You won't believe it! This doomgod has been cheating all along!

[GarrettChan]

I did mention the verification process is very costly and make Classic Doom speedrunning a chore, rather than being fun. Putting up video footages can't 100% solve the problem. On the other side, it seems you don't care about how much damage will this deal what if the run is real.

 

I don't know why you stick to the 4-hour thing and won't let go. It seems ZeroMaster's TNT Nightmare! run has similar difficulty to me with even more randomization and uncertainty. ZeroMaster didn't actually perform in real time (if there's any video, I would love to see), so maybe eventually every Doomgod will have this problem.

Edited August 29, 2018 by GarrettChan

[GoneAway]

1) The vast majority of people don't cheat, so the default assumption should be that a run is valid. What's the worst case scenario? Somebody gets free e-fame with a cheated run? That rare possibility is worth the openness of the community and ease of entry for new players that is provided by our standards and the beauty of the demo format.

 

2) Pretty much all serious discussion about doom speedrunning goes on in the discord (including about this thread). I admittedly don't take as much care when posting here because the real talk already happens elsewhere. We have in the past and will continue to have discussions about suspicious runs in there (important distinction being discussion, not throwing accusations).

 

3) What makes a run suspicious? Typically things like "this player just started, how are they so good?" or "wow, the quality of this player's runs skyrocketed recently" or "there are a lot of strange moments in the run that look like segmenting is going on". There has been a lot of serious discussion about some players in the past couple years, so I think for you it's just a matter of not seeing it actually taking place. In every case I've seen it was decided there just isn't enough clear evidence of misdeeds (though obviously some people may feel more strongly than others).

 

[Player177]

It looks like you have to try hard to get caught :)

[geo]

1 hour ago, Player177 said:

It looks like you have to try hard to get caught :)

Took Saturday Night Live like 25+ years to finally catch someone lip syncing... which also revealed the real mind blower, that it's live to tape.

[Liberation]

Innocent till proven guilty surely? 

[Grazza]

23 hours ago, Memfis said:

In the chess world you can get like a one year ban from all tournaments for "raising unfounded suspicions". The result, of course, is that there is obviously some cheating going on even at very high levels, but it's too scary to bring up and so nothing really changes.

Wrong on many counts.

Top-level events are very tightly regulated and anti-cheating measures are very strictly applied. No electronic devices allowed in the playing area, everyone (and everything) entering it (spectators, players and officials) are swept every time they enter, and players aren't allowed to leave designated areas during play. Yes, measures can be taken in the case of totally unfounded allegations, and rightly so, as these are easy to make and highly damaging (actually, the fact that you have been misled is additional evidence of this). Indeed, there is a clear similarity with what Vorpal did here. When allegations have been backed up by the slightest amount of evidence or based on a well-founded suspicion, they are taken seriously. There is very good reason to believe that there is no cheating in current top-level events. The last vaguely credible allegation dates back to 2005, in fact.

[GarrettChan]

18 minutes ago, Grazza said:

(Anti-cheating)

I'm curious how to cheat in chess. The closest thing I can think of is there's someone or a computer helping you out to calculate the best move right now. Just want to know, maybe there's something more besides this?

[Grazza]

The crudest way is to have an electronic device on your person (and some way to get its information to you - via buzzing or a concealed earpiece) or concealed in a location you can access (e.g. a toilet). Pretty easy to detect. Occasionally this gets used in open tournaments where there is less security than in high-level events - though those with big prize money may nevertheless have a fair degree of security. Suspicions easily get aroused, as it is generally employed crudely. A player who suddenly has a huge leap in playing standard and plays super-GM+ moves without much thought or being able to explain them adequately will immediately attract attention.

 

More subtle methods are possible. Rather than typing a bunch of text, I'll just quote from my Mammoth Book of Chess (3rd edition):

 

But in higher-level chess, the great increase in strength of the engines could potentially make cheating of a less blatant form effective. First we should note that in games at top level, the players do not need a lot of assistance in order to achieve a large increase in effective playing strength. For instance, World Champion Vishy Anand stated that, if he were told at just three points in each game that he had a strong possibility in the current position (not what it is – only that one exists), then his results would improve to the tune of 100 Elo rating points – that’s the difference between a World Championship Candidate and a dominant World Champion. And to convey that information to the player would require just a nod or a wink – no need for any electronics in the playing hall. And three times a game would not arouse suspicion. One possible method for providing much more than that level of assistance is as follows. This has in fact been claimed to have been employed at top level, but not conclusively proven in terms of documented, publicly-presented evidence.
Three people are needed, at least two of them very strong grandmasters. The first is of course the player at the board. The second is some distance away, and not in a public location (possibly in a hotel room). He is working with a computer the whole time, looking at possibilities in the current position of the game going on in the tournament hall (the moves may be available online). The third man is the go-between, who spends some time in the tournament hall, and some time outside the playing area, communicating with the computer man (e.g. by mobile phone). The go-between may need to tell the computer man the moves that have been played, in case the online coverage is unreliable (or has a deliberate time delay in order to hinder cheating). The computer man is an expert in computer-assisted analysis, who has also worked closely with the player at the board, and is therefore familiar with his playing style. At crucial moments, the go-between will signal to the player at the board based on the conclusions of the computer man’s assisted analysis.
Does this sound far-fetched? I don’t think so. Would it be easy to stamp out? Yes, but not without fairly draconian restrictions on the players and/or spectators. In a World Championship match, you can employ methods to defeat this type of scheme (making sure the players can never see the spectators, for instance, while jamming all telecommunications in the area is another possibility), but that depends on the organizers (and often the match will be held in the home country of one of the players, so allegations of partiality can come into play).
But as I mentioned earlier, allegations of cheating are as much of a problem as the possibility of cheating itself. Most readers will no doubt remember “Toiletgate”, which nearly overshadowed the Kramnik-Topalov World Championship reunification match in 2006, and came within an inch of sinking the entire reunification process. After a bad start to the match, the Topalov team made some obviously absurd allegations against Kramnik, based on the fact that he was visiting the toilet fairly often. How exactly this enabled him to cheat, given that the players and whole area had been swept for electronic devices, was never fully explained. But the match officials gave the allegations credence when they responded by putting restrictions on Kramnik for further games. Naturally incensed by this, Kramnik refused to start the next game. The whole sorry tale is recounted elsewhere, so I’ll spare my readers a blow-by-blow account, except to mention that the Topalov team later used a common ploy in allegations of cheating. They compared Kramnik’s choices of move with those of a top engine, and found that they often agreed. Some journalists even found this quite compelling. However, in many positions there is a clearly best move, and any strong player is likely to choose it, whether they be carbon- or silicon-based. If you ran such a comparison on the Capablanca-Alekhine match games from 1927, you’d also find a fair degree of correlation. If you are determined to make such a claim, then you can bias the figures even more by a little trick. Rather than giving the computer a set amount of time on each move, and then seeing how many of its choices match, you can instead set the computer thinking, and as soon as, at any point within the next, say, ten minutes, the computer agrees with the human’s choice, mark this down as “success” and move on to the next move. Or you could just make the results up in fact, since no one will be able to replicate your set-up exactly and so disprove your “analysis”.
Another type of false allegation of cheating can arise when a player is disgruntled after a bad loss. All chess-players know how rotten it can feel to lose a game, and how it takes some self-control to react gracefully. Sadly, some players see fit to strike back with a farcical allegation. One recent example involved one of the world’s finest young grandmasters, who claimed his opponent (also a grandmaster) must have used computer assistance during the game. His opponent had introduced a new move in this game, in an opening that he had played before and so was well-prepared for. The novelty proved successful, and led to a very quick victory. In the course of these few remaining moves (none of them very complex for grandmasters), the player who was being accused of cheating had in fact missed a much stronger possibility that any engine would have pointed out very quickly. Sadly, the accuser did not withdraw his allegation once he had got over his loss, but even added to it by the familiar method of comparing his opponent’s earlier games with a computer’s choices. The clinching argument, apparently, was that the player then went on to perform badly in the following rounds. No wonder – he had been called a cheat in front of the whole chess world, and knowing that some mud will always stick, was no doubt concerned about the effect this would have on his future career.
It would be a sad state of affairs if chess tournaments in the future are to require airport-style security checks, but this might be what is needed to protect the innocent. But with ever-greater miniaturization of electronic devices, even that might not be sufficient. There have recently (2013) been a number of cases where a player had a suspiciously large and sudden leap in playing strength and an equally instant change to a computer-like playing style, but no obvious means of electronic assistance were uncovered, even when the matter was investigated quite closely.

I wrote that in 2009, with some updating for the electronic edition in 2013.

[Kaido]

Prboom+/glboom+ has a 'show progress bar during demo playback' feature , I suggest before the stream begins you show your audience that this option is enabled , this prevents any way of possible demo playback as any demos would have the progress bar shown in the bottom. And ensure that your streaming software extends all the way from the  to the name of your sourceport you are using so people dont switch between sourceport or use an encoded video(god knows who does that) . I also recommend setting and turning on your smooth demo playback to a higher number like 6 or 7 .

[GoneAway]

You could just launch it with a different config when you play back the demo, so that wouldn't make a difference.

[Buckshot]

I love how the title of this post starts with an almost definitive and damning accusation to the point of certainty, but the post itself is completely devoid of any reasonable or irrefutable evidence with literally nothing more than a line implying petty suspicion.

OP would make a great author of some tantalizing clickbait.
 

[Pirx]

Seriously,  some proposed security measures would make recording a chore.

 

I understand that in world class chess there's money at the stake, but in freely downloadable demos for a 25 years old game? E-fame? Wow.

 

BTW,  very interesting post @Grazza

 

Perhaps this will require splitting threads,  but what's a computer -like playing style in chess? 

[Vorpal]

I didn't create the title, and I'm wrong to use the word "probably" which I tried to walk back, but I won't edit posts to change their tone because then anybody coming in late will have trouble following. I'm not wrong in having a suspicion though... the arrow that my suspicion points in can be wrong, but the act of having it is fine and I reject the idea that only experts can have opinions. However insulting as I was though, some thin skins in here, like I don't attack a person if they shit on a map I liked or spent 500 hours working on or walk into the sunset over it. A good flamewar is the oldest legacy of doomworld, lol

 

I think kraflab cut through the bullshit though and gave me what I was looking for, that runs are presumed legit in this community because webcams are too big of an ask for speedrunners, which I think is odd and unique among speedrunning/hiscore communities. I can consider it case closed there, or rather, case nobody cares. [edit for clarity: that nobody cares about verification]

[Grazza]

Pirx: Please note that I wasn't in any way suggesting that anti-cheating measures are needed for Doom. It's just that chess had been brought into it, and I felt the need to set the record straight. There is also a relevant parallel with the ease with which farcical allegations can be made and in terms of the immense harm they do.

 

Actually, just yesterday we had a prime example of the type of move that would give a crude cheat away. This was with Caruana's Ra2!! in his Chess Tour play-off game against So. To a weak player using a computer, it isn't obvious that this is a hard move to find. A computer finds it very quickly, because it thinks in a very different way from a human. To a human it is immensely hard to find. Few players in the world might have spotted it. World no. 2 Caruana nearly didn't. I don't know if Wesley had seen it until it was played (I wasn't in Saint Louis on this occasion). And yet a clumsy computer cheat would probably have played it instantly, and thought it must be an obvious move.

 

Computer-like "style" isn't so relevant, because they play so far above human level nowadays. But they actually vary quite a lot.

 

Vorpal: No, case isn't closed. You ignorantly waded in, dropped a huge steaming dog turd on the carpet, and then asked "Can't we all discuss this calmly?" It doesn't work like that. It's the public nature of it that is so harmful. In the past, when there were legitimate concerns about actually suspicious demos, discussion and detailed analysis took place behind the scenes before any public statements were made in those cases where it seemed overwhelmingly necessary. Once public, there is no undoing these things.

Edited August 30, 2018 by Grazza : minor word order fix

[GarrettChan]

Grazza: Thanks for the quotation from your book. It's very interesting to read even I only know the rules and some very basic strategies of Chess.

 

Vorpal: Do you mind explain why you only stick to this demo rather than other "extreme" demos? If you insist in verifying one demo just for your site, you should probably verify all of those records you put on your site. I don't understand why you don't care about other records. Maybe a 5-minute demo requires more time than a 4-hour demo.

[Bauul]

Can anyone link or outline examples of actual speed-running cheating techniques to compare to?  Either for-demonstration-purposes-only or actual cases of legitimate cheating.

 

I'm quite ignorant about speed-running, so I'd be fascinated to know the things that one would look for to identify some kind of cheating.  Like what was it about Ancalagon's run that specifically made it look like it was tool-assisted?

[leodoom85]

Just now, Bauul said:

Can anyone link or outline examples of actual speed-running cheating techniques to compare to?  Either for-demonstration-purposes-only or actual cases of legitimate cheating.

 

I'm quite noobish at speed-running, so I'd be fascinated to know the things that one would look for to identify some kind of cheating.  Like what was it about Ancalagon's run that specifically made it look like it was tool-assisted?

Here we have a future cheater :D

[Bauul]

1 minute ago, leodoom85 said:

Here we have a future cheater :D

 

Curses!  And I would have gotten away with it if it wasn't for you meddling kids!

[Nine Inch Heels]

6 hours ago, Vorpal said:

why ARENT you guys concerned about proof? (thus far, runner past-history is the metric currently carrying street cred)

AV D2ALL was on the table every so often, and there's enough demos available to watch and learn from. Never at any point has there been a discussion about whether or not the run would be "humanly possible" as far as I remember. That said, it's not like Anc's run came entirely out of nowhere, even if it might seem that way for you.

6 hours ago, Vorpal said:

or what is proof? 

or is proof relevant?

Clearly the better question here is whether or not proof is relevant, because if people wanted they could just switch to a port that doesn't officially support TAS features and run that on a virtual machine, which would quite literally mean people could turn back time even with GZDoom, for example, willy-nilly, which would be undetectable, and it most likely wouldn't be that difficult to do for somebody who wants it that bad, either. Quintessentially, you cannot prevent people from cheating in some way shape or form unless they're streaming live, and even then there's ways to fool the audience into thinking you're playing live when actually you're "finger-syncing".

 

What matters the most is who performed the run, and how it has been executed (taking into account the runner's skill, and the runner's past achievements, not just the AV ones, no, all of them. So the one thing you yourself say you can not/dare not judge is actually the only important thing here, and in the light of this, you would have done yourself and everybody involved a huge favour if you had discussed this "behind closed doors" first.

 

6 hours ago, Vorpal said:

Is the default position to assume all runs are legit?

The default position is of course to assume that all runs are legit, unless there is a genuinely good reason to doubt a run's authenticity, and circumstances such as source port used, hand-cam or not, are never ever sufficient to label a demo cheated or "probably TAS", especially not officially. Basically all of the circumstances you gave out as "reasons" apply to way over 90% of all the demos ever recorded for doom, which should have been a strong enough signal for you to not just jump to "conclusions" that imply "foul play".

 

With all that in mind, the notion that each and every attempt needs to be streamed live, or somehow verified in order for it to not be considered "fake" is slightly misguided. If this was a "high profile game" like Ocarina Of Time, or Mario 64, that didn't have what feels like a million-and-then-some WADs to choose from, then maybe you could justify making significant efforts to detect the supposed cheaters when you have a world record demo coming in, but with that many WADs being made and demos coming in you're better off accepting the possibility that somebody could cheat rather than going on a ghost chase.

 

Let's not forget that without Doom's "open door policy" WRT to speedrunning you maybe wouldn't even have as many AV demos to watch to begin with.

[Looper]

10 minutes ago, Bauul said:

Can anyone link or outline examples of actual speed-running cheating techniques to compare to?  Either for-demonstration-purposes-only or actual cases of legitimate cheating.

Just check any Doom 2 Nightmare demo from Steffen Winterfeldt to see how a noob cheater plays.

[Vorpal]

7 minutes ago, GarrettChan said:

Vorpal: Do you mind explain why you only stick to this demo rather than other "extreme" demos? If you insist in verifying one demo just for your site, you should probably verify all of those records you put on your site. I don't understand why you don't care about other records. Maybe a 5-minute demo requires more time than a 4-hour demo.

 

There were like 5 minor bullet points earlier that gave me pause, and combined together triggered a kneejerk reaction. The demo at this point is hardly the point of discussion, so much as the norms and procedures of the doom speedrun community, which y'all are impressing upon me inbetween our exchanges of venom

[GarrettChan]

8 minutes ago, Vorpal said:

 

There were like 5 minor bullet points earlier that gave me pause, and combined together triggered a kneejerk reaction. The demo at this point is hardly the point of discussion, so much as the norms and procedures of the doom speedrun community, which y'all are impressing upon me inbetween our exchanges of venom

Oh, did I insult you out of nothing? If so, I apologize for that. I totally don't understand what you meant here. Probably because I'm a non native speaker.

 

I still stick to how to get an effective proof from the start, but the thing is that there's no such a way to do it.

[Dragonfly]

@Bauul

 

This video isn't relevant to Doom speedrunning specifically, but speedrunning as a whole.

 

 

Edited August 29, 2018 by Dragonfly

[Memfis]

Hm, I wonder what Grazza would say about articles like this:

http://ruchess.ru/en/news/report/cheating_part_1/

http://ruchess.ru/en/news/report/cheater_part_2/

Or:

https://chess24.com/en/read/news/morozevich-on-go-computers-and-cheating

 

To me the idea that suspicions should be kept private for the most part sounds strange. Logically, it seems like the best way to determine truth is to have a honest open discussion where nobody has to hold back. But I'm probably underestimating the power of uninformed public to spread false rumors and ruin people's careers.

[Nine Inch Heels]

2 minutes ago, Memfis said:

Logically, it seems like the best way to determine truth is to have a honest open discussion where nobody has to hold back.

Except this isn't what happened, because Vorpal simply decided to "Leeroy Jenkins" this, slap a label on the demo, visible on an "official site", and then the talking began.

[Vorpal]

52 minutes ago, GarrettChan said:

Oh, did I insult you out of nothing? If so, I apologize for that. I totally don't understand what you meant here. Probably because I'm a non native speaker.

 

I still stick to how to get an effective proof from the start, but the thing is that there's no such a way to do it.

 

No you didn't insult me, by "our" I mean collectively in this discussion, myself and others have been slinging shit at eachother.

 

25 minutes ago, Memfis said:

To me the idea that suspicions should be kept private for the most part sounds strange. Logically, it seems like the best way to determine truth is to have a honest open discussion where nobody has to hold back. 

 

Yeah this is my "default position". A private tribunal with no standards creeps me out and changes my view of doom speedrunning in general from "grand competition" to "mild curiosity".

[GoneAway]

Today I learned open community discussion and consensus equals private tribunal 🤔

 

The doubling down on stupidity is mind boggling.

[Nine Inch Heels]

6 minutes ago, Vorpal said:

Yeah this is my "default position". A private tribunal with no standards creeps me out and changes my view of doom speedrunning in general from "grand competition" to "mild curiosity".

Just because you don't know what those standards are doesn't mean that standards do not exist, and if you had the decency to discuss this with the runner in question as well as getting some opinions from other people before publically discrediting an achievement, we wouldn't be sitting here right now.

 

That is not a difficult line to draw: Discuss first, voice your concerns with a demo, then put a label on a demo if you still think it needs one. Putting a label on a demo first, and then "discussing" it is the wrong way to go.

[4shockblast]

I should point out that despite Ancalagon's request earlier to take down the demo link from the AV site, this has not occurred, and the words "probably TAS" were replaced with "unverified," which is practically equivalently insulting compared to the previous note as it again singles out this demo for no apparent reason (who verified the other demos anyway?). I have made these requests known privately, but I would like to also indicate in this thread that I would prefer that my demos are removed from the site as well. This situation is utterly ridiculous, and I do not wish to be associated with this shit.

[Vorpal]

Unverified is truth though. Anyway I don't really see the page having much of a future, especially if takedown demands are a thing. For now I'll just leave it as a scarlet letter because I'm in a capricious mood after baring my honest opinions