Three Lane Highway is Chris' weekly column about Dota 2.

There's a certain kind of comment I see below these Dota 2 columns that has always made me think. They show up every other week or so, and usually run along the lines of 'I don't get it', 'boring game', or 'who even plays this?' Standard anonymous point-scoring, for the most part. Comments by people who have an opinion and don't care if it's true or relevant. These are bad-faith lines of enquiry: the commenter doesn't care for an explanation, they just want to be seen asking the question.

It's an interesting question, though, and one that has stuck with me. How has this genre of game, unfriendly, competitive, complex, time-consuming as it is, become so dominant? It feels like we've skipped a step: gone straight from 'check this out' to 'I'm sick of cash-in MOBAs' without the exploratory middle-period in the genre's life. That this is the most popular form in PC gaming at present goes largely unquestioned: the reasons why—and the lessons we might learn from interrogating them—have been more elusive.

This, also, from a genre of game that on the surface lacks any of the game mechanics that you might think of when you consider compulsive loops or systems designed for player retention. Your progress in a given game of Dota 2 is reset when the ancient explodes. Any collecting or levelling you might do in the game is purely cosmetic and entirely optional. Your 'score'—your matchmaking rating—fluctuates based on performance and is not intended to be grown beyond the level suitable for you. Indeed, popular as trying to game the MMR system might be, doing so is totally contrary to the system's purpose—unless you genuinely are improving as a player, in which case the system is working as intended.

MMOs are popular because the entertainment value of moment-to-moment play is matched by tangible, measurable progress in areas that are visible to other players: your character's level, their gear, your guild's progress through a series of raid bosses, whatever form it takes. Your play serves a greater purpose, and it's this purpose that keeps you coming back. The same is true for Farmville, for Fallen London, for any other example you might pick: persistent progress brings you back.

I don't believe that the potential to jury-rig a progression system within Dota 2's existing structure accounts for its popularity. Collecting cosmetics, gambling on professional matches and grinding out MMR are marginal pursuits within the hobby, not intrinsic to its appeal. That appeal, then, is something fundamental to the act of play itself—the simple (or not so simple) reality of the game, divorced from whatever hunter-gatherer instinct it might otherwise be tickling.

There are, I think, two principles at work here. The first is what I'd call the genre's 'emotional efficiency'. Dota 2 is, functionally, a competitive micro-RPG that provides opportunities for individual and collective heroism. The emotional 'payout', here, is the feeling that you or your friends have achieved something genuinely noteworthy; that you are special, powerful, skilled, fortunate.

The previous best example of a publisher-popular 'cash-in' genre, the MMO, chases the same feeling across years of player commitment. In this example, the feeling of power comes from toppling a raid boss following weeks of preparation. The problem that MMOs face is that this feeling is ultimately an illusion. The same boss will be defeated in the same way by many other people. The only players for whom genuine heroism is an option are those who achieve world-first raid wins, a tiny fragment of the population equivalent to the number of people who play Dota 2 professionally.

Even though the vast majority of people will never become pro players, each Dota 2 match (or League match, or Smite match) is its own competitive space, distinct from every other instance of that competition that has ever taken place. You will never encounter the same combination of players, characters, items, scenarios twice. Acts of skill or power are legitimate, in this context, because they are unrepeatable. Your heroism is not an illusion because you only get one shot at it.

Furthermore, these games achieve this feeling with a relatively small number of tools—a pool of characters, a set of game mechanics, a single environment—and each instance of the game takes a manageable amount of time to play out before the board is reset. Contrast with the MMO, where offering new opportunities for heroism requires constant work by the game's developers: new areas, monsters, missions, narratives. A lane pushing game can achieve the same thing with a single new character, or with a balance patch. The format is efficient, which makes it manageable for players to consume in vast amounts and practical for developers to create and maintain.

The MOBA is the emotional payout of an MMO in tablet form; minus the years-long social investment, plus the compelling quality of an experience that can be repeated over and over and over in the space of a single day.

The second key principle is that persistent progression mechanics and compulsive loops are not entirely absent—their expression simply takes a different form. You might talk of World of Warcraft being addictive because it combines fixed ratio reward schedules (reliable gold and experience from quests, exploration and farming) with variable ratio schedules (the chance for random loot or lucky wins on the auction house.) The casino analogy feels more applicable here than it does with Dota because the rewards being offered are analogous to the physical rewards most associated with gambling: gold, coveted items. The fixed ratio schedule gives people an incentive to keep investing their time, and the variable ratio convinces them that they'll one day win big. The result is a traditional compulsive loop.

This same combination is present in almost any other game with loot, and that's where you'll find it talked about most often with regard to games—my friend Matt Lees discussed the same subject in this interesting video about the console MMO Destiny. These principles are equally applicable, however, to the psychological processes operating within players themselves. 'Reward' need not necessarily mean gold or items: it can refer to more nebulous things, like the satisfaction of learning or the accolades that follow skilful or imaginative play.

Dota 2 combines fixed and variable ratio schedules in the way it distributes information. Every time you play, you are steadily learning new things about heroes, items, combos, techniques and so on. The amount that the game asks you to learn is vast, but manageably so: there is always the sense that there is more to learn, but not so much that you can't play right away. For the WoW player, their experience is a bar that runs along the bottom of the screen, steadily growing; for the Dota player, it is a feeling—also steadily growing.

All competitive games have this aspect, but the Dota formula is distinctive because of the amount of creative space it provides. It is possible, at every level of play, to be the one exceptional mind that comes up with a silly combo, a play, a draft that wins a game and earns you a story to tell. These aren't high-level or professionally legitimate expressions of skill, but chance wins that make for good YouTube nonetheless. This is how a variable ratio reward schedule can manifest within a competitive context: not in terms of an epic item from a trash mob, but in the eureka moment that produces Blink Armlet Dagon Terrorblade, or the Five Man Bird Bomb.

No other competitive game, I would argue, provides this balance between steady progress and the chance for creative triumph. Only the very best StarCraft players will ever earn the right to define their own meta and win; Counter-Strike players are judged by extreme degrees of finesse within the binary framework of a gunfight. There is absolutely skill and beauty in both of these examples, but they lack the aspirational quality that comes with creative space—space created by Dota's lack of structural grace, by the sheer volume of rules and exceptions that make it so daunting on paper. Its complexity isn't just a ladder that you climb: it's a sandbox for you to make your name in.

These games are interesting because they fall far outside of the format that we'd associate with accessible or addictive games while operating, I'd argue, on many of the same principles. In offering an emotional payout in place of an ultimately-thin sense of progress, they suggest that it's possible for 'compulsive' to be divorced from 'unhealthy'; 'time-consuming' from 'a waste of time'. There an awful lot of people playing these games, and I find it heartening to think that the majority of them are getting something out of it beyond a few numbers on a character sheet.

You might not get why these games are as popular as they are, and you might be justified in sighing next time another major publisher announces that they've bought a studio you like and set them to work trying to beat LoL: but there is a message communicated by this popularity, and it's one that says better things about gamers than any essay on compulsive mechanics or player retention ever will. It's that true emotional payout is everything: deliver the feeling, earn the player.

To read more Three Lane Highway, click here. Thanks to Eskil Steenberg for the discussion that helped to flesh out a few parts of this article.

Each Friday the PC Gamer editorial team convenes at a secret mountain lodge to discuss the deeper ramifications of the week that was. The gl wein expenses are killing us.

THE HIGHS

Chris Thursten: Well strike me down It finally happened: I found another lane-pushing game that I like. The Smite regionals in Cologne are to blame: I went in a novice and came out wanting to give the game a try, and now my lunch breaks are consumed by Arena. This is well-timed, really: 6.82-era Dota 2 doesn t reliably fit into 45 minutes any more, whereas I can play a couple of games in one of Smite s lighter modes in that time. I m still not quite feeling brave enough to head into full 5-on-5 play, but that s partly because I m really enjoying the 80% Arena winrate that my transferrable Dota 2 skills (such as they are) have granted me.

More to the point, I like the little things that make Smite less of a fraught experience than the game I m used to: the tactical necessity of retreating to base, the rhythm of pokes and ganks, the way ults operate as much as a make play button as a oh god guys we need to teamfight we re wasting time button. I m sure the game has those elements, but at the moment it feels like a holiday in a more straightforward land. There s a chance I d enjoy League of Legends for similar reasons, but I ll never see the point in committing to a game where your account level dictates anything about your potential power in-game. Smite earns a lot of good will by dodging that, and by offering an affordable pack ( 20/$30) that allows you to unlock every character, past present and future.

Shaun Prescott: A truly disturbing shooter I haven t actually played GAME OF THE YEAR: 420BLAZEIT vs. xxXilluminatiXxx [wow/10 #rekt edition], but I ve seen enough gameplay footage to know that it s a work of art. Created by Melbourne developer Andy Sum for the recent Seven Day First-Person Shooter Challenge, watching GOTY is like listening to Lou Reed s notoriously unlistenable Metal Machine Music LP. It s a distillation of everything ugly and garish about modern blockbuster gaming: the unmitigated bloodlust, the corporate synergising, the flagrant macho stupidity of it all. While the parodic game is ostensibly funny , it s also weirdly disturbing: this is a part of our culture that seems fairly commonplace, until it s paraded before us in such a concentrated fashion.

Phil Savage: The unbearable lightness of being I recently completed a big, sprawling RPG—one that took up a solid week-and-a-bit of my life. Afterwards, I was in need of something smaller, lighter and altogether more silly. The answer, it seems, was bears.

At this point, I think I've played more of Far Cry 4's map editor than I have of Far Cry 4. It's great. More importantly, it's dumb. I'm not sure it's meant to be dumb, but the way I've been using it is. With a few clicks you can set up a ridiculous scenario, and then head into play mode and watch it play out. For the most part, I've been stacking animals. Or blowing up animals. Or blowing up stacked animals. One day, I might make a real map. More likely, I'll try to jump over a pyramid of bears with a tuk-tuk.

Tim Clark: More bear love I suppose it s uncool to admit that I m enjoying what s essentially a clever marketing campaign, but the truth is the Good Ship Cool long since sailed for me, and I m straight up loving the daily drip feed of cards from Hearthstone s Goblins vs Gnomes expansion. We got to reveal one ourselves earlier today—Hail, Iron Sensei!—and you can read my ramblings about all the recent cards here. So far easily my favourite is the Druid s Anodized Robo Cub. Not the flashiest creature in the game by any stretch of the imagination, but a wonderfully flexible early drop for a class that badly lacked exactly that. Also, the art is adorable. And if you re not choosing at least a couple of cards based on irrational love of the artwork, you re Hearthstoning wrong. 

Tyler Wilde: Goat MMO Simulator Bears are fine, but goats are where it s at this week. A new free expansion turns Goat Simulator into a (fake) MMO, where you can quest to pick up apples and infiltrate a sheep village, but mostly just headbutt people into ragdolls. I m still with Andy on Goat Simulator—it s silly and all, but I haven t found much actual entertainment out of it—but this update was too clever not to try. For one thing, you can play as a walking microwave. Just like I always dreamed. It also does a great job of simulating MMO chat convincingly enough that, for a minute, I questioned whether it was fake. Even better, it inspired a very stupid article, so I have it to thank for that.

Tom Senior: Dying a thousand deaths I have bled to death in an abandoned apartment complex. I have died of severe hypothermia in a field. I've been hit in the leg so hard I've crumpled to the ground and been mauled to death by a desperate stranger. You too can experience the panic and misery of the post apocalypse in NEO Scavenger.

This low-fi survival game only lets you carry what you could feasibly carry in two normal-sized human hands, which forces painful decisions between whether you ought to keep a shard of glass for defence, or a blanket to fend off the elements. A bag, shoes, a roof over your head—these are as gold dust in the cold plains of future Earth, and if you survive long enough to find them, you've got a shot of unraveling the mystery behind the planet's devastated state. There's horror out there in them hills, from cannibals to killer robots. It's rather good so far.

THE LOWS

Chris Thursten: Phantom controversy The tendency for hardcore games communities to transmute passion into disappointment and rage is a source of constant bemusement, for me. It sometimes feels like you re not allowed to love something unless you re also convinced that it could be ruined at any moment—that every change, or absence of change, is a disaster. This attitude gives developers no room to move, and can turn exciting reveals sour in moments.

Valve released a new hero for Dota 2 this week, Oracle, along with an event associated with a premium item for Phantom Assassin. This is the first event to substantially interact with the playing of regular matches, and as such its implementation warrants some scrutiny. But that s not what I d call the reaction on Wednesday, when a leaked list of the event s features (many of them false) caused much wailing and gnashing of teeth. The community set to building THE END IS NIGH signs over a set of assumptions that didn t even turn out to be true: such a waste of energy.

Today s the first day of the Foreseer s Contract update and people seem to be having fun with the new systems. There are even reports that the event is encouraging greater cooperation between players. Valve s tinkering might just have yielded something fun and innovative, but it was still deemed a disaster before the truth was even out.

Tom Senior: Mob-o-geddon I m in a positive mood this week, so my low is secretly a second high (shhh, don t tell anyone), in that I ll use it to express frustration that I m not playing Lost Ark at this precise moment in time. I m a big fan of the ARPG genre, but as much as I enjoyed Path of Exile s cerebral charms, I ve always wanted someone to push the genre to new heights of lunacy. Lost Ark lets you sail around the world fighting huge ghost ships. One of the classes lets you shoot huge dragons at mobs, or summon god-sized elemental beings to stomp on enemies. Watch the trailer and just look at how big those mobs are. Look at how pretty it all is. Damnit, hurry up and get on my PC.

Tim Clark: Crying foul  Continuing with Tom s definition-bending theme, my low eventually became a high. We took some flak behind the scenes this week from Ubisoft, who weren t pleased that we d referenced the recent buggy launch of Assassin s Creed: Unity when we explained why our Far Cry 4 review was late, and that in the interim buyers should probably exercise caution until the state of the PC code could be confirmed. I mean, god forbid a company s blockbuster game should be mentioned in the same breath as another blockbuster a game it released a few days previously.

In the end I think our scepticism was justified, given that the publisher had to create a live updates site to deal with the brace of patches needed on release. For the first 24 hours I couldn t get the game past the menu screen without it crashing, on a PC that ought to have been capable of handling it on Ultra without breaking a sweat. Andy had a happier time, as his review in progress notes. The full verdict will be up early next week. Here s where things cheer up for me too: With both patches applied and a new driver installed, Far Cry 4 is now running fine. Better than that, in fact, it looks absolutely sensational—as this video which the other Tom made shows. I can t wait to make the inhabitants of Kryat s acquaintance. A weekend of extreme taxidermy awaits.

Shaun Prescott: Official VR support for GTA 5 seems unlikely With the arrival of GTA 5 s first-person mode on new generation consoles, and its release for PC in January, it seemed inevitable that VR would come into the equation somehow. Indeed, it seems like the only valid reason a studio would bother retrofitting a whole new perspective into an already immensely popular game. Alas, comments this week by Take-Two CEO Strauss Zelnick seem to indicate that VR won t be officially supported (Take-Two is Rockstar s parent company). Zelnick thinks the industry isn t ready for VR yet, and while there s little doubt that enterprising modders will get it going, it s still a little disheartening.

Phil Savage: Expecting the Inquisition Last week, we successfully landed a probe on a comet. In space! This is a thing that people envisioned and built and made happen. Given this—and, while we're at it, given the entire scope and wonderment of human achievement—how come we still can't release a video game on the same day worldwide?

Dragon Age: Inquisition has only today been released in the UK. The USA and others have had it since Tuesday. A few extra days might not seem like much, but it's a story based game and the internet exists. People are rightly precious about spoilers, because discovering the story is part of the pleasure. And yet, the risks are now out there. YouTube is filled with cutscenes of late-game missions, wikis are being updated with newly learned fates, and pricks are simply being pricks. We don't communicate across national lines any more, but our media is still held back based on arcane contractual tradition. That these restrictions are so easy to circumvent reveals them for what they are: artificial restraints with no justification or benefit.

Tyler Wilde: I want to walk in Dragon Age: Inquisition I m really enjoying Dragon Age: Inquisition, but one thing is bugging me (well, aside from the stuttering, but I m hoping the latest Nvidia drivers solve that.) I can t walk. This is a big deal. Seriously.

With a controller, a light touch on the analog stick triggers the walking animation. But I am not using a controller, and there s no key to toggle walking. It s driving me crazy. I spent an hour making my character just right (she looks a bit like Mireille Enos in The Killing), and now she bolts around everywhere like an idiot. Mireille Enos doesn t run everywhere! She walks purposefully. If she s going to have a conversation with someone, she doesn t sprint into their faces, bashing nose against nose. YOU WANTED TO TALK? No. She walks. Calmly. Like a human being.

Thank goodness I m not the only one asking for walking, and Creative Director Mike Laidlaw is on the case. Unfortunately, BioWare sort of needs to make sure the game is working for everybody before addressing a design issue like this. I guess that s more important. But I can t wait for the walking patch.

Three Lane Highway is Chris' weekly column about Dota 2. Well, mostly about Dota 2. Image via ESL's Flickr account, credit Helena Kristiansson.

I'm going to try something a little different this week. It might be alarming, and I don't want you to panic. This week's column is not going to be entirely about Dota 2. It's going to be about a related game, Smite, and a recent experience that has helped express something I've been considering about competitive games in general. A lot of the things I've learned and written about while playing Dota could apply to Smite, League of Legends, Counter-Strike, or any other team game: this is simply the furthest I've ever taken that notion.

The Smite pro scene is currently undertaking a run of regional championships to determine which eight teams will earn a place at the $1.3m World Championship in January. Last weekend I attended the European finals in Cologne, a modest event by modern Dota standards but a reminder of how exciting it can be when an e-sport is still finding its feet. The Smite scene is young, and it lacks the history, orthodoxy, and to an extent the politics of professional Dota 2 or League. You should be watching Smite right now because this lack of precedent means that upsets are more likely, bringing the personalities of individual teams and players to the fore.

(It also means that there are a load of UK players in the scene, which was such a novelty I'm still not sure I believe it. I'm so used to encountering my countrymen exclusively as casters and personalities that meeting a British pro player elicits genuine delight and surprise, like seeing a cool dog in the pub.)

Nobody has successfully codified what makes a professional gaming team effective—or an amateur team, for that matter. There's a lot of chaos in the system, and few formalised practices regarding coaching, management, or training. Underprepared teams sometimes win through superlative skill; prepared teams sometimes win through collective discipline. Sometimes there's a little of both, and sometimes there's no pattern at all. I followed Alliance because of the attitude expressed by Loda during this interview; I also understand why old-Na'Vi and their particular swagger drew the approval it did (and does).

E-sports are, I think, a few years away from approaching the kind of stability encountered in traditional sport—and I don't think there's anything to be done to speed up the process of getting there. The only power we do have, as fans and players and managers, is to decide what we're going to focus on: what approaches to play that we'll celebrate, which attitudes we'll try to imitate. It's in this way that it's possible to exert a degree over control over the tone of a game, whether that's just for you or your friends or for the community as a whole.

There is, I think, a natural but problematic imbalance in the kinds of top-level plays that garner attention. Specifically, it tends to be solo players and virtuoso moments that hold the most sway over the minds of the community. It's easier to build narratives around individuals than teams, and it's easier to imagine yourself as the game-winning lone wolf than it is to see yourself as the guy who really knows how to ward. This is the essential root of every drafting disaster that has ever happened to you in ranked matchmaking. It's why people seek other individuals to blame to preserve their own ego. It's why the community flips out about solo MMR in what is fundamentally a team game.

The reason I want to celebrate teams that win the other way is because it sends a message to the community: things do not need to be the way they are. When a team's collective strength is their greatest attribute then it suggests a better way for players to interact with one another. The best recent example of this that I can think of took place in Cologne, in Smite, when Aquila (a Cognitive Gaming team) stormed the regionals through determination, preparation, and teamwork.

Aquila are more or less the dream, as far as e-sports success stories go. They missed out on inclusion in the Smite Pro League by a single game and had to fight their way into the regional championships through the Challenger Cup, also known as the amateur league. They had their own coach and analyst before being picked up by Cognitive but are otherwise fresh to the circuit—Cologne was their first LAN. While certainly underdogs, there was nothing scrappy about their approach to the game.

"I don't think any other teams prepare in quite the same way as we do" team captain Nate 'Ataraxia' Mark told me after their first match. "We're much more formal."

And what a first match. Notable chiefly for featuring the most profound upset in competitive Smite, Aquila vs. Cloud9 turned on a dime when the latter refused to press their advantage in the second game to close out the set 2-0. C9 were maybe two or three auto-attacks from ending the match when they backed off, and while the obvious narrative here is that they threw the game, they were able to throw it because Aquila kept their shit together in the face of first-round elimination.

"No-one was saying anything bad" Ataraxia said. "Everybody kept positive, which is fantastic—especially [support player] KanyeLife. He never gave up. He was always upbeat, always positive."

That ability to keep their heads in the game ultimately turned that game around, and the momentum from the win carried them through the decider. Then they defeated Fnatic 3-2, claiming a spot at the World Championships; finally they beat SK Gaming 2-0, earning the European title. Both of those victories were at the expense of teams that prioritised their star players—Xaliea for Fnatic, Realzx (and increasingly maniaKK) for SK.

Aquila's playstyle stood out to me because, unique among the Smite teams I watched, they valued objectives and efficient rotations over raw individual fighting power—something that the game emphasises given its focus on action. They are extremely disciplined (a little like 6.79-era Alliance or recent Virtus Pro, Dota fans) and I wasn't surprised to hear that they're known for how seriously they take match preparation.

"I'm kind of strict on the guys" said Aquila coach Job 'CaptCoach' Hilbers. "Around some of the other teams, there's not really a strict format. I set up a training schedule, make sure that everybody's on time and that they get everything they need. I make sure that they can fully focus on the game."

There's this notion that creeps around the competitive gaming scene that preparing beyond a certain point is trying too hard: that it's better to win without the effort, if you can, and that strat spreadsheets should be the preserve of teams whose shirts are already covered in sponsor logos, rather than the ones that want to get there. CaptCoach sees it differently: preparation is something that comes naturally as a consequence of the team's hunger to win.

"We keep very close track of what other teams are doing" he says. "And something that a lot of teams forget is that you yourself also need to look at what you are doing. We analyse ourselves like we analyse others so we know what they're going to see—what our weak spots are. Then I want to eradicate those."

Self-knowledge is, I think, the difference between a good team and a potentially great one: it's the answer to ego, and ego is pretty much the root of all evil in a competitive context (the caveat here being that evil is fun.) "There are teams where you have star players that get pulled ahead" CaptCoach says. "I try to keep everybody level, make sure that everybody's comfortable. That's my most important job."

I wish that 'comfort' was more highly regarded, as an attitude to competition. Comfortable players don't rage or look for somebody to blame: they don't demand particular heroes or roles or lose their minds when told what to do. It is, inevitably, harder to identify heroes to worship when what makes a team special is the fact that they have no heroes; but it's also important. The community would do well to celebrate the successes of teams that genuinely and seriously treat each other as equals—no matter how cheesy those prematch fist-bumps can be.

To read more Three Lane Highway, click here.

Three Lane Highway is Chris' weekly column about Dota 2. Artwork is 'Tidehunter' by MikeAzevedo.

I'm playing through Dragon Age: Inquisition at the moment. It's great—probably my game of the year—and you should  read about that elsewhere. Its greatness isn't what I want to talk about right now. Yesterday, I made a decision in the game that actually broke my heart. I picked a dialogue choice from a list and something happened that made me so unhappy I had to take a break from the game. I thought about it, and talked to Phil, our News Editor, who has finished the campaign, and then I went back and unmade that decision. Normally, I would never do that. But this time I decided that I didn't want to be sad and so I loaded a save and picked the other option from the list.

That is how 'regret' most commonly manifests in your experience of playing a game. You go back and replay the section and alter your decisions, or play with more skill, and you bring about an outcome that you're happy with. Even in competitive games, where exact repeats are less likely, there are still scenarios that you can vow to approach differently: next time, they won't get the flag out through the back entrance. Next time, I won't fall for a Dark Templar rush.

Dota 2 is a little different, at least in my experience of it. The individual components that make up a given competitive scenario are so complicated that you never really get repeats: you might experience a base race more than once, but it's very unlikely that you'll ever experience the same base race again. Your hero, your friends' heroes, your enemies heroes; who's alive, who's dead, who has buyback; items, ultimates, cooldowns. All of these things matter enormously. They all factor in to the decisions you make, and when you make the wrong decision it can be as heartbreaking as watching a tragedy unfold in a singleplayer RPG.

Last weekend I played another round in the games industry amateur Dota 2 league,  The Rektreational. My team, Venomancer? I Hardly Know Her, doesn't really practice. Actually, there's no need for 'really'—we don't practice. I jump into games with Pyrion, Shane and Phill from time to time and play with Pip a little more regularly, but as a stack our only experience of playing together is these matches. This means that the first games are always a little rough.

Our first game against League of Legends (best team name in the tournament, there) was more than a little rough. And we should have won, but I made a big mistake. With their carries dead but our ancient exposed we charged down mid and went straight for the throne. As Tidehunter with a Refresher Orb, I was sitting on a double Ravage: I felt confident that we could push through the three remaining supports and take the game. But they weren't there.

LOL pulled an Alliance-in-TI3 and bought Boots of Travel, closing in on our ancient as we took down their tier four towers. I should have teleported back then and there, but I didn't have a scroll. Stupid. I hesitated, thinking there might still be defenders in their base, and waited slightly too long before running to their fountain to buy a town portal and jump back. I ran from the fountain, Blink Dagger on cooldown, to try to get their supports in range of Ravage. I was maybe 200 units away when our ancient exploded—and among those tumbling fragments of our game-critical glowy rock garden were little shards of my self-regard. I was gutted. And you can't go back and fix that.

I'll never encounter that scenario again, and I'll never get to unmake that mistake. It just happened that way. One of the things that can be so frustrating about learning Dota 2 is that you can learn to recognise the mistake but the process of implementing that knowledge takes years to bear any kind of fruit. Another example: last night, in a JoinDota League match, my team successfully ganked a way-too-farmed Spectre carrying an Aegis. As soon as she fell we got ready to fight again. A thought wormed through my mind: their Tidehunter probably has Ravage. Brewmaster probably has Primal Split. They've got time to get here. We're probably already--

dead.

Whose fault was that? I genuinely do not know. Perhaps we were only delaying the inevitable. Perhaps I should have said something, or said it more clearly. Maybe the other members of my team should have seen what I saw coming too, and overcome the natural human urge to cooperate enough to back off on their own. Trying to unpick that game-ending catastrophe is like trying to unpick any other act of irrecoverable bad luck and failure—the unshakeable minor grief of missing the bus to somewhere really important. Except with wizards.

You can vow to do things differently next time, but that's only realistic to a point. You've probably learned something, but you'll probably never know: next time will always be different, and the reasons for your success and failure then won't necessary have much to do with how much you do (or don't) learn now. 

Instead, the conclusion I've arrived at is this: that making a mistake is, in some ways, its own reward. The feeling of recognising a mistake, horrible though it might be, is the only sign you are going to get that you are improving at the game. You were not a good player in that moment, but you are aware of some of the reasons why. You can never guarantee that you'll fix those problems, so you have to learn to derive satisfaction from your awareness of them. There are no repeats. But there is always the ongoing linear progress of your understanding, a mass of experience that can only ever be grown, never diminished, by the things you regret.

I still wish I'd teleported earlier.

To read more Three Lane Highway, click here.

Photo courtesy of MSI Gaming's Facebook page.

Despite fewer matches being played, today saw some of the longest and most exciting games of the entire tournament. For that reason, I'm going to get straight into the match reports: with day one's tech issues largely resolved, this was a day spent watching the most promising teams in the event Virtus Pro, Rave, Immunity and CDEC resolve themselves into a final ranking order. The results were definitely not what I expected.

You can find yesterday's write-up here, and VODs will be available via JoinDota's YouTube channel in the future. You can also find them inside the game client.

The semi-finals

I wrote yesterday that letting Rave.Chrissy play Timbersaw was a mistake that few would make again after his performances against Wired Gaming and Virtus Pro. In the first game of their semi-final best of three against Rave, that's exactly the risk that Immunity took. They picked up Batrider and Skywrath Mage two heroes that they favour alongside dominant early-game support Ogre Magi, Ember Spirit, and the increasingly popular Necrophos. Rave had Vengeful Spirit (now the defacto Batrider counter) as well as Centaur, Witch Doctor, Timbersaw and Terrorblade.

After Rave secured first blood following an over-eager tower dive by Ember Spirit I thought Rave were in a good place to dominate the game, provided that they could capitalise on the pushing power of Terrorblade and Timbersaw. Immunity's teamfight nous proved too much for them, however, and a succession of favourable trades as well as outright victories forced Rave to commit four or five heroes whenever they wanted to achieve anything. After that, however, the game entered a stable phase with neither team willing to risk an all-in engagement. This was a fairly common occurrence today, suggesting that the current patch has an issue with inertia around the thirty minute mark.

Immunity slowly ceded map control to Rave, but Necrophos' buyback-disabling Reaper's Scythe is designed to insure against exactly this scenario. After pickoff kills on Vengeful Spirit and Witch Doctor left Rave open, a follow-up Scythe kill against Terrorblade gave Immunity the time they needed to end it.

Parallel to the escalating stakes of the set, Rave and Immunity were by this point embroiled in an ongoing competition to see who could yell 'NICE!' louder after almost anything of any note happened. The teams were sat two banks of desks apart, and what started as a bout of self-aware self-hype turned into a running gag that was a lot of fun to watch. Claiming their game-ending Terrorblade kill caused Immunity to collectively yell 'NIIIIIICE!' in a way that was faintly harmonised in the manner of an amateur barbershop quintet. Digital sports.

Rave earned a series of increasingly enthusiastic 'NICE'es in game two, running the Medusa that they like to draft whenever they feel like playing Dota for a really really long time. This time the plan was to create space for the Medusa to farm, a strategy that Rave are obviously comfortable with. Immunity ran Void, Dragon Knight and Queen of Pain but struggled to deal with a roaming Puck and Chrissy's Timbersaw, once again. In the end, a contested Roshan attempt turned into a disaster for Immunity when their initial defense fell apart, and one teamfight later their base was left wide open to that well-farmed Medusa.

Immunity allowed Rave to pick up Medusa again in game three, this time supported by Earthshaker and Magnus. They took their own lategame carry in the form of Spectre, however, and an extraordinary performance by sLiCKz saw that hero pick up over 750 gold per minute in an eighty-five minute game. Both teams wanted it to go late, and go late it did, but Rave were relying on perfect Reverse Polarities to maintain their control and this simply wasn't as reliable as Immunity's Centaur and Brewmaster. Right at the end it looked like Immunity might throw the game when they went a little too aggressive on the high ground, but as soon as they pulled their discipline back together they managed to close out the game, and the set, much as they'd done to Bravado a day earlier. Nice!

On the main stage, Virtus Pro faced Chinese favourites CDEC. VP let Death Prophet get through the first round of bans previously, she'd always been struck out by this point but answered with a novel combination of Sven and Mirana that did a tremendous amount of work in the early part of the game. As Virtus Pro tend to do, they dominated in the first ten minutes while CDEC struggled to find kills without trading losses in return. Similarly, VP's Roshan discipline was used to good effect to secure their advantage going into the midgame.

Then, slowly, the game shifted. During another extended passive period both teams made headway towards the others' high ground, but it was ultimately CDEC's Slark that had the durability and damage to actually take objectives. And whenever VP committed to chasing Slark around the map, Death Prophet would be left to bowl over towers with Exorcism which is why the hero is first-ban material. Then, as CDEC got ready to push for mega creeps, the game crashed for everybody. GG was called by VP right then, instead of waiting for a remake simply to confirm the inevitable.

CDEC picked Slark again in game two, this time with a Void/Skywrath Mage combo. VP's answer as would become a trend was to draft for durability, with Ogre Magi, Tidehunter and Necrophos resisting the kind of pickoffs that caused so many problems in the first game. They forced bloody exchanges even when CDEC should have had the advantage such as when Chronosphere first became available. The game remained close for a long time, but VP's talent for getting the most out of teamfights and using Roshan to their advantage gave them the momentum they needed to win.

Game three took the trends of the first two games and stretched them to extremes. Both teams drafted fighting lineups VP taking Slark this time, along with Viper and Tidehunter, CDEC taking Enchantress, Batrider, Lycan, Puck but this played to VP's advantage more than CDEC's. They had a fantastic, aggressive start but couldn't quite close out the game before the now-traditional midgame doldrums set in. In this case, however, the mid and late-game took the form of a series of brutal, inconclusive teamfights as both teams stacked up on BKBs and slowly wore them down over the course of an hour. In the end, dwindling BKB durations made double Ravages relevant once again and a pickoff kill on Puck provided the space that VP needed to finally end the stalemate. They took the set 2-1, and advanced to face Immunity in the final.

Third place match

I was expecting CDEC to face Rave for the title, not for third place. There was a slightly muted air to this best-of-three, which took place on the community stage while StarCraft II took over the main stage. Rave struggled to find form in the first game despite flanking their Medusa with pace-controlling heroes like Silencer, Sand King, and Timbersaw. CDEC used a midlane Troll Warlord to boost the already-substantial power of Faceless Void, ushering the team to a confident win. In the second game, Rave experimented with Naga Siren, Slardar and Batrider but any split-pushing and ganking they might have aspired to do was effectively curtailed by Ember Spirit, Nature's Prophet, and Vengeful Spirit. The set ended 2-0, with CDEC claiming third place in the tournament but you could tell that neither of day one's best teams expected to find themselves in this position when all was said and done.

The grand final

Virtus Pro clearly believe that this patch is about teamfights, and today that belief was proven to the tune of $30,000. They received fierce argument from Immunity, however, who attempted to build their victories around one-off killing power and hard-farming carries. In game one of the best-of-five, this gambit worked. VP allowed Immunity to pick up Batrider, Queen of Pain and Silencer, with Ogre Magi to secure the early game and Phantom Assassin to grow steadily more terrifying as the game wore on. A controlling combo of Puck, Vengeful Spirit, Tidehunter and Wraith King offered some resistance, but by the 35 minute mark Phantom Assassin was simply too powerful to ignore. VP made them work for it, but before too long sLiCKz was landing crits that'd make a grown man cry. Or an ice ghost explode, depending on whether you are Ancient Apparition or not.

Tidehunter was VP's first pick once again in game two, as Sedoy's offlane hero pool settled into Bulldoggian levels of consistency. This time they pulled in Viper, Slark, Sand King and Vengeful Spirit too, effectively sending the message that they'd rather not die at all if Immunity would be so kind. Immunity stuck to their Batrider and Queen of Pain, but they failed to shut down Slark with any degree of efficiency. Just as CDEC had done to VP earlier in the day, VP used Slark's mobility and carry potential to control the map, and ultimately the game. With the set drawn 1-1, it was time for Immunity to rethink their strategy.

Their initial picks in game three were representative of the tournament in general Ogre Magi, Void, Batrider, Mirana. VP responded with their own comfort zone Tidehunter, Necrophos, Vengeful Spirit, Ancient Apparition but delivered the first surprise pick in the form of Legion Commander. I loved this decision. She shuts down Batrider and synergises brilliantly with all of their other heroes, both as part of a full group and in pairs. Immunity were left to choose a mid, and I predicted something like Invoker to bolster their damage and control. They took, to my suprise, Windranger.

At first, it looked like a disastrous decision. That predictable ganking and teamfight power coming out of VP dominated the first half of the game, with Legion racking up a respectable amount of duel victories. Immunity had a lategame advantage on paper, but VP were in a position to deny them the opportunity to use it. Then, both VP and the game itself lost momentum. The teams entered yet another long period of trades, with VP generally claiming greater map control but being forced to back off after every engagement. Then, sLiCKz's Windranger entered her own dominant phase. During one base defense he landed four perfect Shackleshots in a row, and when Windranger utterly destroyed Legion Commander in a duel-gone-wrong (thank Daedalus and Aghanim's Scepter for that) the game suddenly looked winnable.

And it was. Windranger's extraordinary damage output posed a serious problem for VP, who had grown overconfident. Chronosphere became a way to set up targets for Windranger to destroy, a status quo that will probably cause pub Voids everywhere to faint from the scandal of it all. But when Immunity started to venture beyond the river, Necrophos happened. Losing Windranger to an upgraded Reaper's Scythe was all it took to place the game firmly back in VP's hands, and this time they didn't squander the opportunity to end it.

The beginning of game four made a fifth look inevitable. Against a strong Immunity line-up including Ember Spirit, Razor, Batrider, Skywrath and Sand King VP took Lycan, Enigma, Tidehunter (who'd have thought), Witch Doctor and Venomancer. Put it this way: I've seen Venomancer show up as part of a push strat in a BO5 grand final on two prior occasions and neither TI3-era Na'Vi or TI4-era ViCi seemed particularly happy about it afterwards. VP needed to win the early game convincingly and got off to just about the worst possible start by losing Enigma to a five-man Immunity rotation before the game even began. Ember Spirit, Razor and Batrider all got off to good starts, with Batrider managing a solo kill on VP's safelane Lycan that made Witch Doctor look like a neglectful parent.

Napping clusters of StarCraft II commentators sighed and anticipated the extra hour or so they'd have to wait before the SCII final. I started to wonder what had possessed VP to gamble this hard. Then, VP proved their point about teamfight lineups to the exclusion of all doubt. A Refresher emerged on Sedoy's Tidehunter depite his rough start, and well-placed Poison Novas and a string of perfect Black Holes gave Immunity no space to move.

'No space'. That was a space joke. Black Hole. Never mind.

MusiCa's Ember Spirit picked up a tremendous amount of farm two Daedaluses, a Battlefury, and so on but it wasn't enough. In the final engagement of the tournament, three successive Sleight of Fists yielded precisely zero critical hits as RNGesus turned his gaze away from the floundering team. And while teamfight ult after teamfight ult forced Immunity back into their fountain, a comfortably-farmed Lycan did what Lycan does when he gets inside your base. Immunity's ancient exploded and VP collected their giant cheque as the speakers blared the impressively literal 'We Are The Champions' over the cheering crowd.

Consistently, VP struck me as the most disciplined and experienced team in this tournament and that bore out in the end. Where others lost seconds arguing about their next moves, VP simply acted and they kept their cool as the long days wore on, even if that meant running similar drafts over and over. Of all the teams in contention at MSI Beat It, they are the closest to cracking the current meta: and they cracked it by pressing 'R' at just the right time, over and over and over.

To read more about Dota 2, check out Three Lane Highway.

Photo courtesy of MSI Gaming's Facebook page.

Large or small, top tier or bottom, every e-sports event begins the same way. At some point, the pyrotechnics and techno give way to the barely-manageable chaos of organising however many teams, however many matches, however many technical problems. As a journalist you can gauge how successfully an event is going by how quickly you bond with a stranger over the quest for working wi-fi.

Even by those standards, the first day of the MSI Beat It 2014 Grand Finals has been an experience. It started, honestly, a little rough internet troubles on the community stages where the majority of Dota 2 matches were played, a lack of stream provision when Dota 2 was absent from the main stage, and, as you'd expect, spotty wi-fi. But through ad-hoc improvisation by the JoinDota guys and a quick response to the technical issues by MSI the event came together, and it's been an interesting twelve hours of Dota.

That's down to the roster of competitors. Beat It's continental qualifiers provided room for entrants from the Middle East, Africa, and Australia who you'd never normally see on the world stage. Unforeseen problems with Empire's availability and Arrow's match-fixing scandal resulted in Europe and SEA being represented by Virtus Pro and Insidious Idol respectively, and Na'Vi US and CDEC both represent the promise rather than the current pinnacle of their respective regions. Na'Vi US additionally played with three stand-ins, adding another degree of chance to an already chaotic line-up.

The result was a lot of scrappy, speculative Dota with moments of extraordinary coordination and collective skill. I watched the majority of the matches from the community stages, overcoming my wi-fi troubles by standing directly behind the players themselves. You get a decent sense for the character of a team when you hear the things they yell at each other. Mostly, it's "nice!". Often, "back!". Sometimes, sad silence.

At the time of writing the VODs aren't available, but when they are you'll find them here. You can, however, find stats pages and Match IDs on Dotabuff.

The beginning of the group stages

I figured that Na'Vi US would have Virtus Pro where they wanted them when they drafted Phantom Assassin against Faceless Void, but it didn't quite work out that way for either Na'Vi US or Phantom Assassin. A strong aggressive trilane built around VP.633's Mirana kept Na'Vi US on the back foot, and while they gained some ground during the midgame particularly when they could take advantage of Ravage they couldn't fight their way through VP's variously slippery or durable core heroes (nor the 150ms pings and 30% packet loss that affected both teams for extended stretches of the game.) They might have turned it around had it been a best of three, but it wasn't, and this was the start that both VP (and Faceless Void) needed in their long tournament run to follow.

Void showed up again in CDEC's roster when they faced Bravado Emotion, the winners of the African qualifier. Facing the tournament favourites in a best-of-one in their first match was a rough deal for Bravado, and the result was a very convincing win for CDEC to the tune of 47 kills to 14.

It was a similar situation with APAC qualifier winners Rave and their first opponent, the Lebanese Wired Gaming. In a rare Void-free game, Wired picked up a bunch of strong cores Ember Spirit, Viper, Batrider but were utterly denied by an extraordinary Timbersaw performance by Rave.Chrissy. This was, for my money, unquestionably the best combination of player and hero at the event today.

Fear not! Faceless Void made a comforting reappearance with Immunity in their match against Insidious Idol, facing his sometime fling Invoker and 6.81 life partner Skywrath Mage. The game was dead even until around the 25 minute mark, when Immunity began to pull ahead thanks to the mobility and versatility of Weaver and Queen of Pain. Also, you know, Chronospheres. They finally broke base around the 56 minute mark, but went a little too HAM and lost four people at the cost of most of Insidious' buybacks. Their answer, which I utterly respect, was to wait for their respawns and go HAM again, which won them the game.

Loser's best-of-ones

The losers of the first round were given a single shot to fight their way back in, and this led to some of the day's most heartbreaking games. Chief among them was Wired Gaming vs. Na'Vi US, a game that I know Na'Vi expected to win without too much of a problem. This overconfidence left them underprepared for a strong Templar Assassin performance by WG.Cbs, and the game was even until long past the point where Na'Vi's draft (with the exception of Morphling) had peaked. Then, Na'Vi secured a pick-off kill on Cbs just after he'd spent his buyback gold on the last parts of a Butterfly, and that gave them the foothold they needed to force a conclusion. Wired came very, very close.

Bravado Emotion faced Insidious Idol with the full compliment of Void, Skywrath Mage and Invoker, along with the (more popular than I thought, apparently) right-click carry Queen of Pain build. The structured play encouraged by the Chronosphere combo seemed to suit them, and they found sufficient form to win the game. The change in their attitude was palpable of all the teams in the room, Bravado Emotion had (appropriately enough) the loudest and most emotive coaches. While watching the other concurrent matches I could more or less tell when Chronosphere was off cooldown by the cries of YES BOYS from across the room.

Winner's best-of-threes

Group A winners Virtus Pro and Rave were the first teams to face each other on the main stage, and VP quickly learned something that Wired had learned the hard way a few hours earlier: don't give Chrissy Timbersaw. Having lost game one as a consequence, VP came back in the second by focusing on the pickoff power of Nyx Assassin and (uhuh) Faceless Void. They couldn't quite pull off the Chain Frost-Chronosphere combo, however, calling their Lich draft into question and by comparison, Rave's Centaur-Ancient Apparition-Sand King combination was very strong, and may well have constituted the best plays in the tournament. This earned ryo's Centaur the advantage he needed to give zero shits about anything that VP could throw at him, and crucially bought space for Jeyo's Medusa to farm. Given that this was a game with a Medusa in it, it ran very long it takes time to attack-move through teamfights, man but the energy picked up right at the end, when VP looked like they might just crawl back an advantage with several barracks down. They couldn't quite pull it together, however. Rave took the second game, and the set, thanks to all that space they won for Medusa.

CDEC's first game against Immunity lasted less than 24 minutes, with CDEC pulling out perfect games on maybe's Brewmaster and garder's Skywrath Mage. This match also saw Vengeful Spirit crop up as an effective counter to Batrider, a pick that has become more and more common and suggests that maybe just maybe! - the bat might slip a few more places down the pick/ban list. I have to admit I missed CDEC's second win against Immunity because I stuck around to watch VP/Rave game two and was totally convinced that the game was about to end for about half an hour before it actually did. You know, because Medusa. Short version: Brewmaster is a very good hero, and CDEC like Rave took their second victory to secure a place in the semi-finals.

Loser's best-of-threes

The final games of the day were best-of-threes between the winners of the loser's best-of-ones and the losers of the winner's BO3s. This was, therefore, Na'Vi US' chance to get revenge on Virtus Pro and win a place in the semi-finals and for a while, it seemed like they might do it. The combination of Korok's Necrophos and Demon's Elder Titan stymied VP's aggression for long enough to secure strong momentum, and another powerful Void performance sealed the deal. In game two, VP answered with an aggressive Wraith King/Vengeful Spirit/Ancient Apparition trilane that successfully controlled Na'Vi's safelane Void, and while Na'Vi managed a few impressive pickoffs with the global combo of Korok's Zeus and Demon's Nature's Prophet they found themselves needing control, not nuking power. Na'Vi's play got shakier as the game continued, while VP found form and forced a third.

Vengeful Spirit and Void made reappearances for VP in the third game, along with a clowny-but-effective Magnus performance by VP.G that involved a lot of cancelled Reverse Polarities for some reason. VP also continued their use of Gyrocopter as a complement to an offlane Void, and while Na'Vi US were able to find trades throughout the game thanks to Ember Spirit and Doom it wasn't ultimately enough to stop a slow shift of power to VP. In the tournament as a whole, VP have shown the most discipline when it comes to securing Roshan whenever he spawns and making full use of the momentum their teamfight victories grant them. Where other teams seemed to rely on individual moments of flare to find an advantage, VP's map control came almost entirely from constant, sustainable pressure ironically, exactly what Na'Vi used against them in the first game. By the third game the (constant, sustainable) shoe was on the other foot, and VP reclaimed their spot at the winner's table.

What was striking about Bravado Emotion vs. Immunity was how similar the games were, structurally, to the Na'Vi/VP match that ran concurrent with it. In both sets, the first game ran close to an hour and was won by the team most able to control the pace of the game in this case, Bravado, with their Brewmaster and Timbersaw creating the space for Anti-Mage to farm. Likewise, the second game was short thanks to a return salvo of aggression as Immunity drafted for the early game Windrunner, Witch Doctor, Ogre Magi, Mirana, Centaur and secured a kill advantage that Bravado just couldn't come back from. By the third game, the last of the night, Bravado seemed exhausted. They picked up the Brewmaster again, but didn't take advantage of the momentum that comes with using Primal Split to threaten the enemy around the map. Immunity claimed an early kill and tower advantage and Bravado were left defending against mega creeps for an agonisingly long stretch right at the end of the game. Facing elimination they refused to GG, successfully repelling attack after attack while Immunity were happy to sit outside, let creeps do the work, and take Roshan when it was available. Bravado even managed to claim a kill advantage by the end - but they couldn't claim map control to match it, and their end was slow and hard to watch. They should, nonetheless, be proud of how far they got as should Immunity, for staying in control after a very trying day of play.

Virtus Pro, CDEC, Rave and Immunity face off in the semi-finals tomorrow.

Microtransactions. "Micro." As in, very small bits of money—a dollar here, a dollar there—exchanged for comparably small in-game items: A new hat, perhaps, or some healing potions. But boy, it sure adds up. Riot Games earned $624 million from League of Legends last year, and that was only good enough for second place on the top-ten free-to-play earnings list. In 2014, however, it's shot to number one, and is poised to be the first to break $1 billion in microtransaction revenues.

League of Legends pulled in an estimated $964 million between January and September of this year, according to SuperData research (via VentureBeat), making the magical $1 billion mark almost a sure thing. And it's not alone: Last year's first-place finisher, the military FPS CrossFire, is at $897 million, Nexon's Dungeon Fighter Online is sitting at $891 million, and World of Warcraft has brought in $728 million. All them have a good shot at making it over the hump.

Also noteworthy is just how badly League of Legends is trouncing the other big free-to-play games in North American, Dota 2 and Hearthstone. Both made the list, but way, way back in the pack: Dota 2 is at number nine with $136 million in revenues, while Hearthstone brought up the rear with $114 million. I'm not going to shed too many tears for either of them, but it's a remarkable gap between them.

The World of Warcraft situation is interesting as well. Subscription numbers have slid dramatically over the past several years (although they recently enjoyed a bounce thanks to the launch of Warlords of Draenor), but the spending on in-game items, which is all these numbers take into account—no subscription revenues, in other words—is way up: Its nine-month total is more than triple the $213 million it earned in all of 2013.

In a similar light, Hearthstone may be doing better than it appears, because the figures are PC-only and thus don't account for mobile spending. That still wouldn't be enough to get it near LoL, but it might move it up a few places in the list.

Three Lane Highway is Chris' weekly column about Dota 2.

I'm going to a Dota 2-themed Halloween party this weekend. Yes, that's a little early—and yes, that is the dorkiest thing you've ever heard. Whatever. I'm too busy trying to figure out a Dark Seer costume that isn't going to make me look like Papa Smurf to care. This week I'm going to list some costume ideas that you might want to try out at your own Halloween parties—or, indeed, any other party you attend over the course of the year.

There's more to a good costume than simply what you wear. A good costume comes with an attitude, a persona, a performance. A good costume convinces everybody else in the room that you are the new and exciting image you present. These costumes are geared to immerse you in each hero's distinctive personality and fascinating lore. These are identities for you to adopt, armour to gird you as you march down the midlane of any given social engagement. Enjoy.

Also, maybe don't do any of this.

Dazzle

You will need: Facepaint; a dream-catcher on a stick; a shitload of feathers; a pink hoola-hoop.

You might not be the host, but you're here to help. It's your job to ensure that the party runs smoothly and that everybody has a good time, and that starts by topping up drinks and looking after anybody who doesn't seem to fit in. Use your own deft touch to hem awkward partygoers into the event's social weave: weaken discomfort with a smile, and when the lights go down watch the edges of the dance floor for anybody who looks excluded.

In particular, you are looking for anyone on the verge of regretting that fifth/sixth/seventh cocktail. When you see somebody teetering on the edge of shame, immediately toss your pink hoola-hoop over them and make them dance. The activity will make them more alert and temporarily hold off the inevitable tears, fists or vomit. When the hoola-hoop reaches the ground, however, there's very little you can do for them—and gyrating like a lunatic may very well have made it worse. But you tried, and are therefore blameless.

Terrorblade

You will need: An attitude; a big swishy silvery cape; big pointy silver horns.

You are the single most fabulous and important person in this room. The sole function of this occasion—indeed, of any social event—is to furnish you with a good time. Ensure that you are the biggest and loudest voice, and spend every quiet moment hoovering up all of the snacks and freebies to be had. Consume so comprehensively that your host does not believe that a single party-goer could possibly be responsible. They don't understand. You require all of the munchies, because you are very important.

There will be times when you find yourself running low on drink at a crucial moment. You might be telling an excellent story about how great you are, or single-handedly diffusing a situation while everybody else looks on, impressed. Should you run dry at this point it is perfectly acceptable to swap your near-empty glass for the full glass of a nearby friend. That is, after all, what they are there for.

Phantom Assassin

You will need: eye makeup, a dark hoodie.

Too shy to go full Terrorblade? No problem. As Phantom Assassin, nobody will care that you're there and nobody will remember you when you sneak away. This is where your power comes from. When they are not looking, you will eat all of the snacks. You will swipe bonus drinks. While they talk and have a good time, you will grow more confident, more powerful.

Think of the single best one-liner you can come up with. The most critical, cutting barb; the conversation-ending anecdote to end all anecdotes. Work on it until it's perfect, then invite yourself to a conversation and—boom. They won't forget to look for you after that.

Lone Druid

You will need: A beard; a full-sized bear outfit; a dog; a bear outfit for the dog.

Bring your dog to the party and spend all of your time petting and feeding it. If somebody offers you a drink or something to eat, demand that your dog gets the same thing. Later, do all of your socialising through your dog. If somebody tries to talk to you directly, point out something interesting or amusing about your dog. If they then attempt to disengage from conversation, send your dog after them.

Alternatively: Chen

A variant on the Lone Druid costume, the Chen costume shares many of the same principles. Instead of bringing your own dog, however, you will spend the first few hours of the party in the host's garden trying to turn their own pets against them.

Rubick

You will need: Somebody else's Riddler costume (stolen); a stick with a green rag on the end; a green mask.

One for the social butterflies. Don't feel confined to a single group or aspect of the party. Move freely and attempt to ingratiate yourself with everybody. Become the soul of the event: offer kind words where they're needed, diffuse aggression, endorse good times.

Then, when you are settled, start to steal other people's jokes. Hang out in a group until somebody tells a funny story, then disengage, walk across the room, and tell the exact same story there. You'll get away with this for a surprisingly long time.

When confronted about your rampant patter theft, insist that you do not steal—only borrow. The final and abiding irony here is that the writers of Dota 2 lifted this line from Picasso.

Silencer

You will need: A purple bishop's hat; a dressing gown; a silver frisbee; no scruples.

At the beginning of the evening, focus on approaching groups of partygoers and subtly making them feel uncomfortable. Don't say anything so objectionable that conversation stops, but make sure that everybody feels bad. For example: if you overhear a friend complementing the host on their cooking, point out that many, many people—far more than you will ever meet—are vastly less fortunate, and that many of them suffer even now.

Then, up your game. Actively seek out and create stifling social encounters. Examples include:

"Hey, I thought you guys broke up?"

"Good call, not spending much time on your costume—hers took hours."

And so on.

The party will grow quieter and fracture into isolated groups as people retreat into their comfort zones. Conversation will be muted, tentative. This is when you must bring the evening to a close: stand in the centre of the room, draw a deep breath, and clearly and loudly say something utterly and unforgiveably offensive.

Well done! You have achieved global silence. Nobody likes you.

Diretide

Don't respond to your party invitation. Later, explain that you are busy and that you might—might—make it along at New Years.

To read more Three Lane Highway, click here.

Three Lane Highway is Chris' weekly column about Dota 2.

Watching ESL One New York last weekend made me realise just how long ago The International was. I mean, sure, technically it's only been three months. You could say that, if you were going to be that guy. The guy with a calendar. In Dota time, though, three months is forever. This was one of the first major tournaments of the 6.82b (now 6.82c) era, and the first high-profile LAN in the west since the team shuffle process began after TI4. That great switcheroo is still ongoing, but for me this was the first time the game has felt anything close to stable in its new identity.

It was like watching a different Dota. Not simply because of the patch or the players but the effect that the two have in combination. Dota feels fresh again, a welcome sensation after the deathball doldrums that set in during the latter months of 6.81. This is in spite of a spate of embarrassing technical issues that riddled the tournament with interruptions. Maybe they helped, in fact: you've got plenty of time to ponder the state of the game when it's paused for two minutes out of every twenty.

Spoilers follow, naturally.

The long war

The best sets of the series Alliance vs. EG in the quarter finals, EG vs. Vici in the grand final were defined by games that the casters simply couldn't call until deep into the midgame. This was extraordinarily rare in the previous patch, and demonstrates that the update has been successful in holding off the '8 minute GG' issue (at least for the time being). Longer, closer games mean that greater emphasis is placed on super-lategame carries, as demonstrated by Loda in Alliance vs. EG game two. Vici's Black^ said afterwards that this patch was going to be about three-core strategies with a single farming hard carry, and TobiWan put it a different way when he described Vici's winning strategy as "four-create-space-for-one" (a twist on "four-protect-one").

What was striking about this change for me, however, was that matches haven't become more passive as they've gotten longer. In some cases the early game is less active than it used to be it certainly doesn't outright the determine the rest of the game as it sometimes used to but we're seeing more brilliant, close teamfights in the midgame as teams wrestle to claim map control. There's something to be said for game-determining teamfights and they still happen but I appreciate the number of close trades we saw, particularly in the final. It's funny that the solution to a game that was losing some of its drama was to effectively curtail the long-term signficance of certain engagements, but it really works for me. More fighting is better fighting, after all.

The age of the support

This general shift has interesting implications for support players. I've seen the argument made that 6.82's lategame focus makes supports less relevant as fewer matches are decided by early rotations, but I don't think that's quite right. A combination of interesting Aghanim's Scepter upgrades and more gold going to supports in general means more interesting plays coming out of position four and five later in the game. Vici Gaming more or less won the tournament on this basis, whether that's fy's Skywrath Mage saving Super's Ember Spirit with an Ethereal Blade or the same player picking up Diffusal Blade on Sand King in two separate matches to counter Omniknight's Repel. I'd probably argue that Vici Gaming won the tournament because of fy in general, but there you go.

It's great to see lategame supports being given more to do than 'get your ult off' and 'don't die'. There seems to have been an increase in the skill ceiling of the role in general, and I love that many of the biggest plays are coming out of heroes with no gold in their pockets. It will, inevitably, take a long time for this line of thinking to filter down into pub play if it ever does. But with the rise of Omniknight as a ranked matchmaking 'win' button and Jakiro's steady shift from support to core, there's a rethink coming that should hopefully impact how people perceive these characters and their importance to the game.

Despite this, many people will still first-pick Faceless Void.

Size isn't everything

In terms of the game itself, then, ESL One NY convinced me that the next couple of months are going to be really exciting. The heroes that have clung to their top-tier positions from one patch to the next Brewmaster in particular are generally the ones that are the most entertaining to watch, and I'm happy to see Razor and Death Prophet (slowly) make their way back into the box of spares.

On the other hand, the tournament highlighted a few things about the growth of the sport in general that aren't entirely positive and I'm not just talking about the technical issues. The choice of venue is a sticking point for me. E-sports have always chased the old-media legitimacy that comes with being on TV or occupying a famous venue, but I'm not convinced that 'old-media legitimacy' is something that anybody actually wants. Going for Madison Square garden might have got Dota into the New York Times, but is that the real measure for success here? When high ticket prices and limited spaces are the consequence of a high-profile venue, I'm not sure it's the best thing for the game.

As Dota grows I hope organisations like ESL see the importance of balancing prestige venues with locations that support the volume of people who want to see Dota live and who want to do so affordably. The Commerzbank Arena where ESL One Frankfurt was held was actually a good balance of the two, in that regard. It's possible to have both, but I believe that the entertainment value of Dota is substantial enough that it doesn't need the extra flare all of the time. It's natural for a new form of entertainment, particularly such a generational one, to look to adopt the trappings of the things that preceded it. Traditional sport established our model for what success looks like. But because this is a new and generational medium it is an opportunity to challenge a few paradigms and high ticket prices are one of them.

Which is not to say that Dota hasn't earned its place in Madison Square Garden, nor does it undermine the effort expended to get it there (and to make everything work, most of the time.) There is a real opportunity here, however, to serve the Dota community with events that lots of people can get to to the point where live Dota is a natural part of the hobby, and not an exceptional extravagance available only to the fortunate.

To read more Three Lane Highway, click here.

Watch live video from esltv_dota on www.twitch.tv

The ESL One Dota 2 championship tournament in New York City is now live, with eight teams vying for the trophy and a prize pool worth more than $100,000.

The first-round match pitting Natus Vincere against Vici Gaming is already underway, to be followed by Team Secret versus Natus Vincere US, Alliance against Evil Geniuses and Cloud 9 against Sneaky Nyx Assassins. The tournament is following a single-elimination knockout format, with each match, including the finals, a best-of-three contest.

The ESL One tournament is the biggest Dota 2 event in the West since The International 2014. As broken down by Joindota.com, there have also been a lot of roster changes since that event.

To find out more about the ESL One New York City 2014 Dota 2 championship, hit up els-one.com. You can also follow the action on Twitch.