Over the weekend, Dota Underlords lured 200,000 concurrent players onto its checkered battlefield of swords, sorcery, and slot machines. Meanwhile, Riot’s Teamfight Tactics continued to duel with Fortnite for the Twitch throne. Also, I played Dota Underlords like it was my dang job. If it feels like these Auto Chess games have taken over the world overnight, that’s because they kinda have.
Unlike battle royale, which existed in various forms for years before ascending to video game royalty with games like Fortnite and PUBG, Auto Chess as we know it actually started in 2019. It began with Dota Auto Chess, a Dota 2 mod that came out on January 3. Less than two weeks later, it was one of the single most popular games on Steam despite being a game within another game. Months later, two of the biggest companies in video games, Valve and Riot, have released their own variations on the formula.
There is no great mystery as to how this genre took off. There’s now a well-oiled pipeline that helps propel these things into the zeitgeist. Steam’s Workshop gives exact numbers on how many people are subscribed to a certain mod. Twitch tells us how many viewers are watching people stream it. On both these platforms, games that do big numbers get more prominent placement. Media sites obsessively track these numbers and report irregularities and spikes. These processes and systems have been optimized even more than they were a couple of years ago, when PUBG came out of nowhere to turn battle royale games into a sensation. Popularity has always begotten more popularity, but now it does so in record time. It doesn’t hurt that, in the Fortnite era, players, streamers, reporters, developers, and business execs are all on the lookout for the next big thing. In Auto Chess, it’s possible they’ve found it.
The question, then, is what particular elements make Auto Chess—or the auto battler genre, as people have taken to calling it—so universally appealing as to inspire an initial fervor that not even chess, the sport of kings, or Dota, the sport of people with an annual opportunity to become kings, could rival. It’s one of those things that makes immediate sense if you just play these games, which of course means it’s riddled with intangibles that are nearly impossible to articulate. Time to articulate them.
Despite being attached to notoriously difficult-to-get-into games like Dota 2 and League of Legends, auto battlers are remarkably simple. At the start of each round, players buy units. Players can then combine those units to make stronger versions of the same units. Certain types of units get bonuses for being on the field at the same time. You deploy these units, and they fight each other automatically, with no player input involved whatsoever. Units’ stats and special abilities and the way those abilities counter opponents’ units determine who wins and who loses. The remaining units in the winning army then do damage to the loser’s central health pool. If you lose all your health, you’re out.
There are eight players per match, and each of you square off one-on-one in a mini-tournament sort of format until only one is left standing. Those are the core pillars of the entire genre. You only really need to know those facts to do alright in your first match. (You won’t win, mind you. It took me, like, five hours to win my first Dota Underlords match. But you won’t drown, either.)
MOBAs like Dota 2 and League of Legends require at least a degree of mechanical skill. You need reflexes, timing, and sometimes even accuracy in addition to tactical smarts. You also have to cooperate with other people, flesh-sealed nightmares that leak feelings and opinions everywhere. Auto battlers strip away many of those elements. They’re multiplayer, yes, but you are your entire team, and all you’re doing is making decisions. There are no expectations about the number of actions you take per minute, nor are there teammates around to yell at you for not meeting those expectations.
Technically, yes, you’re squaring off against seven opponents in a battle royale where you can get stomped and embarrassed if you’re not able to think on your feet, but it’s all relatively chill compared to other multiplayer games. You just sit back, make your moves, and watch your tiny action figure army do its best.
Auto battlers use their first impression to inspire intrigue. There are so many different types of units, and they form rewarding synergies. It’s fun to just toss out a few same-type units just to see what happens. Maybe you go with a couple of druids in Dota Underlords, only to realize that one of them also falls into the “fierce” category, meaning they have a stat-boosting synergy with your tanky warrior unit. Suddenly, with just three units, you’ve got the backbone of a strategy. You can get away with experimentation like this, because the early game is pretty forgiving. Units respawn after dying, and nobody’s dealing much damage at the beginning, so even if your fledgling army gets wiped out a few times in a row, your central health pool isn’t in much danger. You can still mount a comeback a handful of rounds later, after you’ve amassed more resources and better units.
Before too long, you may start asking other questions. How, after only a few rounds, do some of your opponents have multiple two-star units? Or even a max-level three-star unit? (Stars represent statistical strength.) How can you do that? In my case, my favorite early Dota Underlords discovery was one of the many answers to that question: those druids. If you have at least two druids on the field, your lowest-level druid automatically gains a level. In the early game, this feels like cheating. Every two-star unit is a game-changer in the first handful of rounds, and you can get an extra one with basically zero effort by plopping down a couple druids. I cackled the first time I did it, and then I cackled even more when I leveled those druids up to two stars the regular way, by combining units. Then the lowest-level druid on the field was a two-star and, because of the aforementioned auto-leveling rule, I suddenly had a three-star world-eating mega-hero relatively early in a match.
Auto battlers are absolutely ridden with strategies like these, and each one feels like the sort of revelation somebody has when they decide to become a professional card shark. “I can get away with that?” you will ask many times while playing these games. “And I won’t get arrested?”
Fun tactics are all well and good, but the secret behind the explosive popularity of Dota Underlords, Teamfight Tactics, and their ilk is gambling. No, auto battlers don’t try to vacuum real money out of your wallet (at least, not mid-match), but the start of each round sees you buy new units from a slot-machine-like system. Don’t like what you see? Drop in a couple gold to re-roll and get a new selection of heroes. Maybe you’ll luck out and get two of a hero you already have, allowing you to upgrade that hero to a two-star. Or maybe you’ll finally get the OP hero who’s the linchpin of your whole strategy (hi there, Kunkka). Then you’ll want to start upgrading them. Maybe you’ll break the bank and get nothing at all. But at the end of your turn, win or lose, you’ll get more gold, at which point the cycle begins anew.
There a countless meta strategies within this system. For one, you can gain interest on gold if you accumulate enough (one additional gold for every ten gold you hold onto per turn in Dota Underlords, for example). It’s often worthwhile, then, to sandbag for a few rounds and build up a healthy chest of glittering prizes. Then you can sink a ton into re-rolls until your army soars right past everybody else’s. All the while, you’ll want to keep an eye out for crucial heroes—or even just copies of the scrub heroes you put onto the field during the first handful of rounds. You’ve got to plan ahead, even if it means lugging around a bunch of useless copies of Tusk, the walrus-punching warrior who is also a walrus, because eventually, in 10 or 15 or 20 turns, you might be able to turn him into a three-star. Then you’ll have a massive upper hand.
The randomness of these games can be insanely frustrating. Occasionally, it can feel like you lost to the fickle RNG gods, idly tugging at the strings of your fate from inside their cushy digital heavens, instead of clever opponents. Most of the time, though, RNG giveth as much as it taketh away, and the rest is about smart planning and intelligent split-second decision-making. Depending on how long you last, games can go for 40+ rounds, so over time you’ll have plenty of chances to get new units. This means that every game—even if you’re planning to employ the exact same strategy as last time—is a new adventure.
This, in a nutshell, is the magic of auto battler games. You never entirely know what to expect, but you’ll slowly learn to control and even harness the chaos. Figuring out the inner workings of this infernal slot machine is diabolically compulsive. You lose one match, but you figure out how to slightly optimize your strategy for a particular build. Or you realize that if an important unit had appeared just a few turns earlier, you’d have wound up duking it out for first place instead of slumming it in the middle of the pack. You play again. Matches usually only last 20-30 minutes, so what’s the harm? Maybe this time, you get different units early and end up accidentally figuring out a better way to implement your preferred strategy. Or perhaps your early-game unit selection forces you to employ a different synergy strategy entirely, but you realize you like that one more. You do slightly better than last time. You learn a few more lessons. So you play again. And again. And again. Suddenly, your whole Saturday is gone, and your previously-made Sunday plans are starting to look awfully optional.
For a few months, the Dota 2 Auto Chess mod was the main auto battler on the block. Released in January by Chinese developer Drodo Studio, it was regularly played by hundreds of thousands of people at a time on Steam. As of now, it has nearly 10 million Steam subscribers.
It wasn’t exactly shocking, then, when Valve announced its own auto battler, Dota Underlords, last month. In fact, the only surprise was that Valve wasn’t working with the mod’s original developer, given the company’s history of hiring mod developers like it did with Dota 2’s own designer, IceFrog. Instead, Valve and Drodo decided to make their own standalone auto battlers with each other’s respective blessings. Drodo’s Auto Chess shed the Dota license (but little else) and appeared on mobile devices earlier this month. Drodo will also be releasing a PC version on the Epic Games Store in the future. Dota Underlords, meanwhile, entered open beta on PC and mobile last week. It immediately pulled in hundreds of thousands of players on Steam. Despite an interface that’s well-suited to mobile, however, neither Auto Chess nor Dota Underlords are currently in the iOS App Store’s list of top-100 free games.
League of Legends developer Riot also caught wind of the auto battler trend and similarly adapted its pre-existing MOBA to suit the formula. It announced Teamfight Tactics earlier this month and begin testing it on PC shortly before the release of Dota Underlords. While Riot isn’t as transparent with its numbers as Valve, Teamfight Tactics has been a huge hit on Twitch, which means it’s making waves. Twitch success, however, is not necessarily indicative of a gargantuan player base, nor does it inherently portend long-term interest.
For now, both Dota Underlords and Teamfight Tactics are surfing on waves of early momentum despite the many kinks that come part and parcel with beta releases. The question now is whether they’ll be able to maintain that momentum, especially given that these are far from fully featured games, given Valve and Riot’s apparent desire to capitalize on the trend as quickly as possible.
Because video games.