Wednesday 23 October 2013

Gone Home - The Amnesia Edition

From time to time we get requests from people who want to license our HPL2 engine in order to make a commercial game. This is quite flattering, but the answer is always "no". Our reply is to simply state that there is no documentation, we do not have time for support and they are better off using Unity or UDK anyway. In all honesty, we also do not have any high hopes of these projects finishing at all. Choosing an engine is one of the very first decisions made when starting the development of a game, and very few games, especially indie ones, go beyond a pre-production phase.

So it came as quite a surprise for me when I learned that Steve Gaynor, one of the people behind the phenomenal Gone Home, had sent this sort of mail to us! I met him briefly at GDC this spring, and was quite amazed to hear that the very first prototype of the game was made in HPL2. He had mailed and asked if the engine would be possible to use for a commercial game, and got the usual response. Fortunately this did not discourage the team from continuing. It also seems like they took our advice since the final version of Gone Home is made in Unity. I really wanted to see the level, and told Steve that I would mail him when I got back from GDC. But as always other stuff happened and I just pushed the thing forward. I swear that I had "Mail Steve about Amnesia: Gone Home" written on my todo list for 6 months!

Then Steve mailed me for totally different reasons, and I decided I really had to get this Gone Home prototype over with. He scavenged his files and managed to dig out the map. This was during the whole SOMA teaser campaign and I did not have time to look right away. A few days ago things finally settled down a bit and it was time to take a look.

The prototype is quite short and very basic; it is really more of a proof of concept. But it still gives a very good sense of the game, and having played the full version, I could recognize quite a bit. It does feel a bit awkward to play an early test like this though. Gone Home is a very personal game, and playing this prototype felt like a meta version of the game's voyeuristic thematics.

We got Steve's mail regarding HPL2 engine on the 14th of January 2012 so this prototype must have been made before that. This means the prototype is over a year and half older than the final game and made almost 5 months before the game was announced. My guess it is the first time that the ideas for the game got some sort of substance.

Here is a few comparisons between the prototype and the final game. Prototype is on the right (as if you couldn't tell...):

The game opening is in the exact same place, on the porch of the house. There is even still a chair and small table with a pot next to the front door!

The first key is still found hidden under an ornament! I think this is a very neat puzzle as it explains to the player that it is worthwhile to do some extra scavenging and acts as a sort of unobtrusive tutorial. So it is not that incredible that it stuck so long. But still, very fun to see this intact.

And here is the first view when entering the house. It is not visible from the screen, but both versions have paths leading both to the left and to the  right, giving the player three different ways to start their exploration. Like the key puzzle, there are good reasons this stuck, but it is still awesome to see it this similar.

Taken together, the prototype is really incredibly close to the final game.

In case you are not amazed by the similarities in such an old prototype, check how Amnesia looked at the end of 2008 (a few months less than 2 years before release):

I also have to note the awesome handy work on this toilet:

If you want to try the level out yourself, you can download it from here:
Just extract the file in the "custom_stories" directory in Amnesia: The Dark Descent, start the game, press "Custom Story" and select and start "Test Game".

Finally, if you have not played Gone Home, do so now! It is a really unique and emotional experience that is a must for anybody interesting in videogame storytelling. You can get it here:

Lots of thanks for Steve Gaynor for saving this lovely slice of history and letting me (and all of you now) try it out!

Tuesday 22 October 2013

Useful Tips for Horror Game Designers

A while back Chris Pruett (creator of the excellent Chris's Survival Horror Quest and currently at work with some creepy stuff at Robot Invader) and I had some discussion about common horror / puzzle tropes over twitter. Now all of these little nuggets, and more that came up during subsequent mail discussions, have been collected into a nice blog post by Chris. If you are ever going to make a survival horror please read this first. Here comes:

No puzzles about equalizing pressure (or any other type of dial) by adjusting switches or knobs. Do not include puzzles that involve reconnecting the power, especially to an elevator. No sliding bookshelves with scratch marks on the floor. Avoid puzzles that involve pressing keys on a piano in a specific order. Do not require the player to collect paintings to reveal a secret image, or examine paintings to decode a correct sequence of buttons. No locked doors with an engraved symbol that also appears on the key. No important documents encrypted with stupid-simple substitution ciphers.

As you design, repeat this mantra to yourself: "I will have no keycard doors in my game." No feeding fertilizer or poison to giant plants. Check yourself before adding puzzles about inserting crystals, gems, or figurines into some ornate locking mechanism. Reconsider any puzzle involving a four-digit number sequence, found elsewhere, that opens a lock.

Do not employ sliding block puzzles. Ever. That includes sliding statues! No!

Deny the urge to take inventory items away from the player without a legitimate reason. When building puzzles that require combining more than two items, you must allow combination of arbitrary pairs of items even before the entire set has been collected.

Do not turn terrifying monsters into puzzles unless your goal is to kill all tension.

It's important to make objectives and mechanics clear, but if you just tell the player what to do and where to go, you've removed the puzzle entirely. Let them think for themselves occasionally. Be especially vigilant when designing any cumbersome door opening apparatus. Remember, your players will only believe so much!

This got old in 1997.

Not all stories have to be about the protagonist's personal demons. Don't blame everything on evil mega-corporations. You don't need a crazy Special Forces unit with an awkward acronym name. Do not include a sequence in which a child must crawl through a small opening to unlock a door for an adult. No more helicopters escaping from mushroom-cloud explosions. Eschew underdeveloped sub-plots about drugs.

Avoid zombies. But if you must use zombies, for the love of all that is holy, do not rely on a virus to explain them. Zombie dogs: no.

Not all vengeful ghosts need to be women. And curses do not all need to spread like a virus. And the virus doesn't have to kill its victims after exactly seven days. Also, ghosts don't always have to be innocent people who died horrible deaths.

It's not very believable that a high-security military research complex would have passwords written down on scraps of paper. If your plot twist involves the surprise reveal of a secret, sinister cult, you should probably stop.

Try to think of ways to put your characters in vulnerable situations that are not limited to making all of your characters petite school girls. Men can be vulnerable too. Plus, I know some school girls that could wipe the floor with your sorry designer ass.

Levels and Characters
There are other ways to block a passage off than having the roof collapse. Make a distinction between locked doors that will eventually open and doors that can never be opened; if you have any of the former, the latter must be barred, or broken, or otherwise obviously forever inaccessible. Be warned, however, that "it's jammed" gets old mighty quick.

No arbitrarily non-interactive objects; either you can interact with all doors or none of them. Ensure that you have more doors that can be opened than cannot. Do not block the player with short fences or other obstacles that should be trivial to bypass.

If a location is supposed to carry emotional weight, do not litter it with ammo boxes and collectibles. Do you want the player to contemplate the horrible living conditions of a young child or rummage through their things looking for loot?

Just say "No!" to items that are of great use to the player's problems but cannot be picked up. No obstacles that could be easily dispatched using the protagonist's arsenal but instead require some puzzle sequence to overcome. Do not provide a stock of limited supplies unless you make the remaining amount clear. Do not put hidden collectables in horror games with large levels, or in games that do not allow you to backtrack. Maybe just skip the whole hidden collectable thing completely.

We don't need any more tentacle monsters in horror games. Especially not tentacle monsters with bright, bulbous weak spots. Avoid close-quarter combat with ghosts that can pass through walls. Never throw the player against a source of infinite damage unless you also provide a source of infinite health and ammo (e.g. infinite enemy spawner).

Little known fact: not all monsters have an irresistible urge to bare their teeth and scream at the player. Nor do they all hunch over with long, bent arms. Crazy, huh!?

Excepting certain types of zombie, it is almost never exciting to see a monster charge the protagonist. Perhaps you can modify your AI to stalk the player and approach him slowly to appear more menacing? Caveat: circling the player and occasionally revealing a weak spot is not a good alternative.

Ask yourself: "how many times have I been to the gym this year?" You're a game designer, so the answer is probably "none." Do you think your game's cultists have it any better? They're too busy summoning an obscure deity to think about their diety. So why did you make them look like they're all bodybuilders and/or silicon implant models?

And while we're on the topic of appearances, does your monster really need that awkward underwear? I mean, you just had him rip a dude's head off in the last scene; I don't think your audience is going to be phased by a little monster nudity. Or heck, just come up with something else. Tiny bits of torn fabric around the midsection of an otherwise naked beast is a cop-out.

Took forever to find pants in my size. And now they're torn.

Technical Stuff 
When you have a body lying on the floor that is significantly more detailed than all of the other bodies on the floor, we all know that it'll come to life and attack us sooner or later. Also, a surprise attack isn't very surprising if the game suddenly starts loading like crazy moments before.

Do not put scary encounters in cutscenes. I know, I know, you want to control the camera and the timing and the sound so everything is "just right." But listen, games don't work that way. Take a gamble. Let the player discover the monster through gameplay.

Navigating save slots, confirming file overwrites, and waiting for flashy menu animations is pretty much the worst possible thing you can subject a player to. Your sense of presence must extend to the game as a whole, even your UI.

If you have item descriptions, why not make them interesting or useful? Everybody already knew it was a trashcan before they examined it.

It may sound a bit unintuitive, but horror games work surprisingly well without rocket launchers. And you'd be surprised how fun mystery games can be when they don't have RPG mechanics shoved into them.

Fail in every other category if you must, but do not fail in this: map and menu screens must not require a loading pause to display. It is bad enough that you have to bring these up in the first place. Oh, and checking the map every two steps is not fun.

10 seconds of loading to tell us that flashlights are useful in the dark. 

Follow these tips and you'll be well on your way to making a horror game that is fresh and original! After which you can make endless sequels!

Friday 11 October 2013

SOMA Officially Revealed


After more than two weeks of teasing, we are happy to properly announce our new game: SOMA. It is a sci-fi horror game played from a first person perspective with the goal to deliver a truly disturbing experience. Having worked on this project for over three years now, it is really nice to finally reveal it to the world!

The game will be released for PS4 and PC, and it will be out some time in 2015.

Here is the official reveal trailer featuring some gameplay:

Making The Trailer
While the footage above is unedited gameplay footage (I played through and recorded it myself), it is not from an actual game level. Just like with our Amnesia teaser video, it is a custom made map that is meant to show off the general feel and tone of the game over a four minute long trailer. This means that it is not an exact representation of how the game is actually played. It gives a pretty good idea of how the gameplay works though.

Coming up with the idea for this trailer was quite hard. We knew from the start that we wanted to have it as recorded gameplay. It was crucial that it showed that SOMA will allow you to play through its central themes. The subject matters of the game are not some kind of wrapper, they are an integral part of the gameplay. So having a trailer entirely made out of uncut gameplay felt like the best way to show this.

With Amnesia doing this sort of thing was fairly simple as the core ingredient was the player running and hiding from monsters. In SOMA it is much more complex. While the game also has its share of monsters, hiding and stuff like that, it is not what makes the game special. What sets SOMA apart is that it gives a first person account of some deep and really disturbing ideas regarding the self, mind and consciousness. These are things that we take a lot of time to build up in the actual game, so showing it off in a few minutes is quite difficult.

Our first idea was to use a lot of dialog to get this across, but that did not feel right. The player will be be an active participant, and do not sit by passively listening to characters having discussions. So instead we approached the themes in a very direct and visceral manner.

A final aspect of making the trailer was not to have too many spoilers. Because of this, the video does not star the protagonist from the actual game, but a totally different (and minor) character. The events that occur in the trailer are not taken directly from the game either. They just showcase the kind of happenings one can expect from the final game. This means that we can give a good overview of what the game is about, without spoiling the actual game experience. But don't worry, the things in the video are very much related to the game's story. Everything you see are hints of things to come.

Live Action Videos
Another big part of the reveal for our game was the release some creepy and mysterious teaser material. You can see texts and films here:
and here:

First I guess I have to settle the big debate: Are these inspired by SCP? And, yes they are! That was actually the pitch for the whole thing: "Let's have some SCP-like texts on the website to give out spooky and fragmented info before the final trailer is released.". The game itself is however not very close to the SCP-style at all. There are some SCP inspiration in SOMA for sure, but it is a lot more subtle and has to do with how we think about monsters and artifacts. They play a larger role now, the foreboding is much deeper and there is a bigger connection with them and the central themes in the game. Actually, a lot of the SOMA's themes are directly expressed through interactions with enemies, an idea that stems from SCP.

Enough SCP, let's move on to the actual teaser texts and films. At first the teasers had a much weaker production values. The idea was just to use plain text and perhaps a few images. As Mikael wrote these he suggested that we might add sounds to them, e.g. a recorded interview, and that felt like a good idea. But that was as far as we thought about going.

When discussing a release trailer for Amnesia: A Machine For Pigs, we came upon Imagos Films and decided to give them a go. Jens contacted them about making a trailer, and they were really excited about it. They even agreed to having a meeting in the middle of the night in order to fit our schedule. Some quick drafts were made for a trailer, but because of various issues, it all fell through in the end. We felt bad about this and promised they would be our first choice when needing some other video material.

A few days after Mikael suggested having voices for the texts, it hit me that we could perhaps use Imagos Films to do some simple clips instead. We mailed them and they said they were up for it, so Mikael converted the transcribed audio from the documents to film scripts and then we set the ball rolling. I did not really expect much from the films, but after seeing the first few WIP shots and production designs, I knew that it could actually turn out really cool. But the end result turned out even better, and it blew my (and everyone else's at FG) mind.

It has to be mentioned how quickly these films were put together. The first script drafts + instructions were sent on the 2nd of september. This meant they had a less than a month to produce the first movie. Then another movie needed to be done a week after that. Given the amount steps needed to go from script to final film, this is insanely fast.

Will there be more movies? We would really like that to happen, but nothing is set yet. Stay tuned for more news regarding that!

End Notes
This reveal has been quite a lot of hard work and there have been way more things to deal with than I first thought. Worst must have been the major server failures that we had on the days before the first teaser would go live. This turned out to be because of a DDoS attack and took quite a while to fix (Jens had to come out of a parental leave to work on it). We had to put off the whole first reveal for a day because of this. It was quite embarrassing, as people wondered what the hell happened and we had to reschedule a bunch of other stuff connected to the reveal (like the PlayStation Europe twitter).

And then of course the final trailer reveal had to be problematic too. Early on the day before reveal (as of writing, yesterday) Tapio, external sound person, had finished editing the trailer, all of David's, external art person, animations were in and all sound was synced and nice. However, as he exported the final version the quality was crap. I had hoped to have a calm day and just code stuff but instead I had to help Tapio search for the error, trying to convert files better, etc. Nothing worked. On top of this, our ftp servers were really slow and sending files took a long time and was a general pain. Eventually, I had to re-edit the entire trailer (adding animations, fades and syncing sound)  using Windows Movie Maker as quick tests showed the final quality was much better there. At the same time we became aware that we also needed an ESRB logo and had to scavenge the net for a proper one. We were now much later than planned and stress was taking its toll. After many grueling hours we finally got it working though. Only to find out the next morning (day of the reveal) that the damn ESRB logo was not shown long enough, so more editing, exporting and uploaded were needed.

In the end it worked out fine though and the whole reveal has been very successful. (As I am writing this we have not seen any reactions to the final video, so I am hoping those are good .. :) )

Now it is time to go back full time on work for the actual game. We are all incredibly excited about SOMA and hope that these 2+ weeks  of teasing and reveal have gotten you all interested aswell.

And do not worry, between now and the 2015 release there will be a lot more unsettling stuff revealed!

Want to end with some links to the non-Frictional people that helped this reveal happen:
The makers of the two live action teaser films.
The coder of our teaser site.
Made gameplay trailer animations, SOMA and FG logo, website graphics and some teasing images.
Made all the sound work for the gameplay teaser.
Who sent out PR and nagged various publications. Jens, who normally does this, has been on parental leave and I would have gone mad if I were to do all that too....
Made trailer-specific animations for the corpse and monster encounter.