Aperture Science isn't the only one with a secret underground base. Deep in the Antarctic lies a hidden subterranean facility, filled with puzzles, lined with traps, and shrouded in mystery. Why is it there? What is its purpose?  Below the Ice, a mod for Half-Life 2, invites you to find out. Just watch your step.

The mod begins with you arriving in the Antarctic, where you quickly stumble upon the entrance to a facility buried in a glacier. There's a sign warning against trespassers, which feels a bit pointless. You've either traveled all the way to the north south pole and aren't going to turn around and go home because of a sign, or you're a polar bear penguin and you can't read anyway.

Entering this facility requires passing a bit of an intelligence test in a number of grid-like puzzle chambers. Many involve the simple pushing of buttons, though figuring out how to reach those buttons, and what those buttons actually do, can take a while. The chambers aren't particularly forgiving if you make a mistake, either. Prepare to be crushed, fried, or fall to your death if you slip up.

Once you've convinced the puzzle chambers you've got a brain in your head, you're granted access to the rest of the facility, which appears to have been abandoned. While exploring, you'll discover living quarters, science labs, and a series of offices. There, you'll begin to piece together the story behind the facility, which ties in to both the fiction of Half-Life 2 and Portal. It's not just a matter of walking around and reading notes: even though you've escaped the test chambers, there are still plenty of puzzles to solve to gain access to the facility's control rooms, observation chambers, and science labs.

The more you progress, the more the facility begins to reveal its secrets, and its true size. While you're navigating the place, unlocking doors, turning on lights, locating missing pieces of technology, dabbling in teleportation, and piecing together its history, also keep an eye out for a series of memory sticks. Find enough of them hidden throughout the mod and it will give you an alternate ending.

There's probably a few hours of play here, depending on your smarts. Some of the puzzles aren't particularly sophisticated: to progress, it's generally more important to carefully examine your surroundings for clues than to be some sort of 10th Level Puzzle Wizard. There's some decent music throughout, and while the map's set dressing is a bit plain, and a few custom textures are a little underwhelming, it's a nice mod if you're in the mood for some gently-paced puzzle solving.

You can grab the mod here and untangle the mystery for yourself. To install, extract the folder into your sourcemods directory (\Steam\steamapps\sourcemods), and restart Steam. Below the Ice will appear in your library. You'll also require Half-Life 2 and Source SDK Base 2013 single-player. You can find the latter by viewing your Games Library in your Steam client, selecting "Tools," and double-clicking it from the list.

The PC doesn't have a platform holder, and that's a blessing. But there's no denying that Valve's role is an important one in the industry, and that its decisions often have a tremendous impact on our hobby. With great popularity comes great responsibility: here's how we'd like to see Valve apply itself this year.

Make TF2 exciting again

Phil Savage

My favourite Valve thing of 2014 was the anticipation around the release of Love and War update. There was excitement, there was speculation, there was a funny Source Filmmaker video. It was everything a TF2 update used to be, with one exception: the update. There was a time when TF2 updates warranted the amount of excitement they received. The class events were significant. They dramatically changed the way you could play a class—giving new, divergent tactics to familiar operators. Also maps. TF2 launched with six maps. It now has over 70.

The Love and War update didn't have that game changing feeling... Sorry, I'll try that again. With the exception of the Conga taunt, the Love and War update didn't have that game changing feeling. The weapons were good, but I don't think weapons—even something as well-defined and desirable as the BASE Jumper—carry the same weight they once did. When there's only one way to play a class, a new item set has massive ramifications. Now there are multiple ways to play, and one more doesn't make as much difference.

In all, 2014 was a quiet year for TF2. Even the long-anticipated End of the Line community update was muted by Valve's decision not to include the Snowplow map. Their reason, supposedly, was that it was deemed too confusing and challenging for new players . And this is from the people who made Hydro. This is my worry with Valve now. It feels like they're afraid to take risks. The most exciting new inclusion of the year was the grappling hook, and that happened without fanfare as part of the Mann Co. Beta initiative.

Imagine if the same energy and excitement leading up to the Love and War update was rewarded by a new game mode, new maps and a goddamn grappling hook—all definitively released, rather than hiding away in beta. TF2 would, once again, feel like an event. Yes, maybe one of those maps would be terrible, but that would only make it more exciting when more maps appeared—these ones learning from past mistakes.

Aside from the Halloween event, every map added to TF2 in 2014 was released into beta. In 2015, I'd love to see the confidence come back. To see the TF2 team declare something ready—not just with a release, but with a week-long build-up that showed how a seven-year-old game could still feel fresh, exciting and essential.

Make the Steam Controller work

Tyler Wilde

It feels like ages ago that we were all talking about Steam Machines and the Steam Controller. It wasn t really that long ago, but Valve went pretty quiet on both fronts after a generating a lot of interest. I m still interested—not so much in Steam Machines, because if I stick a PC under my TV it ll probably be a laptop or something I build myself, but in the Steam Controller.

I m fascinated by what Valve s trying to do with the thing. Back when Evan tried it last year, he found the trackpads unwieldy, but if Valve can really design a new kind of controller that both emulates a mouse (and it ll never be as effective, of course) and works for games I d rather play with a d-pad or analog stick, then we may have a new best PC gaming controller in our hands. It s promising that, after we were a bit unimpressed by the first demos, new mockups keep appearing, which suggests that Valve is still tinkering with the whole design. I hope we see the latest prototype soon, and I m betting we will at this year s GDC.

Grow CS:GO

Evan Lahti

CS:GO became, with plenty of breathing room, the second-most played game on Steam in 2014, hitting 400,000 concurrent players for the first time this month. A lot of that growth is owed to CS:GO s reawoken popularity as an e-sport: more people than ever are watching competitive Counter-Strike, and the recent DreamHack Winter tournament (even with a bit of controversy) was an exciting watch.

But Valve s content updates, patches, and e-sports aid hasn t come close to the support given to its flagship game, Dota 2. CS:GO s support isn t proportional to its popularity, and Valve faces a playerbase that s hungry for anti-cheat fixes, new maps, weapon skins, and ways to engage with the game they re invested in if they want CS:GO to retain its position as the most popular FPS on PC against games like Evolve and Rainbow Six Siege.

Valve needs to continue to keep pace with hackers, and it needs to look to the old days of TF2 for ideas on launching in-game events that don t simply feel like money-grabs veiled in playfulness. But what would really propel CS:GO is a proper, The International-level major tournament—something that Valve owned and operated itself rather than relying on CS:GO s great-but-fragmented leagues to build interest around it as an e-sport.

Beyond that, I d love to see a CS:GO API opened up. Part of Counter-Strike is eliminating bad habits, and right now there s no easy way to track your match history (beyond watching your last few replays) or dig into meaningful stats.

A full year of Dota 2 updates

Chris Thursten

In 2015, I'd like to see the Dota 2 team continue to open up about their plans and processes. At the same time, I really hope that they're able to make this a bigger year for the game than last year. A lot of time has been invested in an engine update that will introduce custom game modes and make it easier for Valve to develop new additions in the future. That's great news - and user-generated content represents a potentially exponential increase in the game's scope. At the same time, I don't think I'm alone in wanting a stronger run of official updates. That means more new heroes in 2015 - potentially the first Dota 2-specific heroes - and proper seasonal events.

Valve may not see it the same way: they could argue, convincingly, that the future of Dota 2 was and should be in the hands of its community. Be that as it may, I think the last couple of years have shown that leaving everything up to the wisdom of crowds creates confused expectations and entrenches divisions. From the pro scene to the potato bracket, the Dota 2 community could use a bit of stability - and a solid year of official updates from Valve could help establish that.

Steam would benefit from a visual overhaul

Andy Kelly

Steam s feature set took a big step up with the addition of curators and, er, the colour blue in the client s basic skin in 2014. I would like to see a bit of housecleaning on the design of it—maybe contemporise the fonts a little bit. Reskinning is obviously an option, but I d like Valve s basic layout to be little more up to date. It s still very similar to when I first signed up years ago. On the one hand, that familiarity is nice, but on the other, Big Picture Mode and the Steam app on iOS show how much cleaner the basic layout of Steam could be with a refresh.

The thing we all really want

Chris Livingston

Let's address the strider in the room. For years now, every single bit of Valve news and every update Steam pushes through has resulted in a fresh round of sarcastic yet subliminally hopeful chatter: Half-Life 3 confirmed!

To which I say: slow your roll, Internet. We've got some unfinished business to attend to. A broken promise. A missing chapter. A little something called Half-Life 2: Episode 3.

I'm confident this is, deep down, what we all really want: another two-to-three hours of content using the same assets and enemies from Half-Life 2. We want to fight more slow, stupid Combine soldiers while they issue their familiar barks. We want to solve yet another giant see-saw puzzle with the gravity gun. We want to stare at a brick wall waiting for a new level to load, then run through the new level for three minutes, then stare at another wall while the next level loads. We want to climb back into that rusty, ugly-ass car and hit the gas only to immediately hit the brakes because the radar is showing a supply cache. We want to spend more time watching Alyx Vance do a bunch of cool things like bashing zombies with the butt of a shotgun or using a sniper rifle or climbing over walls, which we, as Gordon Freeman, cannot do ourselves for some reason. Most of all, we need to hold Valve accountable for their promise, made in June of 2006, that the three episode series will conclude by Christmas 2007. That's the "three" that's important, here. Not Half-Life 3. Episode 3. Make it happen.

j/k, of course. This year, Valve, just announce something, anything genuine about Half-Life 3, even if the announcement is that it's never going to come.

Valve's Source Filmmaker is regularly used to parody Team Fortress 2. Here, instead, it's being used to accurately (and stylishly) portray an entirely different game. With the help of TF2's Heavy, and taking a mere 24 seconds, here is Skyrim in a nutshell.

The short was created by Ferhod, who previously made the Saxxy 2014 winning TF2 film Animation vs. Animator.

As part of this year's Smissmas celebrations - it's a thing - Team Fortress 2 has received a new way to play it. It's called Mannpower Mode, and it features grappling hooks.

Merry Smissmas, one and all!

Mannpower Mode is very much a beta offering, being used to test the water for a bunch of different mechanics, including the grappling hooks. Because there are grappling hooks in it.

Currently you can play the capture the flag-alike Mannpower on modified versions of the Ctf_foundry and Ctf_gorge maps, and as well as the grappling hooks (did I mention those?) there are a few other tweaks and features added.

Nine power-ups, including strength, resistance, haste and warlock, are currently available, though there are more on the way. Each... well, powers you up, obviously, just in a different way for each of the nine.

There's also a grappling hook! Wait, I said that a couple of times already. But the fact is, grappling hooks are brilliant and should be in all games - so it's great to see them added for all classes in Mannpower Mode.

Accessing it is as simple as this: "Click on Play Multiplayer, enable Play Beta Maps checkbox, and then choose Mannpower Mode."

Grappling hooks!

Holy. Shit.

I've watched a lot of Source Filmmaker stuff over the years. I've seen the trailer for End of the Line about a hundred times. But now the full thing is out, and it's good. It's really good.

The community film was directed by James McVinnie, and releases alongside an End of the Line themed TF2 community update—containing hats, a new weapon and some new unusual effects.

I'd give a summary of the film, but why ruin the surprise? It's got a train in it, if that helps. It's also got a very distinct tone. The zanier edges of the TF2 roster have been sanded away, leaving a relatively dramatic piece that nevertheless contains more than its fair share of comedy.

Set aside 15 minutes and enjoy one of the best SFM films we'll likely see for a while.

Last month, Valve launched a promotional contest for the game Lara Croft and the Temple of Osiris. The studio asked their community creators to submit TF2 items related to the Tomb Raider series. In that post, they wrote the following sentence:

If you've been waiting for your chance to put the heavy in short shorts, this is it. (Please don't do this.)

There was only one possible outcome.

"You might remember we specifically told you not to make Heavy short shorts," wrote the TF2 team last night. "However, once we actually glimpsed the majesty of Heavy short shorts, we saw the error of our ways."

In addition to the Jungle Booty, other winning items include the Crown of the Old Kingdom hat and the Tomb Readers sunglasses. All items are now available to those who pre-order Temple of Osiris, and will eventually make their way to the MannCo store along with an additional three runner-up items.

What's the best stealth class in Team Fortress 2?

No, you're wrong.

It's not the Spy. You can't be the best stealth class if you've got a watch that literally makes you invisible. That's cheating.

The best stealth class in TF2 is the Scout. And he is a stealth class. He's frail, but fast quickly killed by a Heavy's minigun or an Engineer's sentry, but able to retreat and reposition for a different approach.

There are two things, specifically, that I love about the Scout. The first is the way he moves. I adore games that offer interesting methods of locomotion. The Scout is fast, and has a double-jump that lets you change direction mid-air. This makes slipping past, around or away from enemies feel great. It's no longer good enough for me to get where I need to go. I need to get there with style.

Stealth is an important part of this. A great Scout will, to the enemy team, appear to be everywhere at once: on their control points, patrolling their corridors and running full bore into their front line. Battles are about the constant flow of position—of where you are relative to everyone else, and of where you need to go in order to be where they least expect. It's not just that you can outrun your opponents; it's that you can outmaneuvere them. This challenge is why I've spent 300+ hours in TF2, and almost 50 hours as its annoying, scrawny shotgun wielder.

The second thing I love about the Scout is that he's a jerk. The Scout's job is to be annoying. There are specific feelings for being killed by each of TF2's classes. Being killed by a Heavy or Pyro feels like the continuation of some natural order. They are forces of nature (or, at least, of fire and meat). Being killed by an Engineer, Spy or Sniper is more cerebral. It's a tactical death—a specific and immediate punishment for a mistake. Being killed by the Soldier or Demoman is annoying, because you'll swear it was a fluke, and also because deep down you'll know that it wasn't.

Being killed by a Medic is, for the most part, humiliating. It's the Medic. The clue is in the name.

Being killed by the Scout is infuriating. The level to which it's infuriating is the result of a complicated formula based on a) how much you are currently sucking, b) how much your team are currently sucking, and c) if the Scout has a Force-A-Nature equipped. I have been specifically and graphically told how infuriating it is in hastily typed strings of four-lettered anger.

I should come clean here: I'm not just a Scout, I'm an unreformed Scout. The abuse is perverse positive feedback. It's how I know I'm credit to team.

For me, both of these loves combined into a single, terrible playstyle when Valve introduced King of the Hill mode, and specifically the maps Nucleus and Sawmill. Both are small—filled with side-routes and escape points. More importantly, both are covered in traps.

A scenario: a Heavy is capping Sawmill's centre point. An opposing Scout is running directly at his back. The opposing Scout has a Force-A-Nature equipped. He gets in close, fires, and watches as the shotgun's knockback flings the Heavy into the spinning buzzsaws. Do you know how angry the Heavy player feels in that moment? Conversely, do you know how elated the Scout player feels?

Based on his abilities, Valve's portrayal of the Scout is perfect. He is a jerk, through attitude as well as deed. It's not just a hint as to how he's best played, but a reward to anyone who manages to kill him. As fun as it is to kill with him, I recognise the catharsis for those who get revenge. In this way an uneasy balance is achieved.

November 16th, 2004 was a red letter day on both sides of the screen. The original Half-Life had redefined the FPS as an immersive experience instead of merely a series of missions, and no-one expected its follow-up to do anything less. Few were disappointed. In City 17, Valve created one of the most coherent and ambitious worlds ever seen in gaming—and if it looks a little primitive now, it s because so many since have followed in its footsteps. BioShock Infinite s opening for instance works almost entirely to Half-Life 2 s now dog-eared playbook, offering greater fidelity and a more exciting city, but recognizably the same style.

What Half-Life 2 really brought to the industry wasn t new ideas… though it absolutely had them… but demonstration after demonstration of how things both could and should be. Alyx Vance for instance was an effective sidekick and a fun character, but it was her ability to make a connection with the player through things like eye contact that elevated her above her others—something shared by fellow Source engine game Vampire: The Masquerade: Bloodlines, especially with characters like Jeanette and your pet ghoul Heather. She could shoot a worried look. She could smile, and have the smile go further than her lips. She could go directly from being your wing-woman to interacting with the world, from having conversations to fixing something, all as smoothly as Half-Life managed to never break perspective as you went from random lab geek to savior of two different worlds. She d even climb and vault over things instead of simply walking around the old fashioned way, easier as that would have been to script.

That sense of flow is what really defined Half-Life 2—sequences bleeding into one another to create the feel of an unbroken journey (give or take a loading screen). It was a game of smooth traversal around the maps, of combat bleeding into story, and each major section, most famously zombified Ravenholm, casually experimenting with the formula. Every cool bit offered something new and most left us wanting more, even if the radical shifts did take away much of the original Half-Life s thematic consistency. Every not-so-cool bit, like the dull start of Sandtraps (a vehicle section that paled in comparison to what Valve was able to pull off by Episode 2) was short enough not to be that big a deal, and something else was always on its way. While admittedly the story sequences are interminable by modern standards, the action was all about peaks and troughs that allowed both intensity and time to savor the craft.

On top of that came the details; a hundred things designed to be absorbed rather than directly noticed. The soundscape for instance, with the Combine announcements using medical terminology to describe uprising—Gordon Freeman as a staph infection —or the Combine s logo—an outreached claw almost, but not quite, absorbing a small world. It s a subtle detail, but one of many told through level design rather than audio logs or cutscenes. Others include visiting rusted playgrounds of a world without children, and seeing the empty seas that have left ships high and dry—the terraforming inflicted on the natural world mirroring the shift from old and human to new and alien that you see throughout City 17.

One of the most subtle, though much borrowed since, is the way that Valve tends to show new mechanics off three times—first with no pressure, then some pressure, and then for real to be sure the player has grasped and understood it, before it becomes an assumed skill. The gravity gun for instance. First you just move a solid box into position to meet Alyx. Then her robot Dog throws other boxes at you for you to catch. Then, it s zombie fighting time.

Speaking of fighting zombies, we can t overlook the physics engine—used in Ravenholm to let you hurl sawblades. Half-Life 2 was one of the first FPS games to go big on physics, for two basic uses. The first was, honestly, showing off. They were a novelty then, which came through in a lot of puzzles like pushing barrels under a platform to be able to cross it. Looking back, they re a little eye-rolling. At the time though, they were pretty cool. It s no secret that Half-Life 2 was at least in part a demo of Source, with these bits standing out even at the time as largely the equivalent of early 3D card lens flare effects. Cool, but gimmicky. When it showed them off, or put us in a vehicle, it was at least partly saying Look what we can do!

The big benefit though was creating a world that felt right, in stark contrast to the largely static worlds of the previous generation; of games like Return To Castle Wolfenstein and Medal of Honor. Again, yes, it s a bit showy to have a guard at the start insist you pick up and throw a cup into a trash can just to shout PHYSICS! What mattered though was that from there you both feel the benefit of them in every interaction with the world, big and small, and immediately start bemoaning their absence in any game that doesn t have them. The rolling and floating of flaming barrels in water. Debris flying off as it felt like it should.

It all added, in much the same way that the original Half-Life s responsive skeletal animation system instantly made conventional frame-based animations intolerably stiff. When Valve called its behind-the-scenes book Raising The Bar, it wasn t kidding. Half-Life 2 was an amazing game, but its crucial, lasting influence was less about the new things it did (as important as they ve been) as showing how the familiar deserved to be done.

Unforeseen consequences

Which of course brings us to its shining achievement—Steam. To sum up Steam s unpopularity in 2004 would leave no words left to describe ebola, lawyers, or Piers Morgan. And not without cause. It was buggy, it was ugly, there was no missing that Valve was outright forcing it down our throats out of nowhere, and the much crappier bandwidth of the day made being told to download games of this size almost offensive in its arrogance. It would be a long, long time before Steam even got close to the service that at least most of us know and love today, instead of its name just getting tacked onto the words ing pile of shit.

But. With Half-Life 2, Valve had a game that managed to get the necessary traction to create the service we know today, and while nobody would claim it s perfect, nothing else has done so much to legitimize and make digital distribution work. Much as it took Apple to break the music industry s obsession with DRM on MP3 files, it took Steam to show the whole industry that the game had changed. The idea that you d be able to redownload your games in perpetuity for instance was heresy to companies that at best wanted that to be another service. Being able to download them onto any machine instead of them being locked to a single PC, or maybe three, or five? That just wasn t done. Valve was the first major company to build a digital download service that people actually wanted to use, that made the experience of buying games online better. Without Half-Life 2 though, who would have used it? Without that audience, who would ever have agreed to let a competitor sell their games? Half-Life 2 didn t just give the FPS a shot in the arm, it changed the entire industry.

It wasn t a perfect game. It was far more a series of cool things than Half-Life s journey, it was heartbreaking to be taken out of City 17 almost immediately in favor of being consigned to the countryside, the story was primitive and a few of the set-pieces dragged on far too long. It held up pretty well for years, but looking back, yes, it s now a bit long in the tooth. Few games though have had a more lasting impact in so many ways, or can be deservedly held up as both paragon and pioneer. Fewer still have done it so well, they re still being copied a decade later.

Now then, Valve, about that Half-Life 3…