When an Australian gamer called “Anti” completed a full playthrough of Grand Theft Auto: San Andreas in a scant four hours, the feat almost seemed impossible. Yet any fans of speedrunning—an activity where die-hard players jockey to complete the game as quickly as possible, with different rulesets forming discrete “categories” of competition—could see this incredible “run” for themselves on the game’s leaderboards. Anti had posted the entire thing online.
An old saying may be coming to mind, and yes: it was too good to be true. A fellow competitor started analyzing Anti’s videos to optimize their own in-game routes, but they noticed that several vehicles in these runs left a faint smoke trail when they accelerated. Since no other runs on the GTA: San Andreas speedrun leaderboard evinced this telltale exhaust, this competitor began to wonder: was Anti somehow messing with the game in order to pull off this record-breaking time?
In the PC versions of the GTA games, after all, the files that control the way cars perform are easily accessible via a plain text editor like Windows Notepad. Game fans know this. And by slightly boosting certain variables to make cars accelerate ever-so-slightly faster, this fellow speedrunner was able to recreate the smoke effect in Anti’s runs. Soon, several runners started complaining to the greater community; someone even created a slick montage full of evidence that Anti had modified the game in order to shave vital seconds from their records.
In August 2018, the leaderboard moderators finally felt compelled to act, announcing a one-year-ban for “Anti” due to suspected cheating. The decision sent shockwaves reverberating through this tiny clique of GTA speedrunners, especially since Anti was an accomplished runner in the scene. At the time, Anti held several world records in each game in the series.
Gotta go fast
Ever since the first two hopeless MIT geeks battled it out in primitive progenitors like Spacewar!, video game players have cast their hobby as a form of heated competition. But while it’s the behemoths like Valve that finance the glittering million-dollar tournaments for esports like Dota, on the grassroots side of the spectrum, there’s a vibrant scene of gamers who pour their hours into an entirely different competition: that of “speedrunning.”
Since the mid-2010s, speedrunning has exploded in popularity as a pseudo-spectator sport, thanks largely to the annual charity event Awesome Games Done Quick (celebrating its tenth anniversary this January), where runners are invited to show off their skills on-stream to raise funds for organizations like Doctors Without Borders. Outside the outsized spotlight of this biannual spectacle, however, the competition rages on platforms like Twitch, where determined runners stream their attempts to break into the all-important leaderboards hosted on Speedrun.com. However, in certain walled-off corners of this tiny world, members of the community have begun to publicly question the legitimacy of certain competitors and records, wrenching open a Pandora’s box of controversy that some runners feel threatens the entire foundation of the hobby as they know it.
In many ways, Anti’s faked GTA runs might stand as a microcosm of the ever-present spectre of cheating that has lurked under the surface of the hobby for years now. Speedrunning fans know the script: first came the announcement, then a half-hearted apology, where Anti admitted that at least some of the runs were cheated. Next came the backlash against the moderators from fans of Anti’s streams, who stated that the apology was a sign of growing maturity. That was, of course, followed by the backlash to the backlash, when the moderators decided to make the ban permanent after analysis of Anti’s older runs concluded that the speedrunner’s confession was far from complete and that the cheating was far more widespread than initially claimed. Eventually, after the initial explosion of drama, Anti’s runs were scrubbed from the GTA charts, and the controversy ebbed away.
According to YouTuber Ben “Apollo Legend” Smith—himself a divisive figure in the world of speedrunning, largely due to his fiery fusillades against some of the biggest players in the space like GDQ—incidents like these are part of a growing trend of trusted, even prominent runners being exposed as fraudsters in sub-communities that they themselves helped build. But while the level of scrutiny that Smith and fellow content creators bring to bear on these alleged cheaters is certainly a new development, the concept of cheating is not. In fact, it’s almost as old as the hobby itself; one GoldenEye 007 runner was faking his runs as early as the late ‘90s. So as the world of speedrunning continues to grow larger and larger today, Smith worries that the methods that cheaters employ will only grow more sophisticated—and, with enough time, perhaps they will become entirely undetectable.
“I think people have probably been cheating as long as speedrunning has existed,” Smith says. “I think it’s only natural. But as more people come into the hobby, that number only grows. It’s a pretty big problem, in my opinion.”
Cutting-edge cheats
While Anti’s approach of modifying game files for a competitive edge might have gone unnoticed for months, Smith says that such strategies are often quite counterproductive when compared to more subtle methods. Throughout the history of speedrunning, by far the most popular form of cheating has remained “splicing”—stitching together several segments of different recorded runs to give the illusion that viewers are seeing a single, smooth playthrough. By doing this, a runner can throw themselves into the breach of a game’s most difficult sections as many times as they want with impunity, which gives them a massive advantage against legitimate competitors.
“Modifying the game is a sort of ingenious method, in a way,” says “Retro,” a UK-based YouTuber whose video on cheating in speedrunning remains perhaps the most popular analysis of the topic. “But there’s an obvious problem with it, which is that if you want to hold the world record, a bunch of people are going to be watching your runs, trying to learn from you. And if they know just as much about the game as you, they’re going to notice whatever discrepancies are there... I think it’s just a matter of time before somebody notices that your car can get down the road a second faster than theirs can, and then you’re caught.”
As Smith and Retro note, most of the popular speed games are played on the closed ecosystem of a console. That environment makes it difficult—but far from impossible—to futz with the code within, compared to Anti’s approach to GTA on PC above. Since all speedruns are verified with video, splicing is instead a cheap and universal method for goosing your times. However, if you actually want to keep your falsified record for any significant period of time, you have to pull it off with grace and finesse. Often, splicers accidentally leave evidence of the stitching process in the audio or video of their runs, the sort of subtle inconsistencies that you have to closely examine the footage to see—a slight tear in the audio or a loading screen that flashes the wrong color for a split-second.
But as splicers have grown more savvy over the years, fellow runners have deployed their vast expertise in increasingly creative ways to prove that their compatriots are, in fact, cheaters. Some proof is merely mundane: for example, in Legend of Zelda: Ocarina of Time, when a player pauses the game, the game remembers the last menu pane they were on, and it displays that specific screen the next time they press the Start button. Therefore, if there’s a discontinuity between the panes at any point in a given run, then you have evidence of a clear splice. Other evidence can be more arcane: in Super Meat Boy, eagle-eyed runners discovered that a running figure on the game’s loading screens always added up to a set number of frames. And when a dedicated soul performed painstaking video analysis on the record-holders, they determined that several of those runs didn’t reach the magic sum and were therefore suspect.
For Retro, this is the major downside to the splicing strategy, and it’s a key weapon in a good moderator’s arsenal. As a result, in order to detect would-be cheaters, there’s a sort of arms race among community diehards to find the minutiae that proves that suspected runs are illegitimate. “The thing is, when you splice a speedrun, you can do it perfectly, but you might still get caught,” Retro said. “You can go in there frame-by-frame, make sure there’s no artifacts or discrepancies, and it might get accepted for a time. But you run the risk that somebody might discover something about the game that no one knows about that proves that your run is faked. That’s exactly what happened to Anti and a bunch of other runners.”
Anything for clout
The methods of cheating aside, the burning question still remains: why would someone go so far to gain an unfair advantage in an activity that only a few dozen people compete in? Though some point to the sizable revenue that top runners can bring in with their daily streams or the thousands of views that GDQ events garner every year, the runners themselves suggest a far more basic motive: social esteem.
“Getting into the mind of another person is very hard, and it’s mostly speculation,” Retro says. “But, for me, the vast majority of the cases I’ve seen, it’s that for these runners, speedrunning is a huge part of their life, and their whole social circle is centered around speedrunning. They’re so deep in the community that it’s almost the entirety of their social interaction. So when they get a good [personal best] or world-record, it boosts their standing in the community, and I think that’s why people do it. They don’t want to hold the record, they just want to feel like somebody. I think it’s kind of sad, in a way. I think a lot of people who cheat just don’t have a lot else going on in their lives.”
As speedrunning grows more public, however, these YouTubers fear that superior technology will soon win out over the devious art of splicing. Both Retro and Smith are particularly wary of an emerging method that they call “TASbotting,” where a player uses an external device to record their controller inputs on an emulator, then calls on the device to play the whole thing back on a real console in the same way that a player piano simulates a human performer. The “TAS” in this case stands for “tool-assisted speedrunning,” which is an entirely distinct form of competition where enthusiasts use save states and other manipulation to create a run that is optimized beyond the realm of what a human can possibly do with a controller. (If you’re familiar with the cinematic masterpiece Real Steel, just imagine that, except instead of robots smashing each other to smithereens, they’re zooming their way through dusty games as quickly as possible.)By playing back these falsified runs on the original hardware, however, it becomes far harder to detect that any manipulation has occurred for video viewers. “It’s basically splicing, except there are no splices to find,” Smith explains. “You just go in, get a run with a decent time, then use save states to optimize, to create the illusion of a single-segment, then string the inputs together and make it look like a world record. The current verification standards just aren’t good enough to catch it. I don’t know anyone who’s gotten caught TASbotting on console, but then again, how would you even get caught? There’s no splices or discrepancies to find. It’s basically a perfect way of cheating.”
As Smith is quick to note, over the years, several runners have been nabbed for this exact thing, trying to pass off tool-assisted runs as the product of their own two hands. However, they were always exposed by the telltale differences that most games display when they’re emulated versus when they’re played on original hardware. In one case, a handful of Super Mario World 2: Yoshi’s Island runners discovered that one of their own was a fraudster during a stream: because his hands were visible on his webcam, one runner eventually figured out that his inputs on the controller didn’t match up with the pixel-perfect on-screen action. “People have this idea that live-streamed runs are automatically real, but they’re not,” Retro says. “You can fake them really easily, if you take the right precautions.”
Rooting out the cheaters
In order to call attention to what they see as a possible scourge of cheating in the community—and, arguably, to boost their own profiles—both Smith and Retro have themselves submitted spliced runs to leaderboards as “tests” to prove that the current rules are too lax. (In the case of Retro, such antics got him banned from Spyro speedrunning entirely.) Over the past few years, Smith in particular has leveraged his considerable YouTube presence to loudly advocate for stricter standards in speedrunning to prevent this sort of blatant dishonesty.
While such measures might vary from game to game—community members have suggested implementing a bot that requires runners to put in a randomly generated string of inputs during dead zones in Super Metroid where the player can do nothing but wait, for example—the most common one championed by these apparent reformers is “handcams,” a runner recording the movement of their hands on the controller as they play. But while the volunteer moderators who actually police these runs agree that requiring these sorts of additional rules would root out the occasional cheater, by and large, they don’t feel that it’s enough of a problem to justify the increased burden on runners.
“This is a hobby and we want as many people to partake as possible,” says “NewanTox,” one of the moderators on the leaderboards for the Spyro series. “This hobby is a lot more fun when there are actually people to compete against, as it’s not all that much fun being #1 on a board with two runners. It’s important to strike a good balance between accessibility and competitive integrity, and I think we’re in a pretty good spot already.”
“Standards should definitely be increased for the whole speedrunning community, [but] handcams just aren’t really that necessary right now. It’s arguable rather they’d really help as much as just being an annoyance,” agrees “GothicLogic,” a moderator on Super Mario 64, the hobby’s most popular game. “I’m not too concerned really with cheating having a big impact on the overall speedrunning community, other than looming disappointment in runners.”
GothicLogic notes that Mario 64 has seen more than its fair share of confirmed cheaters over the years, owing in part to the game’s outsized influence on the speedrunning community in general. (As an illustration of this, when Retro was first learning to speed through the game, he used the record-holder “Akikan” as his model; that winning run was later found to be spliced.) That said, when I suggested that the siloed nature of the speedrunning community allows cheaters to escape scrutiny from game to game—since there is little in the way of a centralized authority for speedrunning, unlike classic arcade high scores and professional sports—both GothicLogic and NewanTox were quick to say that any such initiative would likely run into sharp resistance from the existing community. There’s a fondness surrounding the “grassroots” mentality shared by most runners today.
“If a respected entity, such as speedrun.com, required higher standards, I don’t think it would work,” says GothicLogic. “Every game’s community is different and run by different people that have different standards. They wouldn’t let a greater entity change what they allow, and if it came to it, they would likely just move away from speedrun.com.”
While opinions on solutions may continue to vary, one thing is clear: as long as the community remains fractured in this way, former fraudsters will continue to find respite, even after other games have banned them entirely. For example, “Anti” has now transitioned to Mario 64 running, where they hold a top 10 time in the highly competitive “16-star category.” While plenty of admitted cheaters have apologized and made their way back into the community after a lengthy cooling-off period, some runners feel that these cases represent the weakness of speedrunning today—and its continued vulnerability to all sorts of trickery and dishonesty.
“I just think it’s ridiculous,” Smith says. “I don’t know if global moderators are the answer, it’s a big question. It’s just like the Billy Mitchell situation, and the old-school Twin Galaxies, where you have to wait for people to mature enough to know that fairness actually matters. Anti cheated at GTA, and he’s banned for life, but now he’s a Mario 64 runner? It’s just like, do we have to catch these people at every game you cheat at? If we’ve determined that we can’t trust someone forever within one game, why do we trust them in any game? There just has to be some change, or I think speedrunning is going to lose a lot of credibility in the next few years.”
Steven T. Wright is a writer, critic, and pedant based in the Twin Cities. He almost named a novel after a city in Final Fantasy, but his friends talked him out of it. His previous work on Ars includes "Life in (virtual) pit lane: The war stories of video game car design" and “'The Linux of social media'—How LiveJournal pioneered (then lost) blogging."
"how" - Google News
December 22, 2019 at 10:16PM
https://ift.tt/395oMHG
How the scourge of cheating is changing speedrunning - Ars Technica
"how" - Google News
https://ift.tt/2MfXd3I
Bagikan Berita Ini
0 Response to "How the scourge of cheating is changing speedrunning - Ars Technica"
Post a Comment