Last month, Microsoft announced plans to bring Xbox Wireless tech to PCs and other gaming accessories starting with Lenovo’s IdeaCentre Y710 Cube. Today, the company went a step further with controller support for Samsung’s Gear VR headset. That’s right, the Xbox Wireless Controller will soon work with Samsung’s phone-driven virtual reality device. Minecraft: Gear VR Edition will be the first game to make use of the feature, but eventually you’ll be able to use it on any Gear VR title with controller support.
While the Gear VR version of Minecraft is already available, an update next month will flip the switch on the controller compatibility. You’ll also need to make sure you have the latest update for the Xbox accessory as well, which is already available. When the game update arrives in October, all you’ll need to do is download it and connect the controller via Bluetooth to start playing. Microsoft says it’s working to bring Xbox controller support to both existing and future Gear VR games and promises more details “in the coming months.”
Source: Xbox Wire
Video game journalist and presenter Geoff Keighley is teaming up with YouTube for a new live show. It’ll be broadcast every Thursday at 8pm ET/5pm PT under the YouTube Gaming banner (presumably, you’ll be able to watch it through regular YouTube too.) Rather like Keighley’s E3 show, it’ll have trailers, gameplay footage and developer interviews, all wrapped up in a professional studio environment. Popular YouTubers Nadeshot, iJustine, MatPat and iHasCupQuake will be popping up too (this is a YouTube show, after all) no doubt to raise interest and awareness.
YouTube Gaming is far from the first to tinker with a live show format. IGN hosts Up at Noon, Gamespot has The Lobby — the list goes on and on. YouTube is in a unique position, however, because it’s a service provider first, content producer second. If nothing else, it can ensure the new show is visible to the millions of people that access its site and apps every day. Keighley is a big name, and will bring a level of credibility to the show — but it’s the YouTubers that have the biggest potential to grow its audience. That’s important if YouTube Gaming — the platform, rather than this programming push — is to ever grow into a true Twitch competitor.
Source: YouTube Gaming (Twitter)
Days ago, Bungie released its third expansion, Rise of Iron, for its popular MMO shooter Destiny. A new short single-player campaign, multiplayer mode and six-person raid should keep players busy killing enemy aliens, and each other, for months. Undoubtedly, some of those gamers are parents, and while they pummel and gun down foes in-game, they might look wistfully at their kids sitting on the couch next to them. How can my progeny engage in this rich, bloody, grim universe with me, they’ll ask? The answer, obviously, is a children’s book.
The rhyming alphabet book, D is for Destiny, was envisioned to be the kind of bedtime story read to the game’s last human children huddled in Earth’s only remaining city. “A is for Adventure,” parents will say. “B is for Be Brave,” the City’s rallying cry, an aspiration when the kids depicted on the page are an endangered species. “C is for Cabal,” the alien goombas on Mars with those huge hitboxes that are joyfully easy to headshot. “G is for Grind. Forever.” And so on.
Perhaps this is unjust. Popular science fiction has a delightful history in watering down vicious antagonists for a young audience, best seen in Darth Vader’s transition from galactic butcher and child-killer to asthmatic mascot appearing on kids’ backpacks and lunchboxes. Fittingly, the 40-page D is for Destiny was first made as a holiday gift in 2013 given to parents at Bungie, and the pages are well-illustrated. The book will be released on October 4th and retail for $15.
[Photo credit: Polygon]
“We recognize that the current ecosystem isn’t consistently profitable yet for team owners or for the league.”
That’s how Riot Games’ directors of eSports Jarred “Bradmore” Kennedy and Whalen “Magus” Rozelle laid out the studio’s plans to funnel more money to professional League of Legends teams in 2017. The changes come after four years of explosive growth across the eSports industry, with League of Legends leading the charge, but it also follows a public spat between coaches and founders about the game’s booming economy.
First, the changes: In a blog post today, Bradmore and Magus detailed plans to share revenue with players and teams, starting with changes to the in-game economy. Beginning with this year’s World Championship tournament in October, Riot will share 25 percent of the revenue it earns from selling Championship skins with the winning players, team and league. (Championship skins are in-game outfits and animations inspired by the winning team, after all). The revenue share will be valid for one year. Riot will also retroactively share revenue with previously victorious players, teams and leagues.
In 2017, teams will be able to receive more revenue from sales of in-game goods, including branded items and eSports promotions. However, to address the immediate need for more money across all 13 leagues, each league will set aside a guaranteed minimum for each of its teams in 2017.
“For example, the EU LCS will have a minimum revenue amount of €100,000 per team for the full season, of which 50 percent will go to players as supplemental income on top of their existing salaries,” the post says.
Starting now, 25 percent of the revenue from all sales of Championship skins and wards will go directly to the Worlds prize pool. The 2016 World Championship Finals’ prize pool starts at a standard $2 million, with $1 million provided to the winning team.
“For context, had this been applied last year, it would have more than doubled the prize pool,” the blog post reads. Additionally, 25 percent of Challenger skin sales will be added to the prize pool at MSI, a high-profile mid-season tournament.
These changes follow a public argument over Riot’s financial acumen and its regular release of major, game-altering updates. Andy “Reginald” Dinh, the owner of North American League of Legends team TSM, recently said in an interview with theScore esports that Riot’s habit of constantly changing the game led to player burnout and made it nearly impossible to properly train. He also noted that it was difficult to secure revenue and investors with such a chaotic foundation. Riot Games co-founder Mark “Tryndamere” Merrill answered these claims in a Reddit post that accused Reginald of mishandling his own finances, to put it lightly.
Source: League of Legends
After a summer of test runs, the full version of Minecraft: Education Edition will officially launch on November 1st. When it goes live, the service will require a $5 yearly membership per user or a district-wide license, but the Early Access edition is still free until November.
According to the MinecraftEdu team, over 35,000 students and teachers around the world have been playing around in Minecraft’s sandbox since the program went live at the beginning of the summer. With the official release, the team has built out a few new education-focused features like a “Classroom Mode” that offers a top-down look at the Minecraft world via a companion app. In the app, teachers can manage world settings, talk to students in-game, give out items or teleport their kids around the map from a single interface. As the main Minecraft world evolves and gains new features, so will the education edition, and educators are encouraged to submit feature ideas and feedback.
Finally, for any teachers who haven’t stepped into Minecraft’s blocky world yet, education.minecraft.net offers some starter worlds, tutorials, free lesson plans in subjects ranging from storytelling to city planning, and a mentoring program to connect them with other educators. At launch, Minecraft: Education Edition requires OS X El Capitan or Windows 10, plus a free Office 365 account to use.
If you were hoping that the handheld version of Super Mario Maker played in three dimensions, take a seat. Polygon has stumbled across the GameStop listing for the 3DS edition, the box for which comes with a prominent caveat that it only plays in two dimensions. It’s not that much of a surprise, given how few 3DS titles really harness stereoscopy in a meaningful way — even Pokémon X and Y mostly saved it for battles. Not to mention, of course, that Super Mario Maker is the most two-dimensional of games, and certainly won’t need any extra depth. If you can’t wait to try your hand at becoming the next Miyamoto (spoiler: it’s hard), then it’ll set you back $39.99 on December 2nd.
On the island of Santorini, Greece, a group of AIs has been facing off in an epic battle of Doom.
This is VizDoom, a contest born from one man’s idea: To improve the state of artificial intelligence by teaching computers the art of fragging. That simple notion then spiraled into a battle between tech giants, universities and coders. Over the past few months they’ve all been honing their bots (known as “agents”), building up to one, final death match.
Okay, it was a lot more than one match. But that doesn’t sound nearly as dramatic.
The competition is all about machine visual learning. Just like when you or I play Doom, the agents can only make decisions based on what they “see,” and have no access to information within the game’s code.
There were two “tracks” for agents to compete on, offering very different challenges. Track 1 featured a map known to the teams, and rocket launchers were the only weapons. The agents started with a weapon, but were able to collect ammo and health kits.
Track 2 was a far harder challenge. It featured three maps, unknown to teams, and a full array of weapons and items. While Track 1 agents could learn by repeating a map over and over, agents competing in Track 2 needed more general AI capabilities to navigate their unknown environments. Both maps were played for a total of two hours, with Track 1 consisting of 12 10-minute matches, and Track 2 consisting of three sets of four 10-minute matches (one for each map).
As you might have expected, the winners for both categories came from the private sector. The agent “F1,” programmed by Facebook AI researchers Yuxin Wu and Yuandong Tian, won Track 1 overall, besting its opponents in 10 of 12 rounds. For Track 2, “IntelAct,” programmed by Intel Labs researchers Alexey Dosovitskiy and Vladlen Koltun, put in a similarly dominating performance, taking the victory and winning 10 of 12 rounds. But while Intel and Facebook may have won the overall prizes, there were other impressive performances. Three standout bots, “Arnold,” “Clyde” and “Tuho” came from students.
Arnold is the product of Devendra Singh Chaplot and Guillaume Lample, two masters students from Carnegie Mellon University’s School of Computer Science. Their team “The Terminators” competed on Tracks 1 and 2, and saw success on both. In fact, Arnold was the only agent outside of Facebook and Intel to win rounds. On Track 1, each bot had to skip one round, and F1’s departure gifted round 3 to Arnold. In round 6, though, Arnold won outright, besting F1 by 2 frags. The result never looked in doubt, though, and Arnold ended in second place, 146 frags behind F1.
Track 2 was where things got interesting. Arnold was competitive in the first map, but IntelAct already had a 19-frag lead heading into map two. On the second map, however, Arnold suddenly came alive. It won the first two rounds, closing the gap down to just 11 frags at one point, and ending the map 15 behind. But it wasn’t to be. IntelAct excelled at the final map, scoring 130 frags in just four rounds, and destroying the plucky underdog’s hopes of pulling off an upset. Arnold lost the overall count 256 to 164, again ending in second place.
Behind the scenes, though, all the work as long as several months ago. Arnold is one of the more ambitious efforts in the VizDoom competition, combining multiple techniques. It’s actually the result of two distinct networks. The first is a deep Q-network (DQN), a technique Google DeepMind pioneered to master 49 Atari 2600 games. The second is an deep recurrent Q-network (DRQN). It’s similar to a DQN, but it processes information in a directed cycle, and uses its internal memory of what’s come before to decide what to do next. Arnold’s DRQN has been augmented to help the agent detect when an enemy is visible in the frame.
In a death match, Arnold can be in one of two states: Navigation (exploring the map to pick objects and find enemies) or Action (combat with enemies), with separate neural networks handling each. The DQN is for navigation. It’s responsible for moving the agent around the level when nothing much is happening, hunting down items and other players. As soon as an enemy shows up on the screen, however, it hands control to the DRQN, which sets about shooting things. Combining these two methods, which can be trained in parallel independently, is the key to Arnold’s success.
But Arnold’s creators aren’t interested in pursuing an unbeatable Doom agent. Instead, they saw VizDoom as a nice application to test their ideas on reinforcement learning. Speaking by phone, Chaplot explained that the networks deployed in Arnold can be applied to robotics in the real world. Navigation and self-localization are a real challenge for machines, and the team is now focused on solving those issues. They’ve published their initial findings from Arnold and VizDoom, and are using what they’ve learned to try and create better robots.
Clyde was created by Dino Ratcliffe, a PhD candidate at the University of Essex in the Intelligent Games and Game Intelligence program. A one-person effort, the AI competed on Track 1 only. Though Clyde never won a round, it was extremely competitive throughout, besting Arnold in five rounds and, in one match, losing to F1 by only one frag. It ended the competition in third place with 393 frags, putting it 20 behind Arnold and 166 behind F1.
It could have gone so differently for Clyde. Ratcliffe began development in order to understand “what the state of the art in general video game playing” was for AI right now. He used asynchronous advantage actor-critic (A3C), an advancement in the DQN method that uses multiple neural networks learning in parallel to update a global network.
Ratcliffe told me he took a hands-off approach to training, preferring the agent to learn by itself what enemies are, what death is, what health packs are and so on. “I think it’s dangerous to start encoding your own domain knowledge into these agents as it inhibits their ability to generalize across games,” he explained. “I simply gave it a reward for killing opponents and increasing its health, ammo or armor.”
But a catastrophic failure — Ratcliffe’s PC power supply blew up 24 hours before the competition deadline — caused Clyde to only complete around 40 percent of its training regimen. That meant that it had learned from 30 million frames, rather than the 80 million necessary. The biggest downside of this incomplete training, Ratcliffe explains, is that the agent still occasionally commits suicide. It’s for this reason that Clyde got his moniker — he’s named for the weakest ghost in Pac-Man, who rather than pursuing or holding position, just moves around at random.
Clyde learned a simple form of spawn camping
The fully trained Clyde, which wasn’t submitted, is far stronger. Ratcliffe said he’s observed Clyde using a simple form of “spawn camping,” a much-maligned tactic in multiplayer shooters where you wait at strategic points on a map and kill players as they spawn in. “It notices certain corridors that have spawn points close by and shoots more,” he explained. This behavior is apparently in the competition version of Clyde, but not as noticeable.
Before the results were published, Ratcliffe said he didn’t think Clyde would be competitive, so a third place rosette is definitely above expectations. Ratcliffe has already moved onto a new project: 2D platformers. “I had only started looking into deep reinforcement learning around one week before the competition was announced,” he said. “I pretty much had to learn the whole field in the process of competing, and that was the point of me taking part. So I now have a solid foundation to start my own research this year.” While other agents have mastered 2D platformers, he wants to teach one to learn Mario, and then try to apply that learning set to other games without retraining.
The final prize-winning spot was taken by Anssi “Miffyli” Kanervisto, an MsC student at the University of Eastern Finland’s School of Computing. His agent Tuho (Finnish for “doom”) is a one-person effort, created with oversight by Ville Hautamäki PhD, from the same University.
Some of Tuho’s best performances came on Track 1, where it managed to finish second place behind F1 in three rounds. It ultimately placed fourth, just outside of the prize rankings. On Track 2, it didn’t get close to challenging F1 or Arnold. It put in a solid performance, though, on the first and last map, which was enough to balance out a disastrous showing on the second map. Tuho ended up in third place with 51 frags. That’s despite spending the four middle rounds killing itself more than others.
Kanervisto built a complex agent in Tuho, with a navigation system based on multiple techniques. The most important aspect is a dueling DQN — two networks using different methodology to provide a better end result. Tuho’s shooting and firing system is largely based on image recognition, matching potential enemies against a manually recorded library of images.
It was trained to prioritize movement speed in order to get it running in straight lines, and the result, Kanervisto says, is a “well-behaving model that was able to move around and not get stuck, although it struggled with doorways.” But the entire training regimen took place on his personal computer with an Ivy Bridge i7 processor and GTX 760 graphics card. You typically need a very powerful computer, or better yet several, to train an AI at a reasonable speed. Because of this, he was limited in the size of the network and input image size.
Everyone’s a winner
It may be a mostly false cliché, but at least with VizDoom, it feels like everyone here is a winner. Arnold’s creators will receive €300 for their agent’s performance on Track 1, and €1,000 for Track 2, leaving them with around $1,450 to share. Ratcliffe earned €200 ($222) for Clyde’s third place. Tuho bagged Kanervisto €500 ($558) for its exploits.
Some are going home with prizes, but all the teams I’ve spoken to have gained a lot from their experience. Take Olivier Dressler, and his agent “Abyss II.” Dressler is a PhD candidate in Microfluidics (bioengineering) at ETH in Switzerland, and had no previous experience in AI. I asked him what he’d learned from participating in VizDoom. “Literally all my machine learning knowledge” was the answer.
Dressler based Abyss II on the A3C algorithm, and had to learn everything as he went along. This led to some big mistakes, but lots of gained knowledge. One such lesson came in training. “Shooting is required to win,” he explained, “but shooting at the wrong moment (which is nearly every moment) will results in suicide.” The map was full of small corridors, and any explosion nearby will kill the agent. Just overcoming that is a challenge in itself.
Abyss II placed seventh on Track 1, but from speaking to Dressler before the contest, it was apparent he would be happy regardless of the result. “Given the short time frame I really don’t expect my bot to perform particularly well but it has been an amazing challenge,” he added. “It has even paid off more than I expected and I can use this knowledge very well in my current work.”
VizDoom will also have knock-on effects. Google DeepMind and other leaders in machine learning, despite not formally entering the competition, will also have learned a few things. Doom is a highly complex title, and various DQN, DRQN and A3C-based agents have performed to great success.
I don’t know what methods Facebook and Intel employees used to win the top prizes in their categories, but it’s likely we’ll see papers published from them soon. Regardless, as is often the case with AI, the innovative techniques used to win VizDoom will serve to strengthen every researcher’s knowledge of vision-based machine learning.
The original Dishonored was all about choice. Corvo had a mission, but how he completed it was up to you. Battle the guards head on? Slow down time and quietly sneak past them? Or lurk in the shadows and summon a swarm of bloodthirsty rats? Dishonored 2, which comes out on November 11th, looks to offer a similar, if not greater level of freedom through its twin protagonists Corvo and Emily. To showcase the game’s impressive flexibility, Bethesda has put together a “Creative Kills” video which teases some new powers and how they can be combined to devastating effect.
In one playthrough, Emily is seen chaining enemies together and then pulling them through an electrified gate as a group. In another, she summons a doppelganger on the ledge of a building and then links it to some troops down below. Seconds later, she throws her decoy with telekinesis, launching the enemies like a makeshift catapult. Corvo, meanwhile, has access to “Springrazor” mines which throw out shrapnel upon impact. We see him stopping time to flank an attacker and plant a mine on their back. As soon as time resumes, he kicks the soldier with “Blink Assault,” throwing them like a grenade into a larger pack of thugs.
If you’re a diehard Dishonored fan, you might be interested in the achievement/trophy list for Dishonored 2. Bethesda has been kind enough to publish the lot online, but warns that taking a peek could spoil a few of the game’s surprises. If, however, you don’t care for Corvo and Emily’s story, it’s a chance to start plottin some truly ludicrous assassinations.
Source: Bethesda (YouTube)
Streaming your favorite smartphone apps on Twitch is pretty hard. Through Facebook? Even harder. To simplify the process, Bluestacks is adding a Facebook Live option to its desktop Android emulator. So whether you’re using a PC or Mac, you can now grab your favorite games from the Play store, launch them and stream using the same interface. Bluestacks added a similar Twitch-streaming option just a few month ago — while Amazon’s platform is known for gaming, Facebook clearly has similar ambitions. (You can already broadcast Blizzard titles on the social network, for instance.) With so many eyeballs, its influence is growing fast.
Bluestacks isn’t perfect, however. Portions of the app look a little rough and browsing Android can feel unintuitive with a mouse and keyboard. Some players will, inevitably, always prefer an authentic mobile experience too. Even if you have a Windows laptop with a touch screen, it’ll never be quite the same as using a real smartphone or tablet. Slower-paced games translate pretty well, but anything fast and swipe-heavy can be hard to control. Of course, Bluestacks can tackle other types of applications too. If you want to explain how a new feature works in Instagram, for instance, or react to a movie trailer on YouTube, this could be a solid option.
Overwatch may only be four months old, but Blizzard’s multiplayer shooter has already amassed 15 million players and is gaining a lot interest from eSports fans. Many of the top eSports teams are quickly adding players to their roster as the world’s biggest invitationals find new ways to host them. However, a competition hosted by Blizzard itself may give the game the boost it needs to gain notoriety in a scene dominated by League of Legends, DOTA 2 and CS:GO. It’s called the Overwatch World Cup and we now know the 16 teams that will battle it out for global supremacy.
Over the past month, 50 national teams from around the world put forward their best players to show off their competitive skills in best-of-three online qualifiers. Teams are comprised of six members, who were either picked by a team captain or voted in by Overwatch fans themselves. Many players have made a name for themselves either through casting their games online or already belong to an eSports team and, like a real sporting championship, have been selected to represent their country based on their ability.
The final teams are Australia/New Zealand, Brazil, Canada, Chile, China, Finland, France, Germany, Russia, Singapore, South Korea, Spain, Sweden, Taiwan, Thailand and USA. Team USA is headed by Seagull, who also represents Shaq’s NRG eSports group, while Team Benelux has it’s very own Harambe. The World Cup will begin on October 29th with a round-robin group stage that opens into elimination finals on November 4th and 5th at Blizzard’s annual convention, BlizzCon, in Anaheim.
The World Cup is important for Blizzard to open up Overwatch to fans who haven’t yet entered the world of competitive eSports. By grouping the world’s best players into national teams, viewers can follow the progress of their country through the competition, exposing them to pro players that they may not have heard of. Blizzard will hope that the tournament will be an on-ramp for casual players, who will maintain that interest in Overwatch’s competitive leagues.
Source: Overwatch Blog