Mortal Kombat 3 moderators may be abusing their power
5 years ago
Rio Grande do Sul, Brazil

Greetings,

The Mortal Kombat threads may have been a little headache for the speedrun.com website administrators in the last few months, but I believe that something quite suspicious is happening over there. One of the best brazillian Mortal Kombat 3 players that I know had submitted some recordings of his speedruns (some of them were live streams) in the past few months, and those had been validated by the website, but now we can see that they were deleted. According to the information that I gathered, the justification for deleting his runs was that "his time is too good and even though it can't be proved that it is faked, it must have been because we don't believe it", or something like that, and that was it. Now, if that isn't alone a reason to justify some kind of investigation on this subject, now the world record belongs to the thread moderator/admin of the Mortal Kombat 3 thread. Every time I think about it, it just looks more and more like another Todd Rodgers situation. I believe that an investigation of the situation would be of the best interest for the speedrunning comunity.

Best regards.

Canada

"the justification for deleting his runs was that "his time is too good and even though it can't be proved that it is faked, it must have been because we don't believe it""

I don't know a whole lot about this particular situation, but generally speaking if the moderators have reason to suspect that a run is fake and can't trust it, then it's fine for them to reject that run, even if they don't have definitive proof. In most cases it's nearly impossible to definitively prove that a run is fake (you can only really be certain if someone confesses or if there's some obvious discrepancy), so there has to be some middle ground.

"the world record belongs to the thread moderator/admin of the Mortal Kombat 3 thread"

In some situations this would raise some eyebrows, but in this case there are two other moderators who, presumably, agreed to this ruling. If you really think the one moderator is abusing their power, you should raise these concerns with the other two mods (assuming you haven't already).

"it just looks more and more like another Todd Rodgers situation"

I don't really see any resemblance between this situation and Todd Rodgers, but whatever.

Edited by the author 5 years ago
United Kingdom

Rulings on emulator were made far more strict, due to a repeated number of false/fake speedruns being submitted to the game (and the series). Several runs were then rejected from the leaderboard to uphold the now current rulings based on emulators.

Looking at the logs, that reasoning for a rejection has never been used on MK3.

Edited by the author 5 years ago
Daravae, Imaproshaman and 2 others like this
Antarctica

As a side note this is a bad, very bad, precedent to set

[quote]if the moderators have reason to suspect that a run is fake and can't trust it, then it's fine for them to reject that run, even if they don't have definitive proof.[/quote] Rejecting a run as cheated without proof should get you automatically removed as a moderator in my opinion. You can’t just reject a run because you think it might be cheated or faked. You need to have some form of proof otherwise you could have mods rejecting runs for the hell of it and claiming “suspicion”. If a mod thinks a run needs deeper looking into, it’s fine to convey that to the person so they can maybe clearer up some confusion in the submission, but you can’t just reject it.

Edited by the author 5 years ago
Eggwink, Imaproshaman and 3 others like this
Valhalla

I could make a fake run right now, tell you it's fake, but because I faked it so well you would accept it?

Imaproshaman likes this
Antarctica

No because you telling me it’s fake is the proof I need to reject it.

That was an awful example.

Valhalla

LMAO? So I could submit that same fake run without you knowing it's fake and you would accept it, right? Cause I could fake a run so well, you wouldn't know outside of me giving you that information.

Some games are simple enough to where faking them is stupid easy. Where the only difference between TAS and RTA is how skilled you are.

So how do you catch the fakes? Get it now?

Rio Grande do Sul, Brazil

@Liv I understand your pointing, but when new rules are applied by the moderators and a moderator gets to be number one on the ladder right after these changes on the rules, I do believe that the credibility on the website is, at least, in check. I'm not saying that the moderators DID change the rules for the benefit of one or more of them, but I am saying that this attitude should be at least investigated.

United Kingdom

I think, if anything, their rulings on emulators were just far too lax to begin with. Things like input viewing/capture, capturing the entire emulator window and specific emulators being required should be things most games take into precedent to begin with. If there were ever such a thing as "site-wide rules", these are things I would be heavily in favour of for strictly PC-based emulator submissions.

It's far too easy to fake emulator runs unfortunately, and without really any way to confirm they are fake. A lot of the methods used to detect cheating on console simply will not work, along with an array of tools that are specific to emulator allowing you to get away with it far easier.

So, the only way to really make it more difficult to get away with, is increase the amount of information given to the mods and / or require emulators they are both familiar with and that they know plays the game is a very defined way.

Edited by the author 5 years ago
Antarctica

If I couldn’t tell a run was faked or cheated, then yeah, I’d accept it because what grounds do I have to reject it? If someone asked why I rejected their run, what do I tell them? It looked too good? That makes zero sense and is completely unfair.

The difference between a lot of games and TAS is how skilled you are - this is especially true for older games. In some games it’s easy to catch TAS like tricks, but in others it’s much harder. Catching skilled cheaters is always hard and will always be hard, that’s why spliced runs went uncaught for like a decade in some games like SM64. However people who are that skilled and out that much time into it are few and far between.

Cheating will always happen in something like this, however most times it’ll be easy to spot. But on rare occasions you’ll get someone who does it very good and gets always with it for some time. You can do everything you possibly can to catch it and stop it and that’s what matters. Nobody expects mods to catch every cheated run because realistically that’s not possible if someone dedicated years to faking a run and making it look real. This is also why certain games ban various emulators to prevent TAS tools from being available and considering how easy faking can be on emu.

But if you truly think that rejecting a run for being good, almost too good, with no other proof is acceptable then idk what to tell you. So instead I’ll give you an analogy - imagine a teacher giving a zero on a report because it was plagiarized. When they’re asked what part was plagiarized, they say “I don’t know, but this report was so good I felt it had to be plagiarized so you get a zero”. They can’t do that, you can’t punish someone for something you think is true without proving it.

Get it now?

Edited by the author 5 years ago
Salem, diggity, and Imaproshaman like this
Valhalla

Yeah I wasn't talking rejecting a run for being too good, just pointing out that it's not that simple. Sometimes you have to make the decision to reject a run because you suspect it very well could be cheated.

Otherwise on games that are easy to cheat, the verification process should just be removed entirely.

MASH likes this
Canada

Well, this exploded.

I fully understand where you're coming from @Timmiluvs , ideally you'd be able to definitively prove that any fake runs are fake and no runs would be rejected without proof, but as @Komrade is saying it's not that simple. Example, if a brand new user with no reputation and no runs submits a new world record in a very optimized/contested game that barely meets the proof standards, you just know that run isn't real, even if you can't necessarily spot the exact location where they spliced it (or whatever cheating method was used). So do you still verify that run and have it displayed as the world record just because there's no proof, even though everyone knows it's fake and isn't actually the world record?

That's a pretty extreme example and I don't imagine that happens very often, but nonetheless, that's a situation where I'd say reasonable suspicion is good enough. What exactly constitutes reasonable suspicion and where exactly the line should be drawn is a hard question to answer, and is certainly up for debate, but it's somewhere before 100% certainty.

Edited by the author 5 years ago
Bulgaria

Proof? What kind of proof can you possibly get if someone used something as simple as a macro to help them with one skip for example. If a run seems suspicions to the mods and the community then it is on the runner to prove that it legitimate or it gets rejected to all hell. How can you say "automatically removed as a moderator" with the following red flags that are common in some of these faked runs: doesn't stream, doesn't interact with the community, plays on emulator, seemingly amateur gameplay with the exception of important hard tricks that happen to be done perfectly, is confrontational or makes excuses when asked to provide proof. Yea.... no.

SkyangelPR and MASH like this
New York, USA

I'm a firm believer in "innocent until proven guilty."

Salem likes this
Canada

So if you're presented with a blatantly fake WR submission but you can't find the splice or get the person to confess or anything you're just gonna go "oh well, can't prove it, better list it as WR even though everyone knows its fake"?

It's really easy to say what you're saying, but again, it's not that simple.

TalicZealot and Hako like this
United States

If you can't find how it was cheated then it's not blatantly fake is it?

Salem likes this
United Kingdom

I mean in the case of emulator there wouldn't always be tangible proof. That's why it's smart to buckle down on emulator as much as possible.

It's very easy to poke RAM in emulator and anyone can do it, for example, and this can provide advantageous results that wouldn't necessarily be able to be identified as cheated. If someone got say a 1% chance drop in their attempts over the course of say... 20/30 attempts every single time, would you be the least bit suspicious and require a proofcall, or just pass it off as absurdly good luck? In that particular situation you cannot prove anything.

Edited by the author 5 years ago
Valhalla

A proof check is great, and probably a really fast way to tell if someone is being disingenuous. Hell I'd provide people with the equipment to help prove their run is legit.

TalicZealot likes this
England

Generally speaking, I don't see an issue with just asking runners suspected of foul play to provide a comprehensive rundown of how their run came to be.

If we're talking in terms of, "The run provided is theoretically attainable ingame, but very unlikely in this context" it's often very telling one way or the other when the runner explains the process of how it came about. If they can only provide details in the vaguest of terms, "Well it just happened, I got really lucky, etc" this tends to be cause for reasonable suspicion. Put yourself in the position of the runner whose gameplay is in question. If you're someone who has come relatively out of nowhere with high-level runs, which is often the case with suspected cheating cases (Because how would a 'random' be able to play at WR level??) then you could reasonably expect that the way they achieved such a run is through illegitimate means, whilst not having the very intimate understanding of how the game operates, which is naturally gained through the long process of learning about the game. Thus, they will not be able to articulate exactly what parameters are at work in the run very effectively.

There's already precedent for this in certain communities. My understanding of Exarion's then-meteoric rise in Pokemon speedrunning is that when he was an up-and-coming runner, it was considered very suspicious. He was very closely scrutinised and asked to provide extremely specific details of what was going on in his runs, and he was able to follow through completely and provide any and all information the moderators wanted.

This also serves the dual purpose of, if someone really doesn't have the intimate understanding of the game, and actually has manipulated the game in a way that's not normally possible, then in the attempt of providing detailed information, they will inadvertantly out themselves. It's a win-win situation in my opinion.

drgrumble and ShikenNuggets like this