Autochess – The Rise & Fall of a New Game Genre

I have never seen such a meteoric rise to popularity as Autochess. Starting as a mod for the popular game DOTA 2, it quickly grabbed the attention of thousands, probably millions, of players. The gameplay is unique, the strategy and the economy are complex and rewarding, and it enjoyed the benefit of using all the popular characters that already exist in a very popular game.

Then it fell into obscurity as fast as it rose. As of this writing, the most popular game in the genre is Teamfight Tactics by Riot Games, and the biggest competitor, Underlords by Valve, is almost a meme. So, what happened?

Well, a lot of things happened, and the best way to investigate is to go over chronologically – which is what we will do today.

Genesis – The Humble Beginnings

In January 2019, a game mod developed by a single developer, Drodo, came out in the DOTA 2 community. Labelled as “Autochess”, it had a grid board that is similar to chess but the characters were fighting automatically. Furthermore, you decided what characters you want to place on your board! Through a unique economy system that is both simple to grasp and allows for a lot of strategy, it rose to popularity very quickly. Players were spending a lot of time understanding the intricacies of this new game and drawing attention to the original game it spawned from, DOTA.

It was almost like the start of the original DOTA – a mod for a popular game (Warcraft 3) that spawned its own genre (MOBA). League of Legends is the top game of not only this genre, but the entire world in a holistic view of popularity and longevity. This triggered the fear of missing out (FOMO) of all the big companies, who wanted to be the top dog in this brand-new genre.

I believe this FOMO is one of the core reasons why the game shot up in popularity and crashed soon after.

The original developer was invited to Valve HQ, and after discussions, Drodo and Valve decided that working together is not feasible for them, but they would be supporting each other in developing games in this genre. Drodo went on making a standalone game, while Valve started working on their own title.

The Battle of 2.5 Kings – Developing a Game and Failing Fast

During this time, Riot Games also started working on their own title, Teamfight Tactics in the Autochess genre. There were now two giants, Valve & Riot, rushing out to build their Autochess game while the original developer went the route of building a cross-platform game while partnering with Epic Games for publishing.

This, in my opinion, was the point of failure for Drodo and the original Autochess. I can not know the exact reasons why they could not work together with Valve, and I fully respect his decision to build his game his own way and retain ownership, but he made a few key mistakes. First of all, making a standalone game is far more difficult than making a mod. Second, there was no next-iteration in his development; at the end of the day, it was just the same game/mod with a separate installer. Third, going on a separate platform created more hurdles for players to jump into the game, and no matter the power of your partner (Epic Games), players dislike hurdles and hate having to open new accounts on new platforms. The one massively positive decision was the idea to go to mobile platforms for his game; however, even then he fell into the age-old trap in converting PC games to mobile – games that are not designed initially with mobile in mind have massive difficulty catering to the needs of the mobile audience.

While all of this was going on, Riot and Valve were rushing out their prototypes and trying to get their Autochess games in front of the players. This was a very unique development period; the game design, character models, and story (or lack of it) were already ready and it was just writing the backend code to replicate it. Riot used League of Legends characters and Valve DOTA characters, on a minimal scale. However, as they are established game companies with brilliant minds amongst them, they knew what needed to be done – their games needed a differentiating factor. Furthermore, they had to come up with that differentiating factor as fast as possible as many other games were popping up everywhere. Instead of choosing the best idea overall, they had to go with the best idea they had at the time, and fully commit to it. They came up with these ideas:

Teamfight Tactics / Riot: Adding a carousel and item system.
Underlords / Valve: Adding a special character (an Underlord) to choose for players before the game

Riot’s idea was better, because it was safer. They added entirely new systems to the game, which meant players, if they wanted, could completely ignore it and still get to enjoy the game. Valve’s idea on the other hand hinged on impacting the existing system. An imbalanced Underlord or a bug in the system would completely prevent the players from enjoying the game.

Have you noticed how Underlords truly died once the actual Underlords appeared in the game?

During this development period, both TFT and Underlords saw huge player counts, but Riot increased its lead day-by-day as they ironed out all the issues with the systems and the code, while Valve struggled with making their addition to the game work, feel balanced, and more importantly, fun.

Death of a Genre & Finding the Cure

So now we understand how TFT “won” the Autochess development race, but what happened to the actual genre? Currently the biggest name is TFT, Autochess is nowhere near it and Underlords are completely abandoned. It came to this stage in the span of a few months, even before the end of 2019. The answer lies in two important factors that none of the developers above could think about in the rush to get to the market as fast as possible: Long term engagement and monetization.

Long-term Engagement:

A typical match of Autochess takes around 45 minutes. After that game is won or lost, what is the thing that would get players to press “play again”, except the fun they get out of the game? In some cases that is enough, but in the case of a Free-to-Play game that definitely is not enough. Riot introduced the seasons eventually, and introduced a Battle Pass that gives benefits in LoL for playing TFT. Autochess tried introducing new game modes, and it was actually a great idea to introduce a fast-paced mode, especially for the mobile audience they originally wanted to pivot to.

There are of course some simple progression systems like ranked modes etc., but such a groundbreaking innovation in game design kind of needed a new and unique system of progression in the first place. Maybe if the popularity of the genre stayed as high as it was in its hay day, there would be more time and brainpower dedicated to coming up with such a system but even TFT feels like it is in “maintenance mode” and the innovations are kept within the limits of relatively small changes in seasons, rather than new systems.

Monetization:

Ah, the necessary evil that players hate with a passion – how games make money. What players fail to realize though, is that as games make money, they get more investment and more long-term support. For your favourite game to survive the test of time and maybe even come up with sequels, it has to generate revenue.

If I had to give one reason and one reason alone for the fall of the Autochess genre, I would say it is the difficulty in monetization.

Autochess is a wonderful game, that is impossible to monetize with the existing popular methods of monetization. You can’t sell characters as it would break the game. You can’t charge a subscription for it as it would immediately make a lot of people switch to a competitor – especially in the early days. Cosmetics are an option, but Riot games made the board an actual part of the game mechanics, so customization is difficult. Battle Pass is the one with the highest potential of working, but even then Riot had to include benefits for their original game to make the Battle Pass even remotely worth buying.

As a result, the genre is struggling to generate revenue, which in turn prevents more resources to be diverted to it, which means not enough time to create a new monetization method,… It is a vicious cycle.

Then what is the solution? An entirely new method of monetization is needed for the Autochess genre. I am going to share my idea for a monetization system. It may not be perfect (in fact it most likely is not), and even if it was perfect it may already be too late to introduce it after 3 years. However, I would not want to just complain and point out flaws without a solution.

The Silver Bullet

Among many different strategies, I wanted to focus on one that brings the game design closer to something that we can monetize with the existing methods. I want you to think about the different game modes in Magic: the Gathering. There is a standard mode, and then there is a Draft mode.

Autochess Draft mode: This mode is played with a subset of all the past and current characters in the pool. 8 players come in, and each player takes turns picking one character from an option of 3. Everybody sees what options are provided, and sees the current pool of heroes. The system loops through players until the number of characters is equal to the pool size, with the correct distribution of Tier 1-2-3-4-5 characters. Then the actual game starts, but it will require a completely different strategy than a preset pool. Maybe an old overpowered synergy is chosen by the players, or maybe there is no synergy and the player that can make the best of the situation will have the advantage.

How to monetize this mode: There is a ticket system, in which tickets are sold for a small price. Each ticket gets you 3 play sessions, and if you manage to end in the top four in two out of these three sessions you get another ticket. If you are a good player, you can go “infinite” and play the mode without paying anything. On average, players will be buying a ticket every 10-12 matches. What is good about this system is that it is very easy to balance, so we can generate revenue without being predatory by easily changing a few parameters – maybe you need to win 3 out of 5 sessions per ticket, or you need to finish in the top 3 instead of 4.

Again, this is just my idea of monetizing in Autochess, without the need for massive new development or the creativity to come up with a brand new monetization system or game mode. As I mentioned, it may not be perfect but offering a solution to the biggest challenge of the genre would benefit the genre and the many players that love it. Including me; I used to enjoy Autochess a lot and would love to get back into it knowing it will get the resources and investment the genre deserves for months and years to come.

Conclusion – A lot of Lessons to Learn from One Genre

I hope you enjoyed this story of how a new genre of game took the world by storm and then failed to capitalize on it partly due to how fast everything escalated and partly due to the challenge of monetization for such a unique game. Regardless of how long TFT remains alive, the Autochess craze is an event that game developers and publishers can learn a lot from.

Monetization is crucial, no matter the size of your company or audience. It is too important to be an afterthought.
FOMO is very real, and correct navigation of it is difficult even for the biggest game companies. Development must be fast, but choosing which corners to cut and which areas to give due attention can and will make or break your game.
Solo development is quite possible, and you might be the one to come up with the new Autochess. Solo publishing is a whole other beast, and choosing the correct publisher and platform will be the difference between you establishing yourself as the leader of the genre or watching from afar as others build successful games from your idea.

There are a lot of micro-lessons too to go along with these. That is why I find the story of the Autochess genre fascinating, and everybody should spend some time to research and understand: “What went wrong with Autochess?”