MIBR were dressed in their electric yellow synthetic jerseys every time I saw them at IEM Katowice. Even if this was your first Counter-Strike tournament, even if you were walking into an esports hall completely blind to the culture, the national allegiance of the six young men on stage was aggressively clear. On the Brazilian flag, that yellow symbolizes wealth; it lit up the uniforms the Brazilian soccer team wore during each of their five World Cup titles, too. Last year, when this team switched parent companies from SK Gaming to Immortals, they resurrected a familiar name: MIBR, Made in Brazil, sheathed in the same colors that have delivered glory to their countrymen so many times before.

According to Marcelo "Coldzera" David, a player who at times has been the best Counter-Strike marksman in the world, that christening was a no brainer.

"We create a brand for Brazil. That's why we brought back MIBR, we want to represent our country, we want to make our country proud of us," he says. "To create a legacy. A Brazilian legacy."

The first teams fielded under the MIBR name date all the way back to 2003, during the Counter Strike 1.6 days, a time when esports was still largely underground and punk rock. As a nation, Brazil was in the halcyon stages of an economic boom, and its citizenry was falling in love with first-person shooters. "Since I was young Counter-Strike was always in the LAN houses. It was a game that didn't require an extremely good computer at the time," says Augusto César, a fan swaddled in a Brazilian flag, in the IEM Katowice food court. "For us it was a very simple game to play."

Two decades later, the country fields one of the best CS:GO squads in the world. The modern incarnation of MIBR captured a major title last year at the ZOTAC Cup, and five premieres and an additional major in 2017, under the SK banner. They represent a glacial power shift in the fabric of esports. Scroll through the attending teams at Katowice, and you'll see that between the Americans, the Ukrainians, and the French, MIBR is the only organization representing South America. 

You could feel it in the air. MIBR's success is an exception to the global rule. The fans know it, the scene knows it, and it summons up a one-of-a-kind passion. Katowice is a dinky Polish mining town at the southern end of the country, and the Spodek Arena is a communist-era UFO-like relic built in 1971 that serves hockey games and B-list festivals. Still, miraculously, the Brazilians showed up in droves for the boys.

They dotted the seats and the outer hallways, and most, like César, brought with them their national colors. That makes sense; if MIBR is going to wear a patriotic yellow, then it behooves their supporters to follow suit. But their loyalty took on a different texture than any of the other esports organizations at IEM. Sure, the Danes root for the Danish wunderkinds in Astralis, and it was genuinely heartwarming to see the underdog Finns in ENCE make a deliriously joyful run to the finals, but MIBR are the only ones to literally emblazon what they're fighting for in their name. Brazilians hear that call no matter where they are.

"I didn't play Counter-Strike for a while, I wasn't interested in it until Brazil started to get big. Like, 'Oh, Brazilians are really nice at this game, I need to play it too,'" continues César. "It's what got me into esports. Brazilian teams succeeding. I think most Brazilians are like that."

One of the things I love most about sports—the local, tribal pride, and its corresponding politics—has always been de-emphasized in competitive gaming. The Overwatch League's home cities may be slowly changing that. Generally, teams take their identity from an overarching brand or sponsor—we pull for our favorites the way 14-year olds pull for LeBron, regardless of what team he's currently playing for. But MIBR is a different beast.

The kinship is closer because of the sheer rarity of other Brazilian role-models in pop culture. Since 2016 the majority of the headlines coming out of the country have focused on either a beleaguered Olympics bid, a troubling presidential election, or mounting corruption scandals, but when Neymar Jr. joins Coldzera on Dust II, all of that fades away.

"Brazilians like to cheer for Brazilians that could win. It happens in soccer and basketball, when we have a chance to win a trophy, Brazilian blood heats up. One thing we miss in Brazil is someone who can represent us and show our good side," explains Rodrigo Guerra, a journalist who covers MIBR for ESPN Brazil. "A few decades back we had a famous driver, Ayrton Senna, and everyone woke up at four in the morning to cheer him. Because he's carrying our flag, he's showing how we can be one of the best. We're always named as a third-world country, that we don't have any potential, so anyone who can change that misconception, in any sport, it's very important to us."

I was left with one lingering question at Katowice: Why Counter-Strike? What is it about this game in particular that's found such a home in Brazil? Sure, you can catch a few stray Brazilian squads in DOTA, and fighting games represent one of the true international scenes on the planet, but when you look at League, or Overwatch, or StarCraft, you rarely find much star power outside of the European and East Asian strongholds. I'm not the only one who's noticed that, either.

"Since the beginning of esports in Brazil, everyone was playing Counter-Strike," says Guerra. "StarCraft, or League of Legends, it's not natural to us."

Everyone I posed that question to returned to the same core point. Counter-Strike was, and is, a fixture of the LAN cafe scene in Brazil, and its resonance in the games culture grew out from there. It just stuck. It's an answer that's about as arcane as anything else in the esports industry. I mean, why do South Koreans excel at StarCraft? A variety of historical accidents, that eventually coalesced into a national heritage and sense of ownership. This isn't an exact science. Coldzera, at least, is able to take a unique perspective, since he admits to me that he does keep a League of Legends habit during off-hours.

"League and Dota are amazing games, but the difference between our sport and them is that they're more mechanical," he says. "In Counter-Strike, you have to factor in the randomness. You can kill someone blind, you can spray and kill one more. That's why the game is so great. Things can be going really good, and it can be destroyed in one round. It gives you adrenaline. That's why I think Counter-Strike is the best game ever." 

Coldzera tells me he hears a lot of the same stories from people like César; Brazilians who loved video games, and loved their country, and were brought into the fold by the team's collective excellence. Coldzera himself has a great respect for the first incarnation of MIBR. Today, casters have even coined a term for the team's deliberate, slow-paced gameplay. "The Brazilian Style." You know you've made it when you're part of an institution.

MIBR lost in the semi-finals to Astralis, the eventual champions. When you look over the roster, you begin to see the glimmer of change on the horizon. Coldzera is 24, and both team captain Gabriel "FalleN" Toledo and Fernando "fer" Alvarenga will be 28 at the end of the year. They are still young by every conceivable measure except for esports, and while Counter-Strike is generally more friendly to the stringent burnout problems compared to other games, eventually a new generation of Brazilians will need to take up the banner.

It's a hope that MIBR welcomes with open arms. None of these players would ever give up their spot without a fight, but throughout the weekend, I noticed that they displayed a remarkable solidarity with FURIA, another Brazilian Counter-Strike team who made it into the qualifying Challengers bracket. (#DiaDeFuria, they tweeted, just before Valentine's Day, during one of their first matches of the tournament.)

Furia showed well, but didn't finish with enough wins to move on to the next level where they may have had a date with MIBR.  "They just need more experience. They have a long road, but they're on the right road," says Coldzera. "It's nice to see a new face for Brazil. … It's crazy to see how they [get better] every tournament."

Given that MIBR is the effective stand-in for Brazil's national team, I asked Coldzera what it'd be like to someday go against a Brazilian team in a major tournament—to have the throne challenged by someone in their backyard. As much as he cares for the future of his country's Counter-Strike scene, would it feel any different when he was staring that future in its eyes?

"That's gonna be nice," he laughs. "Brazil wins, no matter who wins."

Valve has released a companion app for subscribers to Dota 2's premium monthly subscription service, Dota Plus. The app lets you keep track of match and tournament results, as well as keeping you informed about your favourite players and teams. And then it lets you place wagers on them. 

The Dota Pro Circuit app, available for iOS and Android, will let you make 'predictions' on upcoming matches in the pro circuit, wagering your own shards—the Dota Plus exclusive in-game currency—in the hopes of adding to your pile. All the information you get about teams and players, then, can be used to make your predictions. 

No cash changes hands, unlike in the third party gambling that surrounds the game, though technically those shards are linked to financial investment because the currency (and thus what you can spend it on) is only available if you pay for the Dota Plus subscription. That's $4 a month. That technicality isn't enough to make it gambling, however; at least not to rating organisations like PEGI, which has given it a 3 rating, essentially meaning it's appropriate for all ages. 

The Dota Pro Circuit app feels like the latest in a years-long stream of game systems which have been pushing the debate around gaming systems and how close they come to gambling to the fore, though. 

Probably the most well-known facet of the debate involves loot box systems which encourage players to open in-game Macguffins to obtain digital presents of variable rarity or desirability. Often these are boxes or chests you can earn over time in the game or choose to circumvent that grind using real money. 

Another facet is the third-party gambling and trading scene around games like Valve's CS: GO where players circumvent Valve's systems in order to use in-game cosmetics as a gambling currency. The real money comes in when players make payments via third party sites and then use the steam trade function to hand over their digital goodies. 

Despite the PEGI classification (and I should stress that PEGI do not set the legal definition of gambling, they're merely an enforcer), I'd argue the mechanics of the Dota Pro Circuit app's predictions are a lot closer to real-world gambling than, say, Overwatch's loot boxes. 

You can study the teams, check the odds, place a bet with a cash stand-in, get that thrill when you get a shard windfall—the most significant difference is that it's all taking place within Dota 2. You can't head to the bank with your pile of shards and make a nest egg. It's a new grey area in a debate filled with grey areas.  

The app's available now, and you can check out the full feature list here.

Update: It appears that, yet again, our hopes have been raised, only to be smashed down upon the cold, hard, rocks of reality. Shortly after this post went up, McVicker deleted the video containing the "five years or so" tease, saying on Twitter that two people have confirmed to him that the email was actually a fake.   

It's unfortunate, but it also aptly illustrates the thirst that so many of us have for a new Half-Life, or even just an indication from someone in the know that it's actually (maybe) going to happen. I still haven't received a response from Newell denying—or, because hope dies hard, confirming—the original report, but I'll let you know if and when I do. In the meantime, it looks like we'll be adding this one to the collection.

Original story: 

Will we ever actually see Half-Life 3? Or even Half-Life 2: Episode 3? Probably not. But maybe! Who knows? Valve boss Gabe Newell does, possibly, and he's even got a time-frame in mind. Maybe. 

The latest thin Half-Life 3 pseudo-tease comes from Newell himself by way of Tyler McVicker, who's behind Valve News Network, an unofficial but well-informed source of goings-on at Valve. In a new video, McVicker shares a number of user-submitted email replies from Newell, some of them silly meme fodder and others more serious. 

The big bit comes at the very end of the video, around the 6:45 mark, when a fan writes, "Was just wondering, will I be able to play [a] new chapter of Half-Life before my life ends? I'm 32 now."  

To that, Newell replies, "Just don't die in the next 5 years or so ;)" 

That is incredibly imprecise, but it fits with what we've come to expect from talk of the next chapter in the Half-Life series: Eternally hopeful, but couched in the sort of cynical doubt that comes from well over a decade of elevated expectations and dashed dreams. Is it possible that a new Half-Life, be it Episode 3 or HL3 or whatever, arrives at some point in the next half-decade? Sure. Is it likely? That's a different matter altogether, especially when the time frame comes with the "or so" qualifier attached, which could mean just about anything. And that winky face? I have no idea.

There's also the question of whether or not Newell actually sent this email: Valve News Network is a reliable source, but just in case, I've emailed Newell to make sure he definitely sent this email. I'll let you know if he replies. 

Thanks, PCGamesN.

Portal and Left 4 Dead writer Chet Faliszek and Riot Games senior technical designer Kimberly Voll have launched a new studio called Stray Bombay Company that will focus on the development of AI-driven "collaborative gaming experiences." 

"As Kim and I talked over the years about the kind of games we want to make, we realized we want to create games that give players a place to breathe and live in the moment," Faliszek wrote at straybombay.com. "Games that tell stories knowing you are going to come back again and again, that change each time you play them without feeling completely random, and that help you feel like a real team that supports each other... not a bunch of folks in each other’s way. And where AI drives not just the enemies but helps drive the entire experience." 

Faliszek is likely the better known of the two among gamers, but Voll has an impressive list of credits too. She's worked on numerous indie games over the years including the 2016 VR release Fantastic Contraption, a must-have for HTC Vive owners, and last year she gave a GDC presentation about how Riot was able to revamp League of Legend's honor system. 

"We think now is the time to change the culture of game development. Make everyone equals, not just in their impact on the project but in how we divide the loot of our success. Relax strict PTO policies because we trust each other to take the time you need. We want to build games that reflect our culture," Voll said. 

"We are supported by like-minded, patient investors with Kevin at Upfront Ventures as our lead, who has known us for years, working with Chet on a gaming startup board, as well as Riot Games. They love and understand games but more importantly, have time and again backed founder-led and employee-owned tech startups from the beginning towards long-term, independent success." 

With the studio announced, the plan now is to "go dark for a little bit as we start laying the foundation of our new world." It sounds like there's already been some decent amount of progress in that direction, though, as Faliszek told PC Games Insider that he had a prototype running in Unity and Unreal before partnering with Voll to launch the studio.   

"We know the direction we're going. As people join the team, that'll help find the game more clearly. We're very iterative, everyone is a designer, everyone participates in the process," he said. "Everyone joining [the studio] will have a big impact on the project. Obviously, we have a plan, there's a framework that we can hang it all off, but everyone will be able to express themselves and have an impact." 

Faliszek also confirmed that he is no longer with Bossa Studios, the Worlds Adrift developer he joined up with in 2017. In fact, it sounds as though that might be a big part of what led to the launch of Stray Bombay. 

"I just don't think it was quite the right match. Bossa is great; I love them. It was the most cordial breakup of making sure that everybody was okay and helping each other the best we could ... I take a lot of notes, I track progress and see where things are so I can have a good perspective on it and I brought some stuff up and we talked it through and it just wasn't good for either side. We decided to break-up."   

Faliszek said that it was up to Bossa to comment on the state of that project following his departure, so I've reached out to ask and will update if I receive a reply. To keep up with happenings on the new project, you can add your name to the mailing list at straybombay.com. 

This article originally ran in PC Gamer 328 back in February. The render above is by our art editor John Strike. Subscribe here and get great features like this sent to your door every month.

It all started with Black Mesa. Firstly, because the stars aligned in Christmas 1998 such that my first taste of PC gaming happened to be Half-Life, the best shooter ever made. Santa Claus delivered a personal computer to our home that year—a Packard Bell Platinum 350, since you ask. A 350MHz Pentium II lay within it. A 3DFX Voodoo 2 with 8MB onboard memory. 16MB of RAM and 6GB of hard drive space. These were formidable gaming specs, and when I was given the luxury of choosing a new PC game to accompany it under the tree I relied on the wisdom of PC Gamer, who sure were keen on this Half-Life game. That Christmas was magical. But that’s not the point. 

I mean to say, it all started with Black Mesa, the Source Engine reworking of Half-Life by Crowbar Collective. When it appeared on my radar back in 2013 I thought it looked like the perfect way to experience the game I’d confidently been calling the best shooter ever made—having played it just once, aged 12—once again. After 15 years of abstinence I’d once more allow myself to take in the giddy delights of creeping past the tentacle beast and watching Barneys plunge to their doom in broken elevators through the lens of this Source Engine remake. 

I lasted about two minutes. That was all the time it took to realise that every slight deviation, every instance of minor creative licensing, was only going to wind me up. The Barneys all had different lines! The posters were slightly different! Some of the rooms were bigger/smaller than I remembered! This wouldn’t do. 

No, this wouldn’t do at all. I made a very serious promise to myself that day, having closed down the perfectly good Black Mesa mod. The only way I’d play my darling Half-Life ever again was in situ: the original game disc I’d kept all these years, running on a Packard Bell Platinum 350. So began a painstaking and indefensibly self-indulgent quest to source nearly worthless PC parts.

The keyboard and mouse were surprisingly easy to get hold of. Having set up eBay alerts for every bit of Packard Bell minutiae I required, I was directed to the very same ’board, complete with redundant multimedia controls, going for a princely £10. It arrived shortly afterwards, smelling faintly of someone else’s house and, well, presumably working. I didn’t have anything to plug its PS/2 connector into to check. I opted for a Microsoft Intellimouse to pair it with because—and I’m ashamed to write this—I can’t remember much about the original mouse that came with my first PC and I’m pretty certain we plumped for the Intellimouse quite quickly anyway. They’re ten-a-penny on eBay too. Easy, this retro PC-sourcing lark. 

Retro fit

The real difficulty began, funnily enough, when it came to finding a specific model of PC released 20 years ago in good working order. Retro gaming PCs are all the rage at the moment, and there’s a growing cottage industry of PC builders who source old parts and practise the dark art of ‘refurbishment’ on them (in reality a can of compressed air and some homemade bleach solution to remove the yellowing on beige plastic). But what if you’re not just looking for a retro gaming PC, but the retro gaming PC? After a year of eBay alerts and fortnightly searches, I hadn’t come close. I was at such a low ebb that I considered hitting the ‘Buy it Now’ button on a Dell.

I also thought about building the machine by sourcing the individual parts, and I must now say in the strongest terms possible: don’t do this. You’ve forgotten everything about hardware standards and compatibility from 20 years ago. You have no idea what chipset that motherboard you’re looking at is, and there isn’t a damn thing on the internet about it to inform you. No one will help you if that Voodoo 3 doesn’t fit, and good luck getting all the right cables to connect your miraculously compatible components which you’ve implausibly found working drivers for. Honestly, forget it. Buy a prebuilt PC which the seller confirms is in full working order. If they include the original recovery disc, that’s a massive bonus. You’ll need to buy your old operating system of choice otherwise, and although Windows 98 isn’t quite as expensive now (£15-£20 from most sellers), it’s an added cost you might initially overlook. If you want, you can even refurbish an old machine yourself by buying a £5 can of air and following one of the many questionable recipes for ‘Retrobright’ solution to bleach parts back to factory fresh—just know that PC Gamer accepts no responsibility for you ruining your floors, bathtub, hands and PC parts. 

It came as quite a surprise when my exact make and model of PC materialised on eBay after a full year without leads. I stared at each shaky smartphone photograph on the listing with an almost pornographic fascination, barely conceiving the needle I’d found in eBay’s discarded goods haystack. The seller had listed it with a guide price of £400 to encourage private offers, and honestly I’d have paid it if it came to it. In the end, though, I sent an offer of £100 and spent the day worrying that I’d lowballed to such an insulting degree that my bridge with this seller was forever burned. He accepted it instantly, because you would, wouldn’t you, if some weirdo came out of the woodwork desperate for your unwanted two-decades-old computer. Thanks again, sync_it, and sorry about deleting all your old Champ Man 3 saves. 

The fates had smiled on me. I’d secured some very specific pieces, and I hadn’t even had to risk the one website that claimed to still be selling my original PC new, 20 years later. Still, two pieces elude me: the Packard Bell Milano 17-inch CRT monitor, and the recovery disc. I’ll keep searching, of course, but I was especially disappointed not to fully immerse myself into 1998-o-vision with a CRT’s characteristic display. Good CRTs are hard to find now—most have either been chucked, broken, or taken to recycling centres to fester away. Of those that appear on eBay, most predate my target Windows 98 era—there’s high demand for Win95 screens and earlier, it seems. They’re also heavy as all heck, which means delivery is a real issue—there are specialist couriers who deal with fragile items and know how to handle a CRT, however. Perhaps there’s someone out there who’s cared for a Milano monitor all these years, someone now ready to part with it. Until I find them, I must subject my retro rig to the indignity of outputting on a 32-inch IPS.

Never mind. The beating heart of the PC was just as it had been. Likewise my peripherals. What a perfect way to remind oneself what PC gaming was really, truly like 20 years ago. That’s no small point—that era’s been fetishised in recent years, evidenced by Kickstarter-funded odes to the Infinity Engine RPGs, and reboots of everything from Thief to Outcast. Not to mention Black Mesa, of course. What’s clear when you press the power button on an old tower PC, hear the Windows 98 welcome chimes, and load a game’s CD-ROM into the tray, is that we’ve forgotten much of the era’s reality.

For example: first-person shooter control schemes were the wild west in 1998. By default, Half-Life’s controls are bound to the arrow keys, of which left and right turn, rather than strafe. At least mouselook is enabled by default. Quake II, released just a year prior, maps the mouse to moving forwards, while A and Z control your vertical view. Barbaric. 

Revisionist History 

On the technical side, we’ve forgotten much about what games looked like when they released—for most people, running on a software renderer in 640x480 and still not hitting anything like 60fps. The Half-Life you see in YouTube speedruns and let’s plays, running at modern day resolutions, bears little resemblance to the one I lost myself in the Christmas it came out. That game is grainier, darker, and somehow more atmospheric for it. Although hundreds of shooters have since aped Half-Life’s setpieces, NPCs and storytelling techniques, Valve’s vision stands as tall and impressive on this retro PC as it did on release. Half-Life was, and is, a place you go, rather than a game you play.

Perhaps the most profound realisation that comes from building an old PC and booting up a treasured memory is that I’d advocate every single PC gamer do the same thing. The nostalgia hit is absolutely worth all that e-trawling, but more than that, sitting at a PC without an internet connection (for goodness’ sake don’t try to go online with Windows 98) reminds you how easily distracting your modern gaming habitat can be. There’s nothing to alt-tab out to and no Steam list of zero hours played shame. That feels like an important reminder. 

We first got a heads up that Mars, god of war, would be joining Dota 2 back in August of last year. The day has come at last, and Big Red is here and playable right now.

"The warrior deity Mars thrives in the heart of strife, guarded by the bulk of a deadly shield as he skewers enemies with his legendary spear," reads his page on the Dota 2 site. "He revels in facing opponents in an arena ringed with loyal spearmen—who guarantee that no one escapes and that whatever odds he's facing, the god of war can dictate the terms of battle knowing the crowd is forever on his side."

Here's a look at his abilities and ultimate:

Spear of MarsMars can hurl his spear, impaling the first enemy it strikes and even nailing them to any trees or walls they happen to be standing in front of.

God's RebukeThis is a nifty and wide-ranging shield bash, which knocks back and damages foes.

BulwarkMars can block some damage from physical attacks when struck from the front or the sides.

Arena of BloodHis ultimate basically builds a scary stoneghenge around himself, which will block attacks and movement from enemies outside it. Enemies trapped inside will get jabbed by spears from Mars' undead warriors, who spawn around the edge of the arena.

We're digging into the PC Gamer magazine archives to publish pieces from years gone by. This article was originally published in PC Gamer UK 2005, and you can find more classic pieces here. For more quality articles about all things PC gaming, you can subscribe now in the UK and the US. 

Great games don’t just appear. They’re moulded and shaped over years. Millions of man-hours are poured into their development. To create even a mediocre game is a massive technological and artistic challenge. To create one of the finest games of all time, one that demonstrates everything this entertainment medium can achieve, to rise head and shoulders above your peers... that takes genius.

Right now, three of those geniuses are sitting across the vast table that dominates Valve’s Seattle boardroom. Marc Laidlaw, Valve’s in-house writer. Bill Van Buren, a producer and designer. Bill Fletcher, senior animator, formerly of Disney. These three quiet, understated men are nothing less than magicians, responsible for bringing to life virtual characters, and at the same time satisfying players who just want to break things. The walls are covered in testaments to their skills: magazine covers, box-art, countless awards and trophies, and trinkets sent in by their legions of fans. The whiteboard behind us is bursting with arcane scribbles: profiles, plans for the future. They can wait.

We have other questions to deal with. What makes Alyx tick? Why does Dr Kleiner keep a pet headcrab? How did Valve bring these virtual people to life? These are all important questions. First though, we need to clear up something about a scientist’s beard.

PCG: I’ve just noticed, Marc, that your beard is an exact replica of Gordon’s.

Marc Laidlaw: Oh, er. It came from...

Bill Van Buren: Actually, the beard is from another member of staff. Gordon’s actually a morph of four or five Valve employees.

ML: Yeah. He’s the average dad developer.

He’s certainly an unconventional game hero. The biceps aren’t there, he doesn’t have rippling pecs, he even wears glasses.

ML: Imagine what he looks like when the suit comes off. He’s probably just about 90 pounds. It’s all exoskeleton and suit support.

Yet when a player enters Half-Life, you’re asking them to take on that role. I wondered, Marc, if in your writing, Gordon has a voice or some kind of internal dialogue.

ML: Not really. Gordon’s just the narrative conduit. I’m interested in everything else. What does all that say to the player? If you’re just a pair of eyes looking into a scene, how do we let the rest of the world tell you they know you’re there, or that they’ve been thinking about you, while you’ve been gone? There have to be hooks—they have to glance at you, occasionally. My favourite is the final scene, where Mossman is talking to Breen, and Breen is looking at you, he rolls his eyes. That’s clearly for your benefit.

I’m guessing that beneath all the storytelling you do, a lot of your work is in choreography... I don’t think anyone really talks about what’s a great ‘choreographed moment’ in a game. When you’ve got an idea for a great scripted moment, how do you begin working it into the game. Maybe you could describe how the teleporter scene in Half-Life 2 came to be?

ML: Well, we started talking about that scene many, many years ago. We always knew we wanted to do a parallel to the teleport disaster in the first game. Originally it was going to be Doctor Kleiner and Barney—you weren’t going to meet Alyx until much later on. Through our process of figuring out how and where we were going to introduce characters to the game, we realised we should bring her forward, into that scene, so you would meet her in the context of characters who you were familiar with. We could make it feel like “Oh, I’m back with old friends, she’s one of their friends, so she must be good, she’s one of my friends, too.” That just created an instant bond with the player.

That scene is all about setting up expectations. It’s very different from the first game, where you didn’t know what was going on. This time every player expected to jump into a teleporter. We set up the idea that something good could happen this time. We’d let Alyx get ahead of you, you see everything that you’re about to do. From the first-person point of view it’s hard to picture yourself in context. When you let someone else do it you have a good mental image of you getting in the teleporter, you know what to expect. Again, playing with people’s expectations.

As for drama, we knew we had to do some suit training, we had to re-introduce the characters somehow. But also, one of the important things was “we want to build characters”, and “we wanted to build other details in to the game”. We wanted to create something you hadn’t seen in a game: a family. We wanted to do a stepmother-daughter uneasiness. Then there’s the non-essential stuff—like the thing with Kleiner, Lamar and Barney. The more of these details we put in, the more real characters feel. They have a life that extends outside the boundaries of the scene, and the game. You get this feeling that “they know each other, they have history”.

BVB: This is all about emphasising the importance of other characters. We wanted to make the point that you’re not going to save the world alone. You’re a catalyst, you’re important, but other characters are setting up the scene. The goal isn’t just to liberate the world, or even to liberate you, originally it’s just to get Eli, the father of this girl who puts trust in you. You can’t do that alone. You need to work with the other people in this world. If you want to do something big, you need to have other people working with you.

ML: In Half-Life, things happen because you show up. We don’t have you turn up, as the thing starts. We have you show up, and the thing begins as the result of you showing up. You’re the first domino that knocks all the other dominos over. What we do is let you arrive, walk around, see a bunch of dominos, you push on them, and they fall. That’s what we like to do.

Marc, you were a published author before you came to Valve. Did you have to adapt your writing style to work in games?

ML: From a writing perspective, everything you want to reveal about characters should come through dialogue. It’s not like there was a bunch of backstory about the characters we wanted to deliver. We knew the characters, we knew a bit about their history, we wanted to make sure that when they were talking, they were referring to this whole backlog of things that have gone before, that they don’t have to explain to one another, just like if you walk into a room where two people who’ve known each other for a long time are talking. They wouldn’t spend any time at all talking about the stuff they already know. But you’d pick up a lot about their relationship just by watching them interact. That was the dramatic thing we wanted to aim for. But that’s not any different to how the scene would be laid out in a movie or a book.

But surely the demands are different. You’re going to be writing to fit the game, aren’t you?

ML: Yeah. Things change in the course of building the game. We’d cut a level, and that would alter a relationship.

Is it always that direction? Are you always writing the story to service the level design? Are you just fitting the story together?

ML: Yeah, that’s part of my job. We’re all interdependent. They might make a decision from a level design point of view, and it’s more like a fun challenge. “How can we do without that plot point? What does that afford us? Now that we don’t have to do this piece, what can we do instead?” Usually I end up with two pieces that are not connected, but right up next to each other and I have to think of a way of linking them. But you see this really, really obvious connection, that was there the whole time. Then you organise around that.

That could have an odd effect. We eliminated the guy who was supposed to be Alyx’s father, and suddenly that changed the whole balance of the cast. Originally, Alyx was a go-between for rebel forces, she was can-do, rugged and tough. Her father was an army general who’d been caught, and planted into a Combine installation. In the end we cut the Combine installation, losing him. We felt this was one character too many; we had a bunch of older men characters who started to feel kind of similar. At some point, I went “I don’t think we need this Captain Vance, it’ll be one less character to worry about.” We weren’t giving enough development to the characters we had.

As we did that, for a little while Alyx was on her own, just floating over here. Meanwhile, we had Eli, here on the other side of the table, unconnected. We just pushed them together, saying “Ooh. We think they’re related.”

Did that cause any conflicts?

ML: It wasn’t like there was big resistance to it. Everybody just went... oh.. yeah, OK, this really helps us. We rewrote her story, basically just filling in gaps we weren’t happy with anyway. If we had her be the daughter of any other character, that would have introduced a lot more complications, whereas, when you meet her as Eli’s daughter, then you instantly understand her connection with Black Mesa, and to you as the player. All those connections are shorthand—they help us in not having to explain a lot of stuff, in the background. That means we can spend the time in our scenes just doing the really direct, emotional thing.

So now you have the script. What happens next?

ML: The script represents things we want to accomplish with the story, and the stuff that the animators, that Bill, wants to get across and spend time animating. This is really abstract, but while this level is being built, we have to roll it out. We have this problem of making a scene come about in a space in the game. That’s pure choreography.

BVB: We actually have a choreography team, that’s what we call them internally, who deal with a character’s audio, animation, acting and scene design. A feature of the way we work is that the scene is constantly coming into focus. We have such an adaptive process, some of it is even collage-like. The more information you get into it, whether it’s the script, the acting sessions, the effects or the animation, the better it is.

We’ll have very good character profiles that inform the script, then some people will take that and bang on it. Marc takes their suggestions and rolls them in, then we’ll go back and do a recording session. The actors then bring in a whole new different level of intelligence and interpretation to the stuff that has been written. With some direction, knowing how it’s supposed to be played in the game, we’ll come back and put it in the level. Then, even more changes happen... “This is too long, we can’t fit it here. We can’t afford this line... but, this line the person did was a much better take, we didn’t realise that in the studio.

Sometimes even lines come up that were accidents, they weren’t even intended to be used... They’re so well said that we grab them and put them into the scenes. Bill, driving the animation/visualisation, knows what he wants to do with them in terms of body language... so that takes it again to another level.

How do your voice actors react to giving ‘game’ performances. Are you after something different from what they’re used to?

ML: I think it’s probably like radio. It all has to be there in the voice. We do a bunch of alternate takes – “give me this performance as if you’re right next to me. Now do it like you’re 20 feet across the street.” One of the things they provide is a range of experiences that we can get back here to see what works. We can do mix and match... It’s not like we go into the studio and get the fullblown radio-play version of the thing. We have to bring it back here and cook it for a while. It’s like the mad scientist in the laboratory. We put all these performances into a pot, some of them bubble away, some of them explode, some of them do something you don’t really expect.

Bill Fletcher: We might be having a quiet, whispered line, but we need it to be heard from across the street, or right up close. We need the same line, but different projections. We look for a different type of voice than other people. We’re not after ‘voice talent’, we want actors. We’re not looking for a funny voice, or a sexy voice, we’re looking for someone who adds character behind the voice, to create someone that you believe in.

BVB: We’ve worked with some of the Half-Life guys before, like Mike Shapiro, and Al Robbins, who did Barney, Dr Kleiner, and the G-Man. But I think in general, we found that it was great at the beginning of the sessions, and each successive take continued to get better. We brought some knowledge about who the characters were, and what was happening in the game, they brought this great wealth of acting experience. It was a flexible, open partnership. We’d have different ideas and make suggestions, they’d do the same. For them, I think it was a great leap of faith. They didn’t have the other person who they were acting with there, they were in isolation, and most of them didn’t have much experience of games at all.

The moment that really stood out for me was when Alyx passes through the teleporter. You see her on the monitor: when she arrives, she gives Eli a little peck on the cheek. I’ve never seen that in a game... it just nails the relationship, and perfectly sets up the later jeopardy. I don’t think I’d ever seen such a kiss like that in a game before.

ML: That was when the scene really came to fruition. It wasn’t in the script. It was implicit in their relationship, but that visual language is not... like... I’ll do everything I can with the words, but I’m finding that we can actually cut lines because the acting and animation supports silence. That’s what I’m shooting for, to get stuff into the animators’ hands that lets them go even further. We know there’s affection between them, we build it up like scaffolding, but when you’re on top of that... A lot of the animation seems to be the extra one percent that tips it over. In that scene, the kiss seems really simple. Even to me, it came out of the meaning of the scene.

It sounds almost like you improvise the final touches. Is that fair? Is there room to improvise?

BF: Not really. It’s impossible to improvise because of the sheer amount of planning that has to go in. Every movement needs so much work.

ML: You know, it seems like improv to me, when I see it. It feels like the little nugget that we’re working toward the whole time. You’re not quite sure what you’re going to do, or how you’re going to pull it off, until all the pieces are in place. You don’t really know what you’re going to work with.

BF: One thing I’m always conscious of, is that Marc handles all the dialogue, storytelling, and all that but I’m thinking “what can I do with that dialogue”, or “what can I do between the dialogue”. I’ll look at each scene and say “OK, there’s got to be a couple of moments here that I have to create. The kiss was something I was thinking about, but didn’t put in until the very end. I knew we had to put in this physical contact between Eli and Alyx. The first time I went through the scene, I couldn’t get it quite the way I wanted. Toward the end of the project, I came back to it, and worked on it again. It seems a lot of people had exactly the same reaction you’re talking about.

You seem to embed a lot of details about characters into their bodies, not just in the writing or dialogue. Like, for instance, the way Alyx wears a battered old Black Mesa hoody.

BVB: When we start creating characters we have ‘character profiles’. Instead of designing them totally from drawings, we cast faces. We’ll look at photos, find people who we think would look like the character. That would draw them into focus. We’d make sketches about what their body types are... Alyx had a sketch before we cast the face, then we ran a big casting process for the voice, which again helped inform us about her as well.

ML: Really, it’s just a bunch of individuals thinking about the hard problems. The people that think the hardest about those are the artists – they’re the ones who came up with the Black Mesa hoody. It comes from being immersed in the world, wanting to do it right. I love it when stuff like that gets thrown up, though, because it’s recognition that we’re all inhabiting the same world. It’s almost like we were all at Black Mesa together, that we recognised that she would have a Black Mesa hoody. The level designers have that level of inventiveness, too, the same level of creativity when they’re working with their tools – you see each person’s contribution as “what’s the coolest thing I could do with the little piece I’m working on”.

When you’re working on a scene, do you use pre-visualisations? Do you storyboard? Lamar beating up Barney, for instance, must have needed a lot of preparation.

BF: We probably should use storyboard. Yeah.

BVB: Yeah. [laughing] That would really help.

BF: It’s just brainstorming, really. Then we just block it in, a bare-bones scene. It’s just characters walking around, hitting their marks, saying their lines, then walking off again. Then we slowly start adding things...

ML: The Lemar sequence started as a little idea in the script. We have a lot of those off-the-wall moments. When I bring them up, mostly people will just go “pfah,”, or they get this little look in their eyes, and they’ll start to do it. My instinct is, if no-one has any really strong misgivings about an idea, I’ll pursue it, until people finally accept it. I think the Lemar sequence happened that way. I had a crazy idea “What if Dr Kleiner had a pet headcrab?” It just seems like a silly idea, but if you keep repeating it and repeating it, when people start working on it, they buy into it. It would just never occur to them that he wouldn’t have a pet headcrab.

That would happen time and again: the little thing that you thought was crazy actually gets people excited. We’re not just building the game because you have to go from point A to point B. We’re building it because there’s fun little things you get to do on the way. The details, the non-essential stuff, is actually half the fun. Things we think would be cool, but we wouldn’t put them on the box. That’s the kind of thing I look for in a game. What’s the extra bit they did, that they didn’t have to do. That’s when you really see the personality behind the whole thing.

It sounds like you really enjoy what you’re doing.

ML: We’re having lots of fun doing this. If I were working on a Star Wars game, there’d be a bunch of in-jokes we could do, but it wouldn’t be as if you were creating the universe as you did them. You don’t get that creative buzz where we’re not just making jokes; we’re actually suggesting things about the world itself. I think that other games are kind of constrained in their inspiration because they’re given a story-Bible. Everyone here knows that if they have an idea that doesn’t quite fit the world, we can sit down and talk about it and adjust the world to make it work.

BVB: There’s a certain amount of empowerment that people here have. You’ve already heard the huge amount of influence everyone has over each part of the game. It’s not one person driving their vision. Do you remember coming down into Eli’s lab, seeing the Vortigaunt chefs? That’s an idea Dhabih Eng [One of Valve’s Senior Artists] had. That wouldn’t have made it into the game at any other company. It’s silly, our game’s supposed to be a scary game, but somehow, we have Vortigaunts in chef hats.

ML: Dhabih did the most amazing models you’ve never seen. We were watching him building this stuff, putting so much detail into something no one will ever see. But he’s so excited to be doing it. Even like a curmudgeon would look at it, and say “Normally, I’d say we shouldn’t be doing this. But this is so cool we have to keep it in the game.” Then we try not to draw attention to it, we just leave it there for people to discover.

I’ve noticed that you tend not to integrate storytelling and player action. It’s rare for the player to actually be able to interact with the scripted sequences. Is that a fair observation? Is it just ‘hard’.

ML: Yeah, it is. We can do it, but it requires involving more people in the development process. We do it in Nova Prospekt, when Mossman is putting Eli in the mechanical thing while all the doors are breaking down... We don’t want to do that in every scene, it’s complicated to pull off, and a high point. It has its place. But the gameplay has to be very, very solid. It is what we aim for, though, but those scenes are definitely the hardest to pull off. Compare that scene, which is just action, to the scene where Breen has you captive, and you’re basically just watching theatre. That’s something the level designers and animators could build entirely, without having to wonder “how strong is the gameplay going to have to be to support this?”

BVB: Gordon doesn’t talk, so the player can only participate by picking something up, hitting a button, turning something on, throwing something, shooting something, moving something, breaking something. We’re really taking a lot of that control.  We try and keep hyperactive players interested with some little things, so they can bounce around, or look and play with certain toys. It’s funny how some people want to go into the scene and set up the perfect two-shot or three shot on every dialogue, and see exactly what’s happening. Other players just want to smack the characters and jump on their heads.

But there are scenes where it’s much more open, where it’s all about the gameplay, yet the characters interact with you. That’s certainly something we want to expand on, and go much deeper into, in future games that we do.

One of the interesting quirks of Half-Life is that you never actually complete the story. You never get to explain everything to the player. Why is that?

ML: Part of it is that we’re always creating the world, like we were saying before. Everyone wants to add something to the world. Because the technology changes all the time, and we want to take advantage of that, we don’t want to limit ourselves to hard statements about the world that, maybe a year down the road, we’re suddenly able to make playable, but we’ve screwed ourselves out of doing that because we cut ourselves off a year ago. It’s a function of the medium, I think. We don’t know what we might be able to do tomorrow.

The Seven Hour War, for instance, why show that when, say, in the future we could actually put you through this experience. In a sense, the world extends beyond the narrow limits of the game. It’s a fact of this medium we work in, that we don’t know what the territory is that we can reach next. It’s a function of this technology curve, new areas that are opening up right in front of us. We’re not creating it—we’re discovering it as we go.

BVB: It never occurred to us to stop and put some kind of expository section in the game, that we couldn’t actually support without the player experiencing it. That’s what we’re always trying to support, to keep you in the world. I think part of the fun is telling that story with the environment.

I thought you didn’t finish the world out of deliberate choice. That you wanted to preserve an air of mystery.

ML: More so with Half-Life than Half-Life 2. There is definitely a desire to make the game play out inside people’s heads when they leave their computer. A lot of the fans of the first game, kept wondering about it, telling themselves stories. I don’t mention it because it’s so much part of our mindset in the way we build these games. If we’re going to answer any question, we have to make sure we’re raising even more questions that are even more interesting.

How often do you play-test these scenes? Surely, as they develop, you become immune to their charms?

ML: The cinematic stuff is difficult. Usually, these scenes are a work-in-progress, so the feedback we get can be a little off. And play-testers will like all kinds of things which aren’t necessarily the point of the scene. They don’t like stuff that’s really important  to get across. At the beginning of the game, for instance, we felt a lot of the combat started too soon. Play-testers loved it, because they had a gun in their hand, and they were already breaking stuff. It was hard to say, “we shouldn’t do this, yet” because the feedback was really positive. We thought it was too early.

We wanted the combat to start after meeting the couple who are being beaten up by the metrocops. For a while there was more before that, you’d go from instantly being in the lab, to beating up cops. We wanted you to witness the cops doing something horrible, and feel like what you were doing was a response to that—not that you were just a killing machine. Of course the play-testers are going to love this, but they don’t know what notes we’re going for.

You’re fighting against their urges, to an extent?

ML: We have to be extremely clear about what we want to achieve before we go in. We’re testing it to see if they can play the game without getting stuck. Not to see if they enjoy the combat.

BVB: No one ever says “Ohh, you know what. I’d love another emotional scene right here.”

ML: Exactly. People would have been happy just to be given a crowbar in Kleiner’s lab, so they could go around smashing glassware. But they wouldn’t have had the moment where Barney drops it to you. And a lot of people love that moment. We would have had great play-tests if Alyx had given you the crowbar as you enter the lab. But holding it back, holding it back... Players won’t say “Don’t give it to me, yet.” It’s about learning to have faith in the whole arc of the thing. We’re pacing it, holding back for the payoff.

I do adore that moment. It’s such an emotional moment, where you’re suddenly empowered.

ML: Yeah. It’s amazing how those little scenes have the big payoff. That’s one people mention over and over again. It’s the first time in a game, as far as I’m aware, that a character actually gives you a weapon. In the first game we wanted Barney to give you a shotgun, but we could never get it to work. It was like “Dammit, characters are going to hand you your weapons this time. You can either catch it, or it lands on the ground. We’ve been wanting this for years. We’re not leaving a weapon on the ground, ever again.

Bill, you came to Valve from Disney. Does working in an interactive environment offer a different kind of challenge?

BF: It’s a lot harder. In films, you have it all laid out for you. Everything is in there for specific reasons, you have close-ups, you have long-shots, we use them to tell the story. In this medium, we don’t have that to fall back on.

We have to make these characters live in this space. I can’t make my life easier and animate just in close-up. You have no idea how close the player is going to be, you could be trying to create a nice, quiet, tender moment but the camera is right across the other side of the room. Controlling the camera gives you so much more power over emotions. We’re not able to do that—what we try to do is draw you across and closer in, we do it so you feel you need to get in and close to the acting.

But you can’t just concentrate on the facial, like you would want to do in a traditional animation. You’ve got to work on the body-language, too. By the same token, you can’t have them bouncing around if the player is really close to them. You choose your battles. I try and keep it pretty low-key in the more emotional moments. It’s that much more gratifying, we’re creating a character, not a movie clip.

There’s something about your work, actually. It’s in the eyes. Sometimes, I swear Alyx fancies me. Are you playing tricks on me?

BF: I can’t say there are any basic techniques. It’s about making them feel natural. I have in my head a shy but very alluring pose.

ML: Characters in the scenes do really over the top, crazy stuff to pretty subtle, day to day ordinary interaction. There’s Father Gregori, who is really over the top and crazy. The more time you spend with a character the more you have to tone it down, though. It would drive you nuts if every time Alyx looked at you, she was leering in. I was really surprised to hear people get quite choked up when you go in the elevator. She’s looking at you going down, hands pressed against the glass. It’s what we were going for... I’m just amazed they responded.

BVB: I think, too, that Ken Birdwell, the guy who set up our character technology, did a lot of work on eyes. That was the first thing he really solved. I’ve never seen a situation where they’ve got the eyes so right it actually feels like they’re looking at you. There’s that thing about the eyes being the window into the soul—it’s true.

BF: Oh yeah. If the eyes are wrong, there’s nothing I can do to make that character feel like she’s looking at you, or she’s grateful to you. Without that, this would be impossible.

ML: It gives the illusion of reality.

BF: Yeah. They can look not quite at you, at something behind or around. They actually focus, too. As you get further, or closer to you, they’ll track and narrow their field of vision. Otherwise there’s this feeling that they’re always looking slightly past you. We’re so cued into eyes. We can project so much via them. We can project intention.

That attention to detail in Half-Life 2 still surprises me. Even now, I don’t think we’ve really seen a competitor to Half-Life, no one’s really attempted to take you on. Why is that?

ML: After Half-Life came out, people pushed on it a bit, but immediately ran away. You could see, in the beginning, there were a bunch of games like that, with scripted sequence events, but no one tried to roll it through the whole game. You quickly run into how hard it is. And, it requires a really limited set of constraints for your game. I mean, the way we do it only really works in this type of game. I mean, we have a mute character. As soon as you put some dialogue in there, if Gordon speaks, or has a conversation tree, or something like that... then you might as well blow all these limitations off, and explore a completely different kind of space. I just don’t think that unless you’re building a Half-Life game, it makes sense to do it exactly this way.

Maybe a lot of games could benefit from emphasis on more non-essential character development. The other thing is that Bill’s comment on scenes coming into focus gradually... that’s one of the problems I have with the way writers are usually used in the industry... where they come in at the beginning, or the end, and they leave, or they take a developer’s script and polish it up. They never get to do that process where the game is continually focused on. We have our eye on every part of the game throughout development, and story is just one part of that. The scenes get to be that good because we’re all focusing on a little bit every time. It’s not like I came to Valve twice, looked at the script, went away, came back... We all have access to each other.

BF: A lot of people don’t see the whole picture, of what it takes. They’ll say, “We want to do great characters, I guess we’ll do it with more polygons... Or we’ll get some famous voice to do it.” It takes so much more than that. It takes everything. It’s all about the good acting, the good animation, the good writing, the good voices, the good technology... it’s all those things.

ML: The usual idea is to want characters that you’d care about as much as you would in a movie, or a book. Then they’ll make a movie. They’ll try and imitate a movie, and do a cutscene, or a bunch of text, so it’s like a book. But the reason I came here was to make games. Having these things happen for the first time in a game is what’s really cool. A lot of the stuff that we play around with is cliché and generic in any other medium. But there’s a whole dimension of it in the game space that’s completely unexplored. We don’t want to make movies. We want to figure out how we do these things in a game.

BVB: There’s a huge barrier to entry, too. It accounts for why Half-Life 2 took so long to ship. We started on our character presentation systems in late ’99, early 2000. Creating the system so characters can address you wherever you are, and other characters wherever they are, blend smoothly into a scene where Bill wants to take total control over how they’re interacting with each other... it’s a massive amount of technology. It’s how they blend walking in different directions. It’s about how Bill makes someone act when they may, or may not be walking. We had to create the system here, then expose it in tools so animators could actually function. I’m so grateful that Gabe [Newell—Valve founder] and Ken Birdwell [Senior Software Engineer] made it a priority to create this technology, so we were able to author for it.

I have this pet theory, that games are the pinnacle of human achievement. They combine every form of media, and squeeze them into a kind of playable state.

ML: Yeah. One person can’t do very much. It’s so granular. That scared me about games, originally, and it still overwhelms me occasionally. If you’re building that world, you have to build it from the atom, up. You can’t just get an actor, record them in a studio. There are so many stages—every little thing in the game has been worked on by multiple people. Are they worth the effort?

ML: Yeah.

BF: Yeah.

BVB: Yeah, definitely.

ML: When I was a writer, I would look at this huge range of people who wrote books. There was this overwhelming thought of getting into Hollywood and working on movies, all these millions of people who’ve made movies, it’s hard to figure out where you’d go to do your thing, and be really distinctive, to go and make your mark. This is the frontier, the edge. By definition, everything we’re doing is pioneering. If you did that in writing, you’d just have no audience, because experimental tends to mean so obtuse that no one enjoys it. We get to experiment in this popular form of entertainment. We’re creating worlds that have never been seen before. That’s totally worth it.

If you've ever wondered what the top of Gordon Freeman's head looks like then Half-Life: Top-Down is the mod for you—it turns Half-Life into a top-down shooter, complete with plenty of camera customisation options.

You can watch it in action above, but I wouldn't judge it until you've seen the last 30 seconds, which shows how you can adjust the camera. On the default settings the overhead camera hugs the ceilings and, as the mod's creator points out, "it makes the camera get really close when the ceiling is low". It's a bit claustrophobic, and makes it hard to see where enemies are. But it's easily fixed thanks to a neat options menu that lets you manually set the height of the camera.

If you tell the camera that it doesn't have to stay in-bounds and bump the height up then it starts to look more like a conventional top-down shooter, and you can actually see the enemies you're fighting (see how it looks in the picture below). Unfortunately it will show everything that's out of bounds in a bright red colour, but I think it's worth putting up with. Beside, modder Sockman111 is working on a solution, perhaps by placing a big object with a texture far below.

The mod also tweaks the auto-aim to account for the fact you can't aim higher or lower on an enemy, and makes it easier to interact with objects if you aren't looking straight at them. 

A word of warning: the creator says they're "not sure yet if the entire game is playable like this". One player in the comments of the mod has also reported a number of bugs. But it's something Sockman111 is working on, and most people seem pleased with the results. 

You can grab it from ModDB.

Believe it or not, learning Dota 2 is easy nowadays. I don’t mean easy in the sense that it is straightforward or comprehensible or painless. I mean easy in the sense that it is slightly less angry at you for wanting to know what’s going on. I mean easy in the sense that people like me will tell you that you’ve never had it so good. We had to walk two miles in the Frostivus snow to find a match, and whittle our own Force Staff by hand and no-one had even heard of Purge and his useful video tutorials. 

Dota Auto Chess—a spectacularly popular custom game mode by Drodo Studio—is a return to that initial bafflement. “You can pick dota heros as your chesses,” says the blurb, “and they will automatically fight for you on a 8*8 chessboard.” 

Now, I have 2000+ hours of Dota on my account. I beat my mum at chess when I was in a hospital bed, stuffed with morphine after a life-saving operation. Neither of these skillsets has proven particularly useful in Dota Auto Chess. 

The broad idea behind Dota Auto Chess is closer to deck-building games than to either chess or Dota. The basic pattern of each round is: earn money, choose whether to spend that money on heroes, position those heroes on the board, then let a fight against the heroes of a randomly chosen opponent (from the seven others you’re grouped with) auto-resolve. If you win, you get a bit more gold and maintain your health bar. If you lose, you’ll take a bit of damage. A match lasts as many rounds as it takes for only one player to be left standing. You can keep an eye on how everyone’s doing via a leaderboard on the right hand side of the screen.

Because nothing related to Dota is ever simple, there are a lot of other variables to keep track of along the way. Managing your gold is vital—you want to balance investing in heroes and getting gold through fighting, with keeping some in your pocket to earn interest, spending to level up your donkey (and thus increase the number of heroes (referred to as “chesses”) which you can have on the board), and re-rolling the hero selection.

Each hero is listed with a species and class. If you have multiple heroes from that species or class on the board you can get boosts. The orc species combo gets you a higher maximum HP for each orc, the mage class combo reduces enemy magic resistance.

As well as that interplay there’s a spatial element. Do you bunch your heroes up or spread them out? Do you try to protect a vulnerable unit or shove them to the front as a meat shield? How can you keep important combos in play by keeping the relevant units alive? That’s one part which felt like it was drawing on my actual Dota knowledge. 

Another part which taps into that knowledge is the item system. Some rounds have you facing off against non-player units—the neutral creeps from Dota’s jungles. If you beat them they can drop little treasure chests containing items which the donkey can fetch and put in its little backpack. You can then ask the donkey to deliver the items to a specific unit, thus bestowing its benefits to that unit. Essentially, it’s the courier function the donkey traditionally fulfils in Dota 2. Knowing the types of items which benefit particular heroes in the main game will give you a headstart here. If you don’t know Dota you might not realise you need to deliver the items to specific heroes at all instead of just collecting them in your pack. 

Then there’s the merge-three minigame. Plonking down three identical heroes of the same level (with one or two class exceptions) will merge them into a single, more powerful hero. This has knock-on effects when it comes to which heroes you buy, when you place them on the board, and how it raises or reduces the number of units on the board.

When I booted the game up for the first time it wasn’t even clear where I was, or how I was supposed to chess. The game tips disappeared offscreen before I’d read the first word and the camera was pointing at a rival’s board, meaning I couldn’t see the result of any of my actions. The resulting panic is how I learned that the boards of each of the eight players are presented as physical islands in a 3x3 grid. Panning around you can check in on other players or enjoy the fact that the middle board is missing, replaced by a small version of the Dota map.

Finding my island is how I discovered I needed to interact with my chesses by selecting a donkey and having the donkey do the chess on my behalf. If you’re familiar with Dota, moving your donkey around is probably also when you’ll realise it’s not actually a donkey. Instead it seems to be the hero Io (as per the lore: a multidimensional wisp billed as a Fundamental of the universe) wearing a donkey costume. You can tell it’s Io because it’s making Io’s Ibiza chillout beeping and blooping noises and trailing particle effects across the chessboard.

The existential question of “when is a digital donkey not a digital donkey” is irrelevant to play, but it’s fun to notice how pieces of the main game are repurposed in these custom modes. Again, it’s a way that Dota Auto Chess feels true to an older form of Dota—the Defence of the Ancients which emerged from the Warcraft III fan-made map cauldron, and whose quirks are often the result of units being turned to a new purpose.

I’m absolutely loving it. In each phase there are a manageable number of choices to make. Making a sub-optimal choice doesn’t feel like a total disaster. It taps into the little jolts of pleasure casual games are good at—the satisfaction of merging heroes, auto-fought battles with over-the top effects and the chance to win, a little leaderboard…

Another joy is the lack of toxicity and the lack of pressure. It probably says a lot about the confusing interface that for ages I had no sense of whether the lack of repulsive messages was because the game elicits less rage or whether there’s just no all chat function. A message in Russian during my fourth match points to the former. But, with or without chat, I often feel massive pressure in PvP games. I don’t want to embarrass myself. I don’t want to lose. I particularly don’t want to be the worst on any leaderboard.

But here, I’m playing a weird once-removed version of PvP. My squad of heroes is mostly pitted against the heroes of a human opponent, but the other person isn’t spectating that match. They’re looking at a different chess board, watching their heroes take on a randomly chosen selection of someone else’s heroes. It might end up being mine, but it might not. Me winning or losing doesn’t affect them directly. We’re just having our armies and choices calibrated and recalibrated against each other. And it’s this distinction which takes the sting out of the competition without damping the pleasure of winning.

But what would a free-to-play game within a free-to-play game be without cosmetic microtransactions? An excellent question, dear reader. Well, you can earn or buy candy—the premium currency and spend it on spins of a slot machine. The rewards from spinning this machine are different couriers. So it’s not pay-to-win, just a different look for your non chess piece character. 

And it’s not pushy either—a real contrast to the Dota client it sits within. While logging in to Dota 2 to access the custom game section, Valve immediately invited me to spend £28 on an outfit for a character I don’t even play. After I refused, it reminded me I can spend £2 to open a seasonal treasure chest. At some point I fully expect the Steam store will stop trading in cash and start accepting the souls of children in exchange for digital hats. But I digress.

The above should give you a sense of both the low barrier for entry (“low” being a relative term and entirely skewed by Dota’s base level of nonsense) and the ridiculously high skill ceiling of Dota Auto Chess. It manages to be similar to and the polar opposite of Artifact’s considered design and overwhelming complexity. 

It’s a joyful, weird, opaque project—a hodgepodge of casual mobile gaming compulsion and PC gaming at its most bloody-mindedly hardcore—spitting personality and spell effects from every angle.

If you want to get into the mod yourself, check out our Dota Auto Chess guide.

Here’s a story you’ve heard before: A mod for a popular strategy game takes the existing ideas of the game and turns them sideways, forming a new kind of game in a new, weird genre without adhering to the usual game design conventions. That’s the story of Dota 2, and that’s also the story of Dota Auto Chess, a recent Dota 2 Custom Game that’s attracting a lot of attention. Auto Chess is worth trying out if you’re interested in strategy games or digital card games, and not just because it’s free. It features mechanics from set collection card games, real time strategies, tower defense, and even from mobile gambling. 

Here’s how to get started while you read this guide: Download Dota 2, then download Auto Chess.

How do I play Dota Auto Chess?

In Dota Auto Chess, you control the composition and placement of a team of heroes on a chess-like board. Each round of Auto Chess sees you buying new heroes and placing them on the board to try to win a fight against either neutral Creeps or one of your seven opponents' teams of heroes. Buy and place is all you do, though: The real-time fights between the heroes are outside your control. That’s the Auto part of Auto Chess. If you win, you move on to the next round and get some gold. If you lose, your courier—your controlling avatar—loses hit points based on how many enemy units are left alive and how strong they are. Then you have 30 seconds to buy new heroes and manage your army before the next round starts in an ever-revolving round robin tournament. The last player standing wins.

When you place heroes on the board, placing more than one of the same kind will upgrade that hero to the next level. The number of total heroes you can have on the board is equal to the level of your courier, and leveling up your courier takes precious gold, so upgrading your heroes lets you get the most out of limited space. Heroes also get bonuses based on how many of that type are on the board. 

Upgrading or getting those type bonuses means you have to collect sets of heroes, and each round you draft heroes from a selection drawn randomly from a common pool with a fixed number of pieces shared among all the players. In effect, this means not every player can share the same strategies, and you have to watch what others are buying so that you know if you’re competing for the same pieces. You might also get items from fighting creeps that help you form a strategy. 

I’ll use the terms early-, mid-, and late-game a lot in this guide. The early game is generally considered to be the first four levels, the mid to be levels five to seven, and the late to be levels eight to ten.

If this all sounds random and surprising and hard to keep up with, that’s because it is—at least at first. So you have to focus on aspects of the game that you can control. 

The Auto Chess economy

Economics are king in Auto Chess, but feature some remarkably unintuitive aspects. The game isn't made any easier by balance updates that come nearly every day of the week, especially when they change which heroes cost what and therefore how many are in the pool.

Each round you’ll get a base gold income. You’ll also get a single extra gold for winning a round. You also get gold for a streak—either winning or losing—that can stack up to three per win or loss in a row. Each round you also get 10 percent interest on your gold stores, rounded down—saving up 50 or more gold at the mid game for late game interest is important. (It’s something you can control!)

Economic strategy, therefore, is to either win or lose for a few games in a row, but never alternate between the two. The poorest players in a game are the ones who alternate between winning and losing each round. Losing on purpose can be good if you’re struggling—it’ll cost you hit points, but it’ll also stack up your losing streak bonus and let you get back in the game before you’re out entirely. I like to just commit to the loss until I get to less than 40 health while focusing on only the most valuable heroes, and then combine those heroes to a mid-game surge and spend big to try to win.

Placing three heroes of the same kind and level on the board will upgrade them into a single, stronger hero of a higher level. Heroes cost an amount based on their power—between one and five—but take the same number of duplicates to upgrade no matter their base cost. So it’s pretty easy to make a one cost hero into a level three. You just need to spend the nine gold to buy nine copies, and there are 45 copies of a one cost hero in the pool, so you’ve got a good chance of finding them in the draft. Meanwhile, it’s expensive to make a four or five cost hero (36 or 45 gold, respectively) into its level three form—not to mention that you’d need to get randomly dealt and then have the gold to buy 90% of the available supply of a five cost hero to level it up.

Your early hero buys might not fit into a larger strategy very well, but don’t be afraid to buy low level pieces you might not use long term because level one heroes can always be sold back for their full value. You’ve got eight reserve hero slots for precisely this reason. You can spend money to refresh your available pool and hunt down the pieces for your combo, but that costs valuable gold and will guarantee your defeat if you do it too much in the early to mid game. Remember that you can lock the pool if you want a piece but can’t afford it until next round—just don’t forget to unlock it.

How do Auto Chess heroes work?

Heroes are like any unit in an RTS or RPG: They have health, mana, damage, armor, and the like. They have an attack speed with its own animations and quirks based on model—these are things very familiar to Dota 2 players and accessible on the game’s wiki. Others are weird and poorly understood or documented at this time, like movement and range. Range is measured in the imprecise way of Dota 2. Suffice to say that each chess space is about 200 range, and most heroes with range can hit two spaces away at 400 range. Dwarf heroes like Sniper have 300 extra range, for a total of 700, so they can hit most of the board from any space. Movement is… less clearly delineated. Some heroes, like assassins, can leap the whole board in a move. Others, like Tiny, plod along one space at a time.

Each hero also has a single ability, which it uses by spending its mana and which it will almost always automatically use as soon as it can. Heroes get mana by dealing or taking damage with their attacks,  so survivable or long-range and high damage heroes generally get to use their powers more often. Higher level heroes have more hit points and do more damage, so they’re also more likely to get to use their abilities.  

Where you place your heroes on the board has a lot to do with this. You have half the 8x8 board to use, that’s 16 spaces for as many as 10 heroes. Putting squishy heroes in the back and tanks in the front is good, but knowing which tank needs to be supported to survive and use its ability versus which tank can just take the hits and go down fighting is key. Putting your tough, upgraded Timbersaw up at the front is good because his ability cooldown is very low, so he’ll take lots of damage and get to use that ability a lot. Tidehunter, on the other hand, needs to take damage to get off his powerful Ravage stun as quickly as possible, but it has a huge cooldown, so he won’t need a lot of mana long term. Keeper of the Light needs to be in corners or at the side of the field because his power is a large line. Compared to the rest of Auto Chess, hero placement is actually fairly intuitive when you’re just getting started—more complicated placement combos and flanking strategies can wait. 

Composing a team isn’t just about upgrading whatever you can buy for cheap, it’s about synergising the abilities of what you do buy and knowing when you diversify out of what you’re already specialized in. Goblin Mechs are strong in the early game, for example, but their effectiveness tapers off against other late game combos. Synergy bonuses from species and class are gained by having more multiple unique units of the same class and race on the battlefield at a time. Three or more Mages, for example, decrease the magic resistance of every enemy unit.

Choosing synergies is tricky, dependent on others’ team compositions, and can be a trap if you’re not wholly committed or if you overspecialize. Three or more warriors are always good, because warriors increase every warrior’s armor by a stacking amount when you’ve got three and six of them, and there are a lot of warriors from every species. Elves are a limited species specialization, on the other hand, and give each other evasion—but only the elves will benefit, so you have to tailor strategy for that. Two undead, meanwhile, is almost always worth it because they debuff every enemy’s armor. Two nagas is invaluable, boosting every hero on your team with extra magic resistance.

Popular strategies for starting players are often Warriors early game into Mages or Assassins late game, going all-in on Knights and Undead using Luna and Abaddon, or the often-dominant Goblin Mech combo that double dips on species and class synergies using the four different Goblin Mech characters. Here’s your real warning, though: Every game of Auto Chess has its own metagame based on picks, and the game itself has a meta based on strategies. Going for the dominant or popular builds can lose you the game when everyone else is going for them too.

All this and items, too

Rounds 1, 2, 3, 5, and every fifth round after that are the creep waves, where you fight a group of neutral enemies instead of an opponent. Killing those enemies can give you valuable items to equip heroes with, and those items can make or break a strategy or win a game. No take-backsies on item equipping, though—once you give a hero an item it’s theirs forever. 

Item management is an advanced tactic, so to start just put them where they’ll do good—mana items on a mage hero, armor items on a warrior. There are also, like in Dota 2, item recipes based on equipping a specific set of items to combine them into a more powerful item. These are obscure and weird to beginners, and some differ significantly from what a Dota 2 veteran might expect. I’d recommend just keeping the recipe list on the Auto Chess wiki open on a second monitor or printing it out for ease of reference. 

At this point, you should get in there and play. Don’t be afraid to try stuff out and win or lose. First start to learn the class and species synergies so you can build teams, then the hero powers, then worry about bigger concerns like watching your opponents or building up a stable of go-to strategies.

Advanced Dota Auto Chess tactics

Once you’ve played your first few games you’ll have space to really start to learn combos. You’ll figure out which heroes you like, which heroes nobody else likes, and how to combine those into a winning team. Here are some key tips moving forward from the basics:

1. Make an early commitment to a simple strategy and go all-in on it. You want to win early games, remember, to get that win streak bonus built up and kill those creeps for items. Commit early, then base your actual, long term strategies around your mid-game hero draws and items. Dota Haven has a good guide on the game phases if you’re struggling with how to build teams for different parts of the game. Likewise, a few lucky items drops can make or break the game for you. If you’ve got lots of lifesteal, build a powerful auto-attack team that really values those items. If you luck into a powerful item recipe, be willing to spend the extra gold required to make the equipped hero into as strong a piece as possible, even if it hurts you economically.

2. Watching your enemies’ moves is hard to do in the 30 seconds you have for strategy, but it’s vital. Knowing what others are buying lets you know what you can buy in order to take advantage of surplus pieces in the pool. A few convenient leveled up heroes because nobody’s buying Knights, for example, can easily win you the game. If you’ve decided to take a few losses to hoard money, spend that time watching enemies’ compositions so you can work against them: You’re going to lose anyways.

3. Placement can really matter when you bait valuable enemy heroes away from optimal positioning of their own. If you know your enemy is deploying a central Tidehunter with its huge area of effect stun, try to put a less valuable unit of your own forward and to one side in order to pull it away from the main body of your force.

4. Balance changes can come fast and furious, often two or three times a week. Check out unit tier lists for a shorthand, but watching the Auto Chess subreddit is one of the best ways to understand what’s going on. For now, know that Crystal Maiden is awful and should almost never be picked. Kunkka is godlike good even if he’s useless for your overall strategy, and should nearly always be picked. Tidehunter is similar, if trickier to use, but also plays into the valuable Naga synergy bonus.