said. “But we did it, and we learned a lot about how to do it right that we’ve incorporated into future sets from home, which is really exciting.”
As a result, not only did the TFT team deliver on Reckoning, but they worked in harmony as a whole. Bourus said the entire organization, from marketing to Little Legend designers, came together as a whole.
“That unification of the team is what led this to feel like one of the most cohesive sets yet,” Bourus said. “I definitely want to shout out the entire TFT team, but also want to call it out for folks that this is a change we actually did differently.”
Innovation never ends
Not only did the TFT design team change the way they worked together, but they continue to change the game itself in novel ways. TFT Set 5 saw the introduction of labs, new ways to experience TFT that meet the needs of different players. The first lab is called Hyper Roll, and it condenses the normal game experience into a shorter time frame, so players can get in and out much easier.
This falls in line with what Bourus said is the TFT team’s ultimate vision: to create the strategy game of all strategy games. But to do that, they need to serve other motivations. Hyper Roll serves those on the go, but Bourus hinted at other considerations — like those that want to play the game as a group. Regardless of who they’re trying to serve, exploring those possibilities was an early goal for the design team.
“We’ve known we wanted to do modes for TFT for a long time,” Bourus said. “It was just figuring out which one was our first one, which one was worth the initial investment to get us a good taste of whether players are interested in alternate experiences.”
According to Bourus, he knew Hyper Roll would be a hit the first time they showed it to Riot employees outside the TFT team. He said “early risers” expressed how it was the perfect way for them to experience the game.
At the same time, Mort stressed that labs are often misunderstood by the community at large. He has seen players talk about them like rotating game modes with fanciful concepts like starting with 50 gold or starting at level nine. In truth, Mort said any lab they spend time developing is something the team thinks has the potential to become a part of TFT forever. That’s why, at the least, Hyper Roll is sticking around through all of Reckoning.
TFT has evolved in other ways, too. Mort said the team has gotten close to optimizing the way they design sets, from the number of units per cost to the ideal shop percentages at every level. But in a more big picture sense, the design team has also realized some truths about TFT that now guide their decision making.
“We’ve realized that the core of TFT is already deep and engaging and complex and interesting, and that Fates, Rise of the Elements and Galaxies were all basically like, ‘here’s a curveball on top of it,’” Mort said. “It kind of took away from the core. So right now, we’re not looking to keep throwing layers on layers on top.”
Mort described this change in philosophy as a willingness to challenge the design team’s core assumptions about TFT itself. He added how important that is as the game gets older and change aversion starts to creep in on everyone involved. After all, it’s riskier to introduce radical changes to the game now than it was a year and a half ago.
“That’s the kind of space we’re getting dangerously close to,” Mort said. “You see this with League of Legends, for example, every time they try to do something big… players have this really gross change aversion. That’s going to be our big challenge.”
Aversion aside, Mort said he doesn’t think TFT can survive for 10 years on just a steady flow of new sets. While he doesn’t know what might be necessary to ensure TFT’s longevity, he can turn to other examples of innovative success.
“I think in today’s industry, novelty has to start to play a factor as well,” Mort said. “Even with something like Magic: the Gathering, you’re seeing this with the Commander style and how that’s been dominating.”
While many more years lie ahead before anyone on the development team has to worry about crossing that bridge, Bourus said he isn’t worried about the game’s future. After seeing the reception for Galaxies and Fates, he trusts Mort and the rest of the team to develop a great game.
“I just have so much faith now in what TFT can and has become,” Bourus said. “Although there’s pressure… I have a lot of confidence. I don’t feel the pressure is on the team to crush it. I think we’ve proven to Riot, we’ve proven to players that we can make awesomeness.”