6 August 2020 Infinity Gauntlet

The Making of Infinity Gauntlet: A Love Letter™ Game

How Z-Man Games designer Alexandar Ortloff gave the classic microgame a Marvel twist

Since Love Letter was first designed by Seiji Kanai and published in 2012, the game has gained massive popularity as the definitive “microgame.” With only 16 cards, Love Letter creates a clever, dynamic experience that I have played over and over with just about everyone I know. Needless to say, I was excited when our studio acquired the publishing rights back in 2018.

When we released the new edition in 2019, there was a deliberate effort to avoid making significant adjustments. The new Love Letter would include a handful of new cards, which made the game playable for up to 6 players (instead of the original 4). These were small, careful changes intended to increase the player count but otherwise maintain the classic Love Letter experience. But with that project complete, the question became: “What next?”

Given the popularity of past Love Letter spinoffs and our newfound partnership with Marvel, the way forward was pretty clear. But it would be some time before the project started to resemble Infinity Gauntlet: A Love Letter™ Game as we know it today. Going into the project, we weren't interested in merely changing the names and art on the cards. A thoughtful reskin has its place, but the Marvel license brought a new kind of story to the table, and we wanted to make sure that the mechanics of the game conveyed that theme. This project didn’t need to be subtle and understated—quite the opposite. From the outset, we gave ourselves permission to adjust the rules of Love Letter if it would help evoke the right feeling.

Getting Started

As I considered various angles to approach a Marvel Love Letter game, I realized pretty quickly that the stories I gravitated toward didn’t fit the competitive free-for-all format. I was more excited about heroes teaming up than fighting each other, so I started exploring alternatives. My first prototypes were variations on fully cooperative play, with a greater focus on deduction. While a co-op Love Letter might work, all of these versions felt too calculated and methodical. I wanted a game about superheroes to feel action-packed, and that wasn’t coming through.

At the time, I was aware of the Infinity Gauntlet comic, but the central dynamic of that story is a bunch of heroes facing off against one main villain. This would naturally fit a one-vs.-many game, but I had been actively avoiding going down that road. One-vs.-many can be hard to balance, especially when accommodating different player counts. It can also be easy to accidentally relegate the “one” to facilitating the fun for the “many.” I didn’t want one player to have to pull punches to maintain balance, so I steered away from the concept entirely.

Lucky for me, this was all happening in April of 2019. I saw Marvel Studios' Avengers: Endgame in theaters opening night, and watching the movie and fan reactions convinced me it would be worth the effort to get this right. The next day I had a new prototype and we got to work.

Good vs. Evil

So, I knew I wanted one team playing heroes and the other side playing Thanos. The simplest solution would be to give each team their own deck filled with cards that fit thematically. Two decks would also let me differentiate the playstyles of the different sides. In the stories, Thanos has all the power, and it takes careful teamwork for the heroes to win out. To capture that dynamic, I needed to give Thanos some advantages.

First, I gave Thanos 2 cards in hand instead of 1. Just 1 extra card made a significant difference in the control the Thanos player had over their plays. It also made the Baron effect from original Love Letter (comparing cards) more exciting—even if Thanos had a high card in hand, there was always a chance the hero might choose the other, possibly lower card to compare against and win. It created moments when the heroes know the odds were against them, but it was worth the risk to keep fighting. With this change, the Baron effect would become a cornerstone of Infinity Gauntlet.

The other advantage I gave Thanos was the Infinity Stones. Thanos would need powerful cards to match up against a whole team of heroes, and I loved the idea of Thanos gathering the Stones and playing them for devastating effects. As a hero player on the receiving end, the Infinity Stones feel unfair, which seemed exactly right for the theme.

Of course, with the Infinity Stones came the “snap.” The answer to “What happens when Thanos gets all of the Stones?” turned out to be as simple as “Thanos wins.” It put the hero players on a clock—they would have to defeat Thanos quickly, because if the Thanos player had time to draw through their deck, they were guaranteed to draw all six Infinity Stones. Thanos was, appropriately, inevitable. That turned the hero experience into a desperate race, with tension mounting each time Thanos drew a card. Since Thanos could always come back from the brink of defeat, they could take minor losses in stride. However, since Thanos could play only 1 card per turn and only ever have 3 cards in hand, there was a buffer in the early turns where Thanos couldn’t snap yet. Once Thanos had a few Stones in play, players could see when they were in danger of losing.

Down But Not Out

There were two other fundamental departures from Love Letter early on. The first was the removal of player elimination. I knew I needed rounds to go longer if the snap was going to be relevant, and that wouldn’t happen if players were getting knocked out. (It was also just more fun for all players to keep playing.) Instead, I gave each team a life track—where original Love Letter rules would knock you out of a round, instead your team lost 1 life, you drew a new card, and the game went on. This made each individual defeat less painful, but also left you with a visual reminder of how close you were to losing.

The second major change was the removal of game rounds, and that was more of a realization than a decision. After playtesting for just one round, we all agreed it felt like a complete experience. If we wanted to play again it was easy to do so, but the arc of the story had played out in full and requiring the players to reset and keep playing felt unthematic. I dropped the rounds (and tracking tokens) and never looked back.

Card Effects

With that solid framework in place, the rest of the game would come down to the individual card effects. From the beginning, there was a challenge: by including 2 decks, I had almost doubled the number of cards and tripled the effects from the original design. I was comfortable with a little more complexity, but this threatened to be way too much. One of Love Letter’s great strengths is how approachable it is to a wide variety of players, so at every stage of design I had to ask myself, “Does this complexity justify itself?” and, “How can I get the same outcome more simply?”

In the case of card effects, I addressed the complexity by drawing parallels between cards, decks, and even between Love Letter versions. Wherever possible, cards of the same number in the hero and Thanos decks have the same effect or serve similar functions. The Infinity Stones are more powerful versions of the non-stones of the same number. Any time an effect from original Love Letter appears, it is at the same number as in the original game. So, a “1” in the hero deck has the same effect as in the Thanos deck, the “1” Stone is a better version of that effect, and they all trace back to the “1” effect (the Guard) from Love Letter. This approach allowed for a cohesive core that would give you an idea of what to expect from a card and build on your expectations if you were coming from the original game. Of course, this couldn’t be true of every card, but that familiarity made the other differences easier to handle.

The initial prototype pulled a lot of effects directly from Love Letter, but playtesting quickly showed that some would need to change. The most impactful change was the Handmaid effect (“4”). In original Love Letter, the Handmaid protects you from effects until your next turn. For the heroes in Infinity Gauntlet, this didn’t do much—Thanos could just target a different player and have the same impact against your team. For Thanos, being invincible for a full turn cycle was incredibly powerful and could essentially waste up to 5 hero turns. As often is the case, the solution came during a playtest.

At that point in development, 3s and 6s allowed players to use the Baron effect from Love Letter. You compared your hand to one of Thanos’s cards, and the lower number was defeated. Players asked for a way to improve their chances, and that led to the creation of power tokens (and prompted us to name the Baron effect “fighting”). If you could improve your odds in a fight, you could be more aggressive, and an opponent fighting you became riskier. This element of protection felt natural for a “4” effect. At this point, the round tokens were already gone, so a different type of token felt right at home. Power tokens would go unchanged through the rest of development.

On the Thanos side, the card with the strangest trajectory was the Time Stone (“6”). Unlike the other Infinity Stones, the Time Stone was going to be the only card of its number in the deck, so it needed a unique effect. I approached it more thematically, and the idea of copying a previous card felt like a natural way to represent time manipulation. In testing, though, it was proving to be too powerful. I explored other versions, including one where the Thanos player regained a life they had lost (which turned out to be even more powerful and less fun). Frustrated, I eventually returned to the original version, and it turned out that enough other elements had changed in the meantime that it was no longer a problem.


From the very beginning, I was confident I didn’t want heroes to openly discuss what cards they saw or tell other players what to do, but it took some time for me to articulate why that was important. What it came down to was maintaining the feel of Love Letter, and keeping players engaged.

In Love Letter, you are constantly trying to deduce everyone’s hand because everyone is your opponent. In Infinity Gauntlet, the heroes have only 1 opponent, so that could mean that they have only 1 hand to deduce. If someone else deduces it for you, your turn is purely mechanical. You lose out the surrounding mind games, and one of the most interesting aspects of Love Letter disappears.

Instead, I wanted players to be deducing the Thanos player’s hand and each other’s. This way, you always want to pay attention to everyone’s decisions. If you can infer what your teammate has, you can set them up for a powerful play, and it’s satisfying when it works out. Marvel’s heroes are mighty, but not always known for their communication skills, so the thematic side didn’t worry me too much.

As All Things Should Be

By far, the most involved part of Infinity Gauntlet development was getting the balance right. Love Letter has a lot of built-in variance, with 1-in-7 guesses that can turn the tide of a game. One-vs.-many games have their own challenges with player scaling and coordination. Unsurprisingly, there was no one trick that made everything fall into place—the answer was playtesting and time. Between external and internal testing, we played hundreds of games across all player counts, and then used the aggregate data to get a sense of where the balance was. Nudge an effect, grind more games, repeat.

Alex's playtest tracking sheets.

One of the biggest challenges I faced was tweaking the balance without introducing additional complexity. In a heavier game, it can be easier to vary setup steps or add new pieces to solve scaling issues. At one point, I tried a version that scaled the number of Outrider cards in the Thanos deck to the number of heroes. This did a decent job of maintaining balance, but wasn’t as “plug and play” as the game needed to be. I wanted you to be able to shuffle the deck, draw starting hands, and play. With that goal in mind, I dug back into playtesting, making minor tweaks back and forth and generally driving my testers crazy. Eventually a combination of scaling Thanos’s life and adding one more Outrider to the Thanos deck (at all player counts) got us to the right balance.

With Great Power…

There are more little stories and decisions than I have space to talk about, but every choice we made was aimed at making the best game possible. Seiji Kanai and Love Letter set a gold standard for a light, quick card game, and Marvel is beloved the world over. I was a huge fan of both, and I wanted to do them justice, so there was a lot to live up to with this project. In the end, I’m really proud of what we made. I can’t wait for players to get their hands on Infinity Gauntlet soon, and I hope it expands everyone’s idea of what a Love Letter game can be. Thanks for playing!