Well this genre just kind of exploded didn’t it? Technically it has existed for decades as a Warcraft 3 custom map, ironically the same place we got the entire MOBA genre from as well, and honestly many of the aspects of this developing genre remind me of the early days of MOBAs as well.
This blog is going to focus on the differences and similarities between all the current Auto Battlers, though now that it’s written I still have other concepts I want to return to and look at more closely in the future.
Dota Auto Chess: The Custom Game
The original Auto Battler (well, original for the sake of argument here) is the Dota Auto Chess custom game mode made by Drodo Studios, who also made the Dota 2 custom game modes GemTD, a tower defense game, and SkateMaster, a sort of pathfinding game. Their original version of Auto Chess is a pretty interesting game, especially since it’s built with the framework of a Dota 2 custom game. In Auto Chess players control a courier, who acts as their avatar on the board and who has a limited range of influence (more on that in later sections). The goal is to be the last of 8 players still alive by assembling a team of powerful Dota Heroes that work well together. Each hero has a race and a class (though some have two races) that determine how they work together. Some are rather exclusive, like mechs who increase the health regen of all your mechs but not your other heroes. Beasts, on the other hand, increase the attack speed of all your heroes whether they’re beasts or not.
The game also focused on a drafting element. The heroes you’ll see in the shop are randomized, but there is a fixed number of each of them in the pool that all 8 players are sharing. This shared pool is key to the multiplayer aspect of the game. In addition to using synergies to make you overall team stronger, each player also wants to pick up multiples of each hero to combine together and create level 2 and level 3 versions. Players must also manage their gold, the primary resource in the game. Gold is used to buy heroes, to reroll the random heroes you see in the shop, and to buy experience to level themselves up. Leveling up a player enables them to place an additional unit on the field, up to a max of 10, and also increases their chances to see more expensive and powerful heroes in the shop. Gold is earned at the end of a round, whether the player won or lost, with bonuses assigned for victory, win streaks, loss streaks, and interest. Interest is based on the amount of gold each player has in their bank at the end of each combat, earning them 1 gold for every 10 stored up, maxing out at 5 total interest.
Additionally players can empower their team with items dropped by neutral enemy creep waves every 5 rounds, though these items are quite random. Some creep waves can pass dropping no items, some dropping 3. The mod also allows many items to be combined into stronger tools. The only downside is that items can never be removed from a hero unless you sell the hero back to the shop, freeing up the item for another hero to take.
It may seem like I went rather in-depth there on the basic mechanics of the original Dota Mod, but frankly everything I said also applies to the three primary Auto Battlers that have been released so far.
The Drodo version, simply called Auto Chess: Origin is currently available on the IOS and Android app stores and has a planned PC release on the Epic Game store. It also already has a 1 million dollar Invitational tournament in Shanghai this October. For the sake of clarity this is partially because Drodo are not the only people developing their version. ImbaTV, an esports company who is hosting the invitational, as well as Long Mobile are doing some portion of the development of Auto Chess: Origin with Drodo. This is part of how Drodo’s 5 person team is also releasing regular updates to the original mod at the same time as developing their standalone version.
Auto Chess: Origin is frankly the closest of the three current games to the original, which does make sense. The character’s names have been changed and many synergy names have been tweaked. For example, trolls have been renamed the glacial clan and orcs are now the red cave clan, however the actual synergy effects are the same. At the time of writing Auto Chess: Origin is missing some heroes from the mod. Notably missing is Io, a hero who acted as a wild card or gap filler for leveling up heroes. Also missing are the gods, relatively recent additions to the mod, Mars and Zeus, though they’re apparently coming in a patch sometime tonight.
The gameplay is almost identical to OG Auto Chess, save for the courier. Instead of dealing with Dota’s adorable donkey and fighting with its limited sphere of influence, a major disadvantage of the mod, the player is represented by and adorable creature. It’s a goblin by default but there are many skins already in the game. This adorable creature is purely cosmetic and doesn’t get in the way or cause problems the way the courier in the original custom game did.
Of course while the gameplay is essentially the same there are some significant improvements in Drodo’s second outing. The first I already mentioned, removing the courier and it’s at best finicky sphere of influence is a massive boon, as are the visual tweaks that make synergies more distinct. Each synergy has a symbol and color associated with it being displayed on one side of the screen making it much clearer what bonuses you team currently has in place and how close you are to getting other bonuses. This version also has a clear indicator of what your chances of pulling certain rarities are, say at rank 8 you have a 1% chance of pulling a legendary hero. OG Auto chess had this as well but it was in a drop down menu.
The draft is also a bit clearer with heroes obviously displaying their synergy symbols instead of the text Auto Chess used. As a bonus if there are enough copies of a hero on your field, your bench, and in a draft hand to level them up they’ll all emit light and dance to make sure you don’t miss it. Oh and I must say that heroes level up WAY more impressively in this game than the other ones.
I expect that I’ll be returning to items as a more direct focus in the future as they are one of the primary points all these games differ but suffice to say that Auto Chess: Origin is almost exactly the same as OG Auto Chess in this regard.
TLDR; Auto Chess: Origin is the closest game to OG Auto Chess, no surprise there, and the most complete feeling game with full voice acting, albeit in Chinese only, and being the only of these game sporting a progression system. Unfortunately that is where I find fault. Drodo already has a battle pass in place with a free track and paid. The free track ends at lvl. 50, while the paid “ends” at LvL. 80. I say ends but after 80 each level just gives you candy, the ingame currency. My biggest issue is still that it costs 680 donuts. Yes really, to get the paid pass, which to be clear gives you more donuts, candy, and cosmetics, you have to use the in-game donut currency. Taking the paid battle pass and getting up to LvL. 80 does net you exactly 680 donuts so you could theoretically go infinite and getting the paid pass give you access to extra objectives to level faster. Free players can see and even complete the paid objectives but they get no reward until they buy the upgrade. See that’s all not that bad, my issue is that to get 680 donuts you need to either buy 1288 for $20 or 628 for $10 and go buy an extra 60 for $1, so no matter what you will have extra of this premium currency, which unlike say MTGA gems I yell about so often, you can’t even earn through gameplay.
Underlords is the Valve developed Auto Battler. Originally Valve actually flew Drodo Studios in to meet with them about co-developing a stand-alone Auto Chess game. We mostly found out about this because after their meeting both teams announced they would not be working together and were instead making their own independent standalone games. Of course this was a fair bit easier for Valve as all the models, animations, and engines that the original Auto Chess was made in are, of course, already theirs. Amusingly enough in mid-May Valve trademarked Underlords and the running theories were that it was either some kind of game focused on the Dota hero Underlord or a rebrand/2.0 version of Artifact.
So, what does Underlords change? They’ve done away with the courier entirely, for one. Players have no on-screen avatar and it’s not really missed. Other than that? Honestly not a lot. Synergies are now called Alliances. Some racial Alliances have had their names changed to make them more generic so future heroes can take up spots. Orcs are now Brawny, Goblins are now Scrappy, but they mostly have the same effects. Of course that’s not 100% true, because at the time of this writing the Knight synergy has been changed twice. The original knight synergy in Auto Chess was absurdly powerful and Underlords quite heavily nerfed it at launch and have since decided to ease up. Similarly the Dragon alliance has had its ability changed as well.
One thing Valve has said and made quite clear is that as Underlords progresses through seasons they intend to rotate heroes, items, and possibly whole alliances in and out, keeping the format fresh. While that doesn’t help in the short term, I appreciate that they’re looking farther ahead with this game. To help with that Underlords does have 2 heroes new to the Auto Battlers based on Dota. Adding Bloodseeker, a human assassin, and Arc Warden, a primordial shaman, gives each those alliances a little more wiggle room, as previously they all only had just enough members to max out each synergy.
Underlords also has clear Alliance markers instead of the original mod’s text labels and even displays how potential draft picks will affect the Alliances they belong to. Of all the current Auto Battlers, Underlords makes it clearest who does what and what kind of effect the heroes sitting on your bench could have on your standing Alliances.
So where does Underlords really deviate? One deviation comes from gold generation. This is a bit minor but interest in Underlords is calculated once a round of combat starts. This means the game sort of encourages you to play conservatively and really think out decisions before combat.But during combat, while you have really nothing to do and any interest you might get is locked in, you’re encouraged to spend more, either on levels or rerolling the shop drafts.
However the major difference between Underlords and its contemporaries is in items. Underlords removes the most painful RNG from the item game. At the end of every neutral creep wave each player is offered a choice of three items. They range from early game minor buffs to global effects tied to alliances to powerful effects that make it so the player sees new drafts or items as though they’re a level higher than they currently are. And as a massive bonus, items can be freely moved between heroes! If you try to move an item in combat it will not take effect until after the combat ends. Sadly, to balance the new item system, items cannot be combined, and each hero only has one item slot available.
The final thing I have to say about Underlords is that it is the most community friendly game I’ve seen in a long time. It’s available on every PC OS, Windows, MacOs, and Linux, and on IOs and Android, with full cross play, AND the party code system means that people who aren’t on your friend list can still join the party. This is amazing for streamers who don’t necessarily want to be friends on Steam with their entire communities.
TLDR; Underlords is the most polished of the Auto Battler games out now. It’s definitely not done and it needs some kind of progression system to really hold interest, but that’s tertiary and this is an early access game. The important thing is that the primary gameplay loop is fun and keeps players invested. Though the games can go quite long because players are attacked by clones instead of direct battles you can end up in strange situations where only two players remain but they continue to win on their own boards.
Team Fight Tactics
Teamfight Tactics is the Auto Battler released by Riot using League of Legends heroes. They have a chance to do something new and exciting, and they kind of did? Before I say anything else, Teamfight Tactics has great bones. It has the makings of an amazing game but it feels like it was rushed and maybe it’s working too hard to set itself apart to start with.
TFT is the most visually different of all the games, which is definitely a mark in its favor. Dota character design has always leaned into a sort of stereotypically “cool and edgy” look. Auto Chess: Origin is adorable but is pretty clearly trying to make existing Dota characters look cute and legally distinct. League on the other….look League has characters that are over-designed out the wazoo, but that gives them a unifying theme, that and many are just happy bright designs, or at least bright and garish (aka: easy to tell apart).
Which makes it so weird to say it feels like they picked a bunch of not visually distinctive characters. In several of the games I played, I found myself trying to draft gunslingers mostly because it was easy for me to see who they were because they all had guns. A lot of the other synergies were not as clear. This, I will be the first to admit, is because I haven’t really played League since 2012 save for brief lapses in 2014 and 2017, so many characters are new to me. But that doesn’t change the fact that a few too many are just sort of grungery grey brown folks.
Now all of that wouldn’t be much of an issue if synergies were clearly displayed on heroes or you could see how your benched heroes could affect your field. Heck, it’s even kind of a pain to see what abilities your heroes on the field have. In most Auto Battlers you left click a hero to see stats, hold left click to move them. In TFT you right click them (and only them or else your courier will come out to play) and even then you don’t get info on damage per hit, or attack speed or anything. You get health, starting mana, synergies, what their Ult is and what items they have. This is all helpful info, but it’s not everything you need. But, as I will keep saying throughout this part, this is very early access so it should improve over time.
Up next on the chopping block of TFT, the Carousel. In theory, I LOVE the idea of the Carousel, not so much as a comeback mechanic, but just as a way to get a hero or two and some items for free. BUT, and it’s a big but, not being able to see what items do or what synergies the heroes belong to hurts a lot, especially for players who don’t know LoL characters (cough, me).
Look, essentially all my complaints about TFT boil down to UI issues, having your opponent’s name and level in a bubble about their head is great in theory, but because TFT kept the whole “your courier grows bigger if you win streak” from the original Auto Chess mod it’s possible for someone to show up for a fight and have their health and level offscreen because they’re massive and the camera is fixed. I love that all your info, your gold, your current draft options, you level up and everything is all nicely condensed at the bottom of the screen, I love that you have a round tracker up top that shows past and upcoming rounds, that’s amazing! But so much of this gets in the way once you start spectating. For some reason in TFT when you look at someone else’s board it acts like they’re on the enemy side, this means again their health and level might be hidden and half of their bench is behind the round display that I love so much.
So I hear you asking angrily, “Cap, is there anything you actually like about TFT?” Yes! I love that there is less focus on heroes with AOE CC abilities, I love that we get to see whole new synergies and entirely new heroes to auto battlers. Like seriously, I fucking LOVE the shapeshifters in TFT with Glacials as a close second. I love their items, in theory, because the idea of literally any two basic items being able to combine is awesome!
But in practice it’s a bit different. Like OG Auto Chess, items can’t be removed once they’re given to a hero, and for some reason items can’t be combined when not attached to a hero. So to find out what combination you’ll get you either need to know already, or commit to placing one item on a hero then mouse over them while holding the other item you wish to combine and hope it works with that hero. Or like me you could, for like the 8th time, accidentally have a situation where you’d be giving a melee attacker a spell damage item, which again is an issue because the game doesn’t tell you who is a spell damage attacker!
My final complaint with TFT is that it kind of hurts that they’re already selling cosmetics (replacements for the river sprite default courier) when they’ve already announced how they plan to make them available through gameplay. They just haven't put any of those plans in the actual game yet. It makes the game feel that much more rushed, though the occasional bugs where heroes just stop moving and attacking certainly don’t help the unfinished feel either.
TLDR; TFT has the bones of a great game, but is currently burdened by a somehow cluttered yet uninformative UI. Having new heroes and entirely new synergies than the existing games is great! Item combination is awesome, but needs a better way of seeing item combos other than committing one item to a hero. The carousel is a great concept, but the players need to be able to see info on the heroes and the items they hold. And maybe it should be clearer when a player’s turn will be. For now it seems to go in waves of two starting with the player in last place. Players need more info on all of their heroes period.
All the Auto Battlers out on the market, save the myriad mobile clones, are solid games. Auto Chess: Origin feels like the most complete game but has a rather parasitic business model as well as ties to some unwholesome Chinese mega corps, but most games do too if you dig deep enough; and it is currently out for only IOs and Android and will be coming to the Epic Store, so likely no native MacOS or Linux support.
Underlords is the most polished of the three and the most rapidly updated, though that’s a double-edged sword at best. Underlords and TFT both need their tertiary gameplay loops (usually a Skinner box type of reward that gets you to play games repeatedly). While TFT has laid theirs out ahead of time, Underlords hasn’t, so that might be a stumbling block for them in the future. Also this is a Valve game, so it’s probably safe to assume it will end up making boat loads of money off the back of people making skins or cosmetics and selling them while Valve takes a heavy cut. On the good side it’s available on every PC OS and most phones (I’m led to believe some people still claim to use Windows phones) and it’s remarkably user friendly now that the “Australia mode” problem on Android is fixed.
Teamfight Tactics, and I know I keep saying this, has the bones of a great game. I think given time and love I will like this more than Underlords, assuming Underlords were to never change. It hurts that I can’t play with as many folks as I’d like, but the game still shows great promise and while it feels very rough now, I can definitely see the diamond in there. I do have one weird suggestion, maybe if the hero store also used 3d models like the other auto battlers. LoL models, especially when seen close up like that, are pretty distinctive but their 2d portraits are a bit busy and action heavy. Also many of the portraits are kind of similar, like how the yordles just look like other yordles. And of course it is still connected to League and that’s going to push people off. It is still Riot, and yeah, kinda fuck Riot management. And it is 100% not new-user friendly. On the plus side though, it is beautiful, like seriously just pretty. It has entirely new heroes and synergies instead of the same ones retextured across three games. It has a low learning curve if you’re at all familiar with Auto Battlers and already know League characters. And finally, with TFT you get to stick it to Valve without supporting Epic? Yeah I know that one is stupid but I’m certain it’s a selling point to freaking someone.