Interview of adelikat for GamersGlobal
This interview was answered by adelikat for a German gaming website called GamersGlobal in August 2010. The interviewer is Christoph Hofmann. Most of the links and formatting were added afterwards.
Please introduce yourself to our readers. Who are you and what’s your association with TASVideos.org?
My name is Andrés Delikat, and I’m an administrator of TASVideos. I’m also an active maker of TAS movies having put out over 70 published movies over the course of the past 5 and a half years . I did not create the site nor did I start the community - that was done by Bisqwit in late 2003. He stepped down from the site administration duties and I took over in summer of 2009.
What prompted you to join the site?
In 2004 I was searching for more advanced strategies and information about my favorite NES classics. I was using nesticle (an NES emulator) to record movies but was frustrated that I couldn’t use savestates while recording in order to demonstrate various tricks and strategies. A friend had told me about "some Japanese guy who could play Super Mario Bros. 3 in 10 minutes" (which turned out to be the famous Morimoto TAS). I was intrigued and when I searched for it, I stumbled upon TASVideos. When I learned they were recording movies using savestates, I was excited to finally find such a feature. When I saw the kind of movies being created with these tools, I was blown away. I joined the site and started making my own movies.
How and why did you become an admin?
As I started to make more and more TAS movies, I became more involved with the site. I started watching and voting on new submissions and soon became a judge. I learned encoding and began publishing movies as well. As time went on I became more involved with admin duties. In summer 2009, Bisqwit stepped down as the admin for personal reasons. Nach and I volunteered to run the site. I took on the duties of admin because I love the community and didn’t want to see it break apart. I never had any intentions of running a community, but I stepped up to the challenge in order to keep it driving.
What are your exact duties as an admin?
Being an admin involves a variety of expectations. I need to manage expenses (such as server rentals, domain registrations, etc) and keep track of donations. I also need to manage the site code and the database. TASVideos is completely run by volunteers on their own time (and for free) so a lot of my time is spent finding and training dedicated helpers, and keep them motivated and on track with the help they give. This also involves deciding and communicating the overall direction of the site and community. I also need to see to it that new submissions are being judged & processed in a timely manner.
How many people are part of the core staff and how’s the work divided?
We have a number of user roles such as publishers, judges, and editors. The site is a wiki engine so any user willing to volunteer can become an editor and create/edit site content. Judges determine the fate of submissions, and publishers get approved submissions published to the site (including making a quality video, streaming it, mirroring it, setting up a torrent, screenshot, and creating a movie description). There are about 15 people involved in higher level user roles (judge or higher), and about 40 additional people in editing roles.
Similar to Unassisted Speedruns at the SpeedDemosArchive, Tool-Assisted Speedruns also have to go through an extensive approval process. As a judge, you too are part of that process. Could you tell us a bit more about how it works?
TASVideos strives to provide movies that are entertaining and of high technical quality. The judge process involves deciding if movies follow our guidelines, meet quality standards, and are approved of by the audience. When a movie is submitted to TASVideos it is displayed as a new submission on the site. Users vote and provide feedback and a judge ultimately decides its fate using their own judgment as well as audience feedback to make the decisions. If accepted, a movie will be encoded into a multimedia format (such as .mp4) and streamed on various streaming sites and then published onto TASVideos.
A TAS per definition has been artificially created using emulators, bots and such as opposed to Unassisted Speedruns. How do you validate such a run?
All submissions require the submitting of the movie file (input log) of the TAS. This is a file that contains all the buttons pressed during the movie. This leaves the movie completely exposed thus easily verified. If a movie syncs on the correct game with an approved emulator, there is no possibility of it being "cheated". The author is welcome to provide a YouTube-video or such along with the submission, but the movie file is what is being judged.
Some TAS exchange speed for entertainment purposes. In which cases is such a trade-off allowed?
The easy answer is: When it is more entertaining to do so. Typically the authors make these decisions on their own. It depends on the game and situation really.
What are the core principles a run is judged by?
TAS movies are art. Good TAS movies demonstrate TASing as art well. Pretty much all guidelines/rules are merely logical consequences of this concept. TAS movies should be impressive, superhuman, unexpected. They should look inhuman and display preposterous feats no human could ever achieve. One means of achieving this is aiming for a fastest completion time, but there are certainly other ways of demonstrating these concepts.
Why are some games not suitable for a TAS?
The concept behind Tool-assisted Speedruns/Superplays is the idea of using emulator features (tools) to go beyond human abilities and instead push the limits of the game itself. TAS movies strive to feature unexpected/impressive possibilities that these tools can provide. Some games provide little to no opportunity for showcasing this. For instance, a game like Where’s Waldo where "beating" the game fast is trivial and the solution for doing so is obvious. In addition it provides no creative uses of human input to showcase extraordinary events.
Who thought of the rules a run is judged by? The founders, the community, both?
Many of the rules were established originally by Bisqwit, the original creator of the site. They were founded on the principles of providing entertaining & artistic TAS movies. Those rules were expanded upon and evolved by the community itself as well. The community has a big impact on the rules, and those rules change overtime as the concept of TAS evolves.
What’s the most debated rule currently in place and why is it so highly debated?
I would probably say the rule on categories. Originally TASes beat the game as quickly as possible. But soon movies that featuring all levels or 100% type goals were being created. Other more unorthodox goals were done as well. TASVideos rules try to limit the number of different categories published of a single game and as well as restricting a lot of unusual goal choices. This rule has lead to some heated debates in the past. Such as Saturn's Super Metroid run.
How long does the verification process usually take?
The process of publication can take anywhere from 1 day to several months depending on various circumstances such as interest level in the submission, controversial situations with the submission, time it takes to encode the movie, and the free time of judges.
TASVideos is actively involved and supports the development of emulators which are especially useful for TAS. You yourself are a coder and are the main developer of FCEUX. What makes an emulator TAS-friendly and accepted within the community?
Tool-assisted Speedruns depend on the use of emulators implemented with rerecording. Without them, TAS movies would not be possible. The bare bones of rerecording is the ability to capture game input into a movie file while using savestates to "undo mistakes". However, while this was adequate in the early days of TASVideos, it has evolved to use much more refined concepts and tools. A TAS friendly emulator is a rather tall order these days. It demands a very accurate emulator, with a polished user interface, a good implementation rerecording concepts, and a host of tools to aid in TASing. Official "acceptance" of the emulator by the site demands that a quality TAS can be created using it. However, it still has to pass the audience test. If it isn’t up to par for the users, they won’t use it, and thus TAS movies will not get created.
Who makes sure that the emulator stays true to the original system and won’t interfere with the original games and ultimately with the run itself?
No emulator is 100% accurate and true to the original system. But over time they become more accurate through development. TASVideos strives to be theoretically reproducible on a console but we are bound by the practical limitations of the software that exists. Developers are continuously making more and more accurate emulators for various consoles. Usually the differences between an emulator and real hardware are relatively insignificant and do not impact the validity of a TAS in any significant way. However, occasionally emulator accuracy will be put into question by certain tricks & glitches. For instance, a Super Mario Land 2 movie used a game breaking glitch to beat the game quickly. But later it was discovered that the glitch was in fact an emulator glitch and not possible on real hardware. Once this was discovered it was rejected, and the bug in the emulator was fixed. Situations like this are quite out of the ordinary however. In reality, the difference between different emulators and a real console are usually insignificant differences such as how games lag.
But isn’t lag a serious problem? A lag of 50-100ms in Super Mario Bros. for example makes an Unassisted Speedrun almost completely impossible. Why is that insignificant for a TAS?
Maybe the game lags for a frame on frame 500 in emulator x, and frame 510 instead on emulator y, or none at all on the real console. The end result is rather insignificant over the course of the entire movie. Such small differences would be rather inpreceivable. Yes, a movie would then not be able to sync on all 3. But a movie could be created on all 3, and all 3 would look virtually the same.
You’ve been with the site for five years now – how has the TAS-community changed in that time from your point of view?
The community has grown larger and more varied. Originally the community centered around classic gamers who were fond of NES, SNES, and Sega Genesis. Since then, rerecording emulators for newer consoles have been developed. With Playsation, N64, Gameboy Advance and Nintendo DS rerecording, newer audiences have gravitated to the community. The site has become a lot more accessible as well, which makes it easier for new users to become engaged in the community.
So how big is the community?
In the grand scheme of things, it is a very small userbase; smaller than the unassisted speedrun and certainly only a small fraction of gamers. But the site is reaching out to a wider audience now than ever before.
To that end: Is TASVideos.org actively promoting TAS at events like MAGFest or PAX or do you keep more to yourself?
TASVideos has not actively promoted any such events. I don’t think that is necessarily from a "keeping to ourselves" attitude.
What is it then? I mean, promoting TAS would certainly help to widen the audience even more, leading not only to more viewers but certainly also to more authors and programmers, wouldn’t it?
I guess it is more of a "we haven’t done that yet". It is has only been 1 year since we took over administration duties from Bisqwit. He did have a keep to himself attiude, but we don’t.
Where do you personally see the future of TASing and TASVideos.org specifically?
I see it becoming less esoteric and more accessible to general audience. TASing is limited to emulators that support rerecording but the numbers of emulators that have such features are increasing over time. I see the average emulator developer being more aware of TAS features and having more of a demand to implement them. Newer consoles are on their way to being TASable.
Most Unassisted Speedrunners we talked to don’t consider TAS a valid form of Speedrunning. Is there something like an ongoing feud or are most of the speedrun-lovers watching both sides of the fence?
Unassisted speedruns are about demonstrating human skill and pushing one’s own ability to its limits. Tool-assisted movies are about demonstrating creativity and critical thinking and pushing the game to its limits. By unassisted speedrun definitions, TAS movies are indeed not "valid". They are their own concept, though both are related. Both fall into a "competitive gaming" genre. In the early days, there was definitely a lot of resistance and bad attitudes from console speedrunners. They feared TAS movies would belittle their talents. That attitude as waned since and TAS movies have not proven to hurt the speedrunning community. On the contrary, both compliment and help one another. I know several tricks I’ve discovered while TASing have been used by console speedrunners. Also, these days there is a lot of overlap between the two communities.
But TASVideos and the Speed Demos Archive for example don’t openly cooperate with each other to that end, do they? Like you said, finding tricks and glitches benefits both sides of the fence.
No we don’t cooperate officially. But like I said, we have a lot of overlap in our communities. For instance, andrewg is both the console record holder of Super Mario Bros and an active TASer of the game, having done a lot of ground breaking work in the discovery of new glitches represented in our published TAS movies. From a website perspective, however, we have not intermingled. I have never spoken with the administrator of Speed Demos Archive. I’m not opposed to the idea, but I guess nothing has necessitated that yet.