Three Lane Highway: the victory switch, and other stories

Three Lane Highway is Chris' weekly column about Dota 2. And Smite. It's about Smite now too, apparently. The amazing fan art above is by Reddit user Pitran.

I've played Dota 2 almost exclusively since July 2012. For a long time it was the only game of its type that I played, and I've spent an order of magnitude more time with it than any other game of, well, any other type. I wouldn't be surprised if the time I've spent learning to wizard exceeds the time I've invested in games generally over the last two years. I held, for a long time, that you couldn't play more than one of these games seriously. I still believe that. Over the last few weeks, however, I've made a concerted effort to learn another—Smite. It's taught me a few things about the genre as a whole, and made me question a few further things that I held to be true about Dota 2.

Here's one new idea: surrender mechanics directly benefit support players. Back in July of last year, I wrote this article about why Dota 2 doesn't, shouldn't give its players a surrender button. I haven't entirely changed my mind about that. I still believe that the 'white flag' option makes these games less interesting overall. The Dota 2 experiences you remember are the late-game upsets, the incredible comebacks. Surrendering truncates the game, closes off possibilities, places hard limits on all of that fascinating complexity. In the abstract, I maintain that if a player is in a position where they must surrender then something has gone wrong with the game's provision of comeback mechanics. What I now realise, however, is that the decision to surrender is, in and of itself, a phenomena worth examining. The possibility of surrender creates new dynamics that alter the way you perceive the story of a match.

Dota 2's lack of a surrender option means that regular matchmaking games always end when one team destroys the other team's base. They can end no other way. It takes carries with good items and smart play by core heroes to do this, and the run of patches following last year's International have attempted to do away with ten-minute death pushes by giving defenders more options. Not only do games run longer, but the most important characters, in the end, are almost always the ones at the top of the farm priority pyramid. Earthshaker might start the ball rolling, but Faceless Void gets to kick it into the goal.

The same is true of Smite, to an extent. The role of the support, in both games, is to control the first half of the match so that it is your carries, not the other guy's, who ultimately succeed. This is where the 'Soccer Mom Crystal Maiden' meme comes from, and why support players are generally so rare—the role requires you to give up a substantial portion of your claim to glory. I've been playing support exclusively since I started to learn Smite because almost nobody volunteers to do it. As in Dota, everybody wants to play a solo roaming hero or carry. They want to make the big, game-ending plays—not the subtle supportive ones.

Teams can surrender in Smite, however, and this alters the prospects of what a support player can achieve. The goal stops being 'how do I ensure we have the best possible lategame' and becomes, in part, 'how do I break their spirits to the extent that there is no lategame'.

I'll give you an example. I've been playing a lot of Ares, a durable support who lacks burst damage but whose ultimate ability can completely turn a teamfight. The spell is called No Escape. Chains fly from Ares towards enemy players in a radius as he leaps into the air. After a few seconds he crashes down, dragging every player chained towards a central point and stunning them. Dota fans: imagine the lovechild of Magnus' Reverse Polarity and Disruptor's Glimpse. New Smite players tend not to buy the crowd control-breaking items that would get them out of dodge, so in these low-level brackets No Escape can act as a game-ending psychological weapon.

Case in point: my last game. The scoreboard is relatively even twelve minutes in. Both teams are almost entirely comprised of junglers and high-damage solo mages. As support Ares, I'm the exception. One of our guys disconnected at the beginning of the game and didn't come back for a few minutes, ceding an early gold and experience lead to the other team that we're only just clawing back. They've grouped up to push down middle lane. I tap two key combinations into the Tribes-style audio command system.

[VD2] Defend middle lane!

[VVVR] Ultimate is ready!

I approach the clustered enemy team from behind, from the jungle. The third-person perspective makes shooter-style sneak attacks a possibility. My blink is on cooldown, but I'm among the enemy team before they have time to do much about it. No Escape connects with all five. During the leap I draw them forwards, closer to our tower. They're dragged into a Chronos nuke; into that impassable ring thing that Odin does; into Loki, who presses a bunch of buttons I guess. (I'm still learning the gods.) Full teamwipe, a five-to-zero victory. They surrender immediately afterwards.

I wasn't the character who picked up the multi-kill, but I, the support, was the character who ended the game. I'd dealt the killing blow to morale in a way that I couldn't aspire to do to the enemy's base.

While I still don't think that a surrender mechanic is ultimately right for Dota, its presence in Smite has demonstrated the role it can play in redistributing power among the team. It allows for demonstrable displays of authority among 'subordinate' player roles, and creates scenarios where victory emerges from something other than a mounting lead in farm or experience. These kinds of psychological early wins play a huge role in Dota 2, of course, but I think the greater emphasis on the power of late-game carries makes them less visible to players who aren't specifically looking for them.

'Momentum' is a word that comes up a lot while discussing the way that teams win games of Dota, and I've written before about the way that this can be thought of both in terms of game mechanics and team psychology. Wins tend to beget more wins, because you've gained a material and emotional advantage. 'Snowballing'. Recently, I've been thinking about this slightly differently. I think there comes a point in the game where your team is in a position to decisively flip the 'victory switch', to turn an advantage into a done thing. This means more than just following the trajectory your momentum has laid out for you—it means identifying an exact methodology for ending the game and then pulling it off. It means closing off uncertainty and confirming victory; if a team's surrender represents a collective willingness to lose, then flipping the victory switch means collectively voting to win.

In that Ares game, the 'switch' could be defined as the moment we planned and achieved a one-sided teamfight victory. In a game where the majority of players on both teams had found themselves taking inconclusive trades in the jungle, a single convincing five-on-five was needed to establish dominance. In a sense, our opponents were right to surrender when they did: that fight in mid demonstrated superior capability stemming from a better-rounded draft, and it is reasonable to assume that we'd be able to repeat that success throughout the game and ultimately win. It was the beginning of the end and therefore, in some ways, the end itself.

Teams throw away their leads when they fail to make their advantage appear insurmountable. In Smite, the version of this I've seen most often is the single-lane death push. The key objective in the game is a Titan which, unlike the Ancient, can fight back against an attacking team. It loses power with every lane of buildings that you eliminate, but players on a roll typically attempt to punch through a single lane and win the game the most direct way they can see. This is often a really good sign if it happens to you, because it demonstrates that your opponent is willing to take risks—they are keeping the possibility space of the match open even as they attempt to end it, giving you options rather than decisively flipping the switch that takes your options away.

In a Smite match like that, that 'switch' might constitute the destruction of a second lane of towers, another Phoenix, or the Fire Giant. In Dota 2 it might be a faked-out split push that baits enough teleports to open up Roshan, followed by a jungle invasion that catches the smoke gank designed to counter the push your opponent believes is coming. These strategies are rarely seen in mid-level pub matches because they require teams to stop, assess what it would take to undo their own advantage, and then act decisively to reduce the chance of that happening to near-zero. It requires a desire to end, not just finish fast.

Learning to play a game with a surrender option has helped me to get better at identifying these moments, because it gives you unique insight into the mind of the enemy team. A surrender call tells you the exact point at which you have successfully drained hope from the equation: where even they agree that the victory switch has been flipped, and flipped by you. Over the course of a couple of weeks you learn the various shapes that moment can take.

That it sometimes takes the form of a play by the guy who buys all of the wards is a bonus, all things considered.

To read more Three Lane Highway, click here.