Meri-Making, August Edition: Mega-crunch!
Meri Mega-crunch & "Table-Read"
Last week we all hunkered down in Josh’s apartment to eat venison, drink gallons of coffee, slather stuff is Sriracha, and--oh yeah--work on Meriwether. We reached a major milestone during the crunch: we have written ourselves all the way to the Pacific Ocean! The major story elements are now written, scripted (meaning that flags and events are activating properly) and implemented in-game. We’re super-stoked about how much content we have been able to include in the game: as of this writing, the word-count for the game dialogue stands at just under 80,000, a word-count similar to a decently-sized novel. You will have control over how much of the dialogue you engage in (to a point), but even if you don’t read every line, you can rest assured that we have worked hard to create a game-world that feels alive by imbuing Meriwether with a substantive narrative. And we’re not done yet! It’s likely that the final word count for dialogue will reach 100,000 words before all is said and done, and that’s not including supporting text, such as the Wunderkammer!
Now that we have reached this milestone, we are doing a “table-read” of the game level by level as the first step in revising the text. We’re using “table-read” to mean that we’re just playing through the dialogue of Meriwether without worrying about any other aspects of gameplay, so that we can get a very clear sense of how well the dialogue is working solely on its own merits. The table-read will continue for the next few weeks and help us improve the quality of the writing, check the continuity of story arcs and missions, and copyedit. Things are going to start coming together very quickly for Meriwether’s overarching narrative from now on.
Google Hangout & Music
Jim is in Texas and so obviously was unable to join our Brooklyn session in person. But he was right there with us on Google Hangout. It was a great way to work together. Just hearing fragments of the tunes he was writing put us in the right mood and helped us all focus. Before you go any further, scroll down to the bottom of this page, hit play and listen to this draft of "West of the Rockies."
New Dialogue Camera System
Kyle’s focus during our crunch was revising the camera system. Our first step was to identify a number of shots used in cinematography, particularly those used for dialogue. We collected references for various types of shots, such as “full shot,” “mid-shot,” “close-up,” “extreme close-up,” “American two-shot,” “over-the-shoulder,” and others. Based on these references, Kyle extracted a set of rules that could procedurally recreate these shots, given an arbitrary set of relevant characters in a scene. This is going to get technical, but that’s game development. :)
The basic technique is to identify a pleasing spot to look at, a pleasing angle to view it from, and a pleasing distance to zoom the camera. With some shots this is easy, as the rules are very specific (for example, the extreme close-up or over-the-shoulder two-shot.) For shots of three or more characters, the guidelines become less specific and require a more advanced algorithm. The algotirthm determines the viewed point based on an averaged position of the characters (so the camera tends to view the middle of the group). Next, it finds an ideal viewing angle to ensure that we don’t have several characters lumped on top of one another in the view if possible. Finally, the algorithm determines how far it must zoom out from the viewed point along that angle to ensure that all of the required points are in view. For example, with a wide shot of three characters the camera must view out far enough to ensure that all characters feet are within view. However, another shot may only require the camera to zoom far enough that the waist of each character is in view. One other thing we learned from looking into common cinematic camera shots is that they tend to "cheat" the angle a little bit by slightly rotating the actors towards the camera. It gives you a better angle of the face while being subtle enough that you don't notice it when switching to an angle on the opposite side. We decided to imitate this technique, although we left out another common aspect of the technique, which is to place the actors closer to each other than feels natural, since we need room in between characters to display dialogue.
The Great Plains
George S. has been busy sprucing up the Great Plains. Those gorgeous prairies are probably the most iconic terrain of the journey. This has been a technical challenge, since games often rely on obstacles like trees to obscure your view to reduce the number of objects that need to be simultaneously drawn on screen. It has also presented some interesting game design challenges, such as how to make a mostly-empty space interesting. Next month we plan to focus more on art, but here’s a work-in-progress to whet your whistle.
We’re lucky to have such a talented team working on Meriwether. We thought we’d share a little bit about some of the teams’ other projects.
George S., our environment artist, was also one of the environmental artists on Saints Row IV, which was just released! If you’re in the mood for some loony, surreal fun, the Saints Row franchise should scratch that itch!
Jiyoun has been very busy on a broad variety of projects, including the artwork for the iPhone game 'Baboon', in which you protect wild animals from poachers in sanctuary. She has also just illustrated a set of stickers and smileys for the new multi-platform texting app HelloPop. She recently had a solo exhibition of her fine art called Roommates at Coohaus Art in Chelsea. She is also preparing a 2-person art show 'Urban Jungle' at BBCN bank, in Woodside, NY, Opening Sept. 12th.
Josh has been refining Killer Queen, a 10-player real-time strategy platformer arcade game that he co-designed with Nikita Mikros. It's a team game inspired by arcade classics like Joust and Mario Bros. In fact, they are hosting a tournament for the game tonight, August 23rd at 7pm, in Brooklyn, NY. And if you're not in NYC, you can watch it live on Twitch.TV at 7:30PM EST! Details here: http://on.fb.me/1d5eSGi
Carlos has sold two stories so far this year. "The International Studbook of the Giant Panda" is an sf story about using robots and virtual reality to help pandas in captivity reproduce (!); it was published in the venerable British magazine Interzone (March-April edition, http://ttapress.com/1603/interzone-245/). His story "More Than Pigs and Rosaries Can Give" was published in the anthology Exotic Gothic 5, Vol. II and features a Cuban spiritualist who pulls ghosts out of the bullet holes left behind after a firing squad. Check it out here! http://www.pspublishing.co.uk/exotic-gothic-5--volume-2-edited-by-danel-olson-1766-p.asp
Jim has been working on a mobile game called Snake Esc. Check out the awesome trailer at http://www.youtube.com/watch?v=NdCTpxtB3eE. And he also just released a Unity sfx pack for digging/mining/chopping sounds (http://u3d.as/5b5).
Barb recently published an article in We Proceeded On, the scholarly journal of the Lewis and Clark Trail Heritage Foundation, about the early years of the publication, which started in 1974.
Op-Ed: On Saving Games
Carlos here. One of the first Kickstarter projects I supported was the Shadowrun Returns computer role-playing game. It’s a spiritual successor to the SNES Shadowrun game that came out decades ago, headed by one of the original creators of Shadowrun, Jordan Weisman. The game has received solid reviews from gamers and critics alike (as of this writing it has a Metacritic score of 76), and I personally would rate it much higher. The quality of the writing is excellent--it struck the perfect balance between campy fun and sudden, surprising insight. Without spoiling anything, I believe the dramatic irony in the game’s denouement is one of the most poignant moments I have experienced in a video game in a long time.
The game has received some criticisms, however. The biggest complaint by far centers around its save-game system. Basically, the game only saves your progress between levels. If you die in a scene, you start over at the beginning of the scene.
There is definitely an “oldskool” feel to that kind of checkpoint system that, for those nostalgic for a ‘90s CRPG experience, might be seen as a plus. But a great many players saw it as a huge problem. Since a single scene takes anywhere from 10-30 minutes to complete, a player could lose a great deal of progress when a battle goes bad--and since Shadowrun Returns is partially luck-based, sometimes things do go bad. Furthermore, without a “save progress on quit” function, some people lost their progress for extrinsic reasons--child needs attention, cake burning in the oven, etc.
I felt the pain of the save system a couple of times. I admit that even though I am a fan of the game’s story, I did not like rereading a heap of dialogue I had just read, just so I could get back to the battle I had just lost. Urge to ragequit, rising!
Still, it only happened a few times to me, and it was in no way a dealbreaker. Most of the time saving between levels was more than adequate for what I needed.
The good folks at Shadowrun Returns, listening to player feedback, have recently stated that they are looking into ways of creating a better save system. But it’s not as simple a problem as it might seem. Saving at any arbitrary point in a game can be complicated and costly, since it requires remembering the state of every single thing in the game, and then being able to load it back up. Also, there is a game design concern. There is a phenomenon known as “scum saving,” whereby players are saving their games constantly and reloading whenever the least little thing in the game doesn’t go their way. This, in a very real sense, can destroy a game’s design, making the game much too easy (especially as time goes on) and therefore much less challenging and fun.
Needless to say, after playing Shawdowrun Returns and following the brouhaha around saving, we here at Meriwether have been thinking carefully about how we should implement saving games. We want players not to be frustrated, but we want to have a system that makes sense in terms of the game’s resources and that does not promote an unfun, “scum saving” gameplay style.
After a good deal of discussion, we think we have it. Recently, and independently of the save discussion, we implemented a day/night system and a concomitant camping system. A day in game lasts a fairly short time--somewhere around five minutes. While that may change slightly, we don’t expect it will change dramatically. So we think that saving the game at the end of each day as you camp is a terrific solution. It will definitely preserve players’ progress very often, but individual conversations, interactions, and other choices you make throughout a given day will still have consequences. We also love it thematically, as saving the game day-to-day will allow you to page back through your saves and note the progress you made each day: which will feel very much like you’re paging through Lewis’ journal.
Of course, we need to test the system thoroughly to know how well it really works, so this is not the final word on saving games! But we want you to know that it’s not just the history we are trying to get right in Meriwether. We are thinking carefully about every part of the game experience to try and make the game feel great, even when we’re talking something as low-level as saving progress.
It’s all coming together: the writing, the art, the music, the programming, the design. Things are only going to get more exciting from now on. We will continue to work and continue to update you on our progress. Until next month!