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

[Cynical]

13 hours ago, rdwpa said:

- Point-blanking those cybs seems believable in a D2ALL. Stationary turrets generally fire way less because of a quirk of the engine (monsters changing direction frequently resets the 'time to fire' counter more often). Plus, and a lot more importantly, you have 200/200 going in and a mega right behind you too, so you can survive two rockets even if the cyb manages to get that many off. I'd be surprised if this was 100% reliable, but it strikes me as 'believable'. 

I'm at work and so can't watch the demo/movie to see it for myself, but is the part in question the bit where you teleport to the upper towers and instantly blast the Cybs?  If so, that's nearly 100%, if you take one rocket you'll immediately restore your health on the megasphere and kill him with the second BFG blast, and that's assuming you need two BFG shots to get the kill, it's not uncommon for them to end up infighting with the nobles and lose a bunch of health before you get there.

[rehelekretep]

1 hour ago, j4rio said:

You know it's bad if, out of all people, it's dew to tell somebody to calm down.

i already liked this but damn lol

[Veinen]

15 hours ago, Altima said:

Second, reiterating the mention of PRBoom+, it's very easy to connect runs. Considering that AV is a vanilla-compatible megawad (and also the fact he didn't use the AVMOVFIX.wad file which I'm assuming is because he used a source port), why not use vanilla for legitimacy's sake?

 

Ancalagon using prboom+ is hardly strange or suspicious. Look at his demo page at DSDA http://doomedsda.us/player383lmps.html: that's 459 demos, including several for vanilla-compatible wads, and a grand total of 0 demos have been recorded with vanilla. Prboom+ is obviously his default engine and the one he is most comfortable with. And since I'm pretty sure he didn't anticipate this shitstorm (I don't recall anyone voicing similar doubts over any of his other demos for vanilla wads...), then what reason does he have to switch to vanilla just for this run? As far as I can tell everyone has considered his prboom runs perfectly legit until now, and clearly the vast majority still sees that way (including me), so why would he have suddenly felt the need to switch to vanilla to "prove the legitimacy" of this particular demo?

[Bob9001]

Ok, lets be real here; We know he didn't cheat so why are you still arguing your point. You're wrong.

[tchkb]

You won't believe it! This accused cheater has been a doomgod all along!

 

Now seriously, does any video footage of that Ancalagon's Chillax run exist? I'd love to watch at least a fragment of it.

[Altima]

10 hours ago, Veinen said:

 

Ancalagon using prboom+ is hardly strange or suspicious.

It's not about him using PRBoom+ (I've used it myself on a few runs), it's about him using PRBoom+ for a run like this because of the circumstances behind it, plus PRBoom+'s general ability to easily be cheated with. This whole thing is just a perfect storm of things that happened. Anc has a run where it's 4+ hours of super aggressive play without any signs of fatigue at all in the run (when a similarly skilled player, Zero Master, was showing a little fatigue and wasn't looking as fresh going into the later maps of his av30 run), never once does he put safety before efficiency, and while there have been signs he did practice it, there weren't any obvious signs he was building up to it. There's no Episode 2 or 3 UV Max movie runs for AV by him. That's typically the normal sequence of events. Runner does E1 UV-Max > Runner does E2 UV-Max > Runner does E3 UV-Max > Runner does full-game UV-Max. Not saying this is always the case but usually when they go for the full game, they at least post up their work of the individual 3 episodes.

I'm not trying to discredit Anc here. He obviously had great route planning, he has spectacular BFG usage, great timings, all of that. There's just things that make me skeptical of it's full legitimacy. As I previously stated, I *do* believe his run is legitimate but I am still going to ask these questions and see what sort of answers I get that might help me personally fill in the gaps as to how this run came to be.

One thing I will ask though, this is more of a curiosity than any sort of accusations. Which is truly the more legitimate way of doing a 30AV Max run? Vanilla where you're forced to use AVMOVFIX.wad which removes a portion of Map20, or through a source port (PRBoom+ in this case) where you don't need AVMOVFIX.wad and can do the portion of Map20 that is removed? Some food for thought I'd love to hear opinions on.

Editing just to clarify that I am only speaking about this run. I have 0 doubt of any of Ancalagon's achievements in other wad files. This is solely about this and I'm merely seeking answers to my own suspicions.

[Da Werecat]

Well, you do it with AVMOVFIX.wad in PrBoom+, remove the footer and claim it was recorded in vanilla. Both convenient and trustworthy.

[Altima]

6 minutes ago, Da Werecat said:

Well, you do it with AVMOVFIX.wad in PrBoom+, remove the footer and claim it was recorded in vanilla. Both convenient and trustworthy.

You. I like your answer. Cheeky :P

[GoneAway]

2 hours ago, Altima said:

...there weren't any obvious signs he was building up to it. There's no Episode 2 or 3 UV Max movie runs for AV by him. That's typically the normal sequence of events. Runner does E1 UV-Max > Runner does E2 UV-Max > Runner does E3 UV-Max > Runner does full-game UV-Max. Not saying this is always the case but usually when they go for the full game, they at least post up their work of the individual 3 episodes.

Are runners supposed to film their lives and keep a diary as well now so we have documented evidence of their intent? 😛

 

I'm curious where you're getting your stats for e2/e3 runs from. For one thing, hardly any d2all runs are done, so you can't have much sample size. How many e2/e3 movies are there in doom 2 for instance, compared to d2all? Most people consider e2/e3 runs to be memes (because they are...).

 

Edit, archi found this in the discord logs, hehe (months earlier):

[Archi]

2 hours ago, Altima said:

That's typically the normal sequence of events. Runner does E1 UV-Max > Runner does E2 UV-Max > Runner does E3 UV-Max > Runner does full-game UV-Max. Not saying this is always the case but usually when they go for the full game, they at least post up their work of the individual 3 episodes.

No.

 

What's the point of doing separate runs if the goal is to do d2all run? It's not even a hard goal, just long.

[4shockblast]

Pointing to PrBoom+ usage as suspicious doesn't make any sense. The vast majority of runners use it right now, and generally, switching ports sucks for any particular runner because of different mouse sensitivity and other settings. It's trivial to fake a PrBoom+ demo to look like it was recorded in vanilla, so it's just as easy to cheat by recording a segmented pr+ demo and claiming it's vanilla. Most active runners and viewers practically don't care either way whether a demo is recorded in PrBoom+ or vanilla, and a good run is an impressive achievement regardless of the port it was recorded in. The source port usage is not relevant here.

[Cynical]

6 hours ago, Altima said:

never once does he put safety before efficiency

Da fuq?  His time is 20% slower than individual level maxes, and that's with carry-overs (which let you shave a lot of time off vs. pistol start).  You don't end up with that if you're not putting a high priority on safety.  Yes, Anc can play very aggressively by "normal human" standards and still be safe, but he's not playing full-force "speed at all costs" in the second half of this.

[Altima]

I find it funny no one bothered to point out I did indeed say "this is not always the case" as this is clearly an example of that. Maybe his time is "20% slower" but there's still no hesitation, no damn brakes on the train even all the way up to Maps 25/26/28.

 

2 hours ago, 4shockblast said:

It's trivial to fake a PrBoom+ demo to look like it was recorded in vanilla, so it's just as easy to cheat by recording a segmented pr+ demo and claiming it's vanilla.

 

But you do agree that it's easy to do, which is the point I was getting at. Not that he did it.

[GoneAway]

Can't play a wad any safer than killing all the enemies. What kind of a speedrun is that? 😩

[4shockblast]

50 minutes ago, Altima said:

But you do agree that it's easy to do, which is the point I was getting at. Not that he did it.

 

What I'm saying is it would be no more or less suspicious than if Ancalagon claimed to have used vanilla because it's easy to fake a PrBoom+ demo as vanilla. If anything, as someone else pointed out in this thread, it would be more suspicious because Ancalagon has no history of using vanilla/Chocolate/etc. for any wads, vanilla-compatible or not.

 

I did watch some of the run, and it seems his play shows familiarity with the wad, which, as mentioned previously, comes from survival sessions, but is far from ultra-aggressive, as his strats are fairly safe. I haven't been able to see any points where it seems unreasonable for him to survive a given fight, but I'm not going to watch the entire demo because I don't have time for that. If there's a particular point in the demo that anyone thinks he does something that seems to require either unreasonable amounts of luck or aggression, then that person should call out that moment for more specific discussion. Otherwise, any further discussion here is entirely pointless.

[Altima]

I guess I'm just of the mindset that people are downplaying his aggressive playstyle for a full 32 maps/4+ hours. My biggest point of questioning at this moment is that Zero Master's 30av run starts to slow down a bit toward the end and you can see a decline in the play whereas Ancalagon's run here is as fresh going into Map01 as going into Map29, there's no sign of wear or tear anywhere. You can only combat RNG in Doom for so long before it's going to do whatever it wants to do no matter how skilled you are. If someone can disprove these points then I will have no further questions.

Again, since it's been a bit, I want to reiterate that I do indeed feel Ancalagon's run is legitimate. That doesn't mean I am without concerns.

One thing I do want to bring up though, Vorpal has a point. We *are* the only speedrunning community that has no verification system or anything of the sort. I speedrun other games and there's always a person on the other end that watches and verifies that the run through the game is a legal run. I realize that this game isn't your Super Metroid, Mario 64, Mega Man, or other high-profile game, but I feel that the point here remains the same. Whether we need something like that or not is questionable. The .lmp format is both great and potentially bad. It's great in the fact that it's a verification system of it's own since it's literally raw game actions dumped into a file format that can be read. However, the fact people can simply *write* a .lmp file can be seen as something that could detract from the legitimacy of the format in itself. Maybe that's what Vorpal is trying to get at in his OTHER point, which I can see being a fair thing to pick out. I realize this is a community of trust but sometimes trust can only go so far before one person ruins it for everyone.

[GoneAway]

35 minutes ago, Altima said:

One thing I do want to bring up though, Vorpal has a point. We *are* the only speedrunning community that has no verification system or anything of the sort. I speedrun other games and there's always a person on the other end that watches and verifies that the run through the game is a legal run. I realize that this game isn't your Super Metroid, Mario 64, Mega Man, or other high-profile game, but I feel that the point here remains the same. Whether we need something like that or not is questionable. The .lmp format is both great and potentially bad. It's great in the fact that it's a verification system of it's own since it's literally raw game actions dumped into a file format that can be read. However, the fact people can simply *write* a .lmp file can be seen as something that could detract from the legitimacy of the format in itself. Maybe that's what Vorpal is trying to get at in his OTHER point, which I can see being a fair thing to pick out. I realize this is a community of trust but sometimes trust can only go so far before one person ruins it for everyone.

It's already been said multiple times in this thread, but again:

 

The lmp is our equivalent of a video. Whereas someone playing another game can splice their video, we can segment the lmp. They have this identical drawback. I'm sure you saw for instance the huge scandal of the super meatboy runner whose run was found to be spliced, which was only detectable due to a quirk of the game, and only noticed years later. A video doesn't somehow miraculously prevent cheating. Likewise, a person checking off each submitted video doesn't guarantee anything. How can you say we have no verification system?

 

The lmp is significantly more powerful than a video though. Let's say I'm speedrunning some fps game, not doom, on my PC, and I want to cheat. Instead of dealing with splicing, I'll just try to make it easier for myself. I edit the game files so the damage rolls are in my favor. That's going to be pretty difficult to notice and impossible to prove. Thanks to not having a demo format and trusting video footage, which might even include a handcam(!), no one questions it.

 

Can you do that with a system that uses demo files? No, you can't. Demo files also allow for heuristic analysis of player inputs, something a video could never achieve.

[Altima]

1 minute ago, kraflab said:

It's already been said multiple times in this thread, but again:

 

The lmp is our equivalent of a video. Whereas someone playing another game can splice their video, we can segment the lmp. They have this identical drawback. I'm sure you saw for instance the huge scandal of the super meatboy runner whose run was found to be spliced, which was only detectable due to a quirk of the game, and only noticed years later. A video doesn't somehow miraculously prevent cheating. Likewise, a person checking off each submitted video doesn't guarantee anything. How can you say we have no verification system?

 

The lmp is significantly more powerful than a video though. Let's say I'm speedrunning some fps game, not doom, on my PC, and I want to cheat. Instead of dealing with splicing, I'll just try to make it easier for myself. I edit the game files so the damage rolls are in my favor. That's going to be pretty difficult to notice and impossible to prove. Thanks to not having a demo format and trusting video footage, which might even include a handcam(!), no one questions it.

 

Can you do that with a system that uses demo files? No, you can't. Demo files also allow for heuristic analysis of player inputs, something a video could never achieve.

I do remember the segmented Super Meat Boy run. Got a lot of attention. I think there was also a cheated Yoshi's Island run recently as well but I could be mistaken on the date, and this player did it on camera during a stream. He only got caught because his finger presses on camera didn't match the streamed game footage and someone with a keen eye spotted it.

By "no" verification I'm obviously referring to someone who watches runs and verifies them with human eyes. That's all I'm referring to here. Even if we did, it's easy to make a convincing looking run that would pass a human's check. 

I don't disagree that the .lmp is great but as with all things, it has it's draw backs. That's why I mentioned it being a double-edged sword when it comes to "no verification". Given enough time and creativity though, anyone can cheat at any particular thing so I guess in the end maybe it makes the whole concept irrelevant? Nothing is fool-proof.

[GoneAway]

Not sure what you mean. All the demos get watched. If you look at the dsda thread you'll see plenty of reports about mistakes / etc coming from the community. If someone posted a demo that was clearly a tas (or otherwise fishy), it would be noticed, unless it was for something so obscure and meaningless no one cared to watch (in which case it's irrelevant anyway).

 

My main point is that a lmp is no worse than a video, and is in fact better. Doom would be easy to splice as video because of the intermission screens. Only specific games would have immunity from this kind of issue (for splicing you'd want periods of no audio and unchanging video, I assume).

 

 

[Archi]

1 hour ago, Altima said:

fresh going into Map01 as going into Map29

I've watched map32, and it's not as fast as map01. Retreating and losing like 5 seconds to get mega? That's not Ancalagon I know. Just compare map32 in the demo with any of his relatively recent IL demo, like italo doom or dark tartarus.

 

1 hour ago, Altima said:

there's no sign of wear or tear anywhere

If ZM became tired at the end of his demo, everyone should? Alright, I've done an ironman run of perdition's gate almost an year ago, it took me more than 3 hours and I had no problems with fatigue, if only a little, but you will not see it on demo. And before anyone accuse me in cheating, I did it on stream 2 hours after the announcement.

 

1 hour ago, Altima said:

We *are* the only speedrunning community that has no verification system or anything of the sort

Nah. Also so called verification isn't any better than demo analyze. What do people want usually for clarification, webcams and streaming? Oh yeah, totally the thing every speedrunner wants to do to prove they aren't cheating in their hobby.

 

You know what's funny about verification? Knowing how to determine cheaters in some game gives you info on how to cheat without getting caught. And everyone would be happy with your verified, yet cheated run.

[Doomkid]

Well folks, since the jury is still out on this one, we'll be closing the Speed Demos section and obliterating all previously held records. Furthermore, the very concept of 'recording demos for Doom' is going to be eradicated from existence. Why such harsh measures, you might ask? It's simple: We've descended so far back in time and intelligence that we now have people arguing that video footage is somehow just as legitimate (or even more legitimate than) demo files. As it stands, we simply cannot reach a consensus, therefore the entire speedrunning scene is to be quarantined and put on lockdown until further notice. Until certain individuals get their heads out of their assholes all runs can be proven authentic beyond any reasonable doubt whatsoever and demonstrated as entirely legitimate in a court of law, they are henceforth deemed invalid.

 

We apologize for any inconvenience this may cause and hope you continue to enjoy your stay in our lovely Doom community.

[Altima]

Thank you for all of the valid points, Archi. These are the types of responses I'm looking for. Considering the ridiculous things ZM has done, I guess it just personally strikes me as odd that a 90-minute run can show a bit of fatigue but a 4+ hour run doesn't really show any signs of it. Not saying long speedruns don't exist (Final Fantasy 7 Any% No Slots is usually 8+ hours) but I don't know of a game that has runs that can theoretically go on for hours and hours where the intensity doesn't let up a whole lot. At least with an RPG you get moments where you're text mashing which is a bit of a breather, or cutscenes where you can walk away for a moment.
 

[Altima]

12 minutes ago, Doomkid said:

Well folks, since the jury is still out on this one, we'll be closing the Speed Demos section and obliterating all previously held records. Furthermore, the very concept of 'recording demos for Doom' is going to be eradicated from existence. Why such harsh measures, you might ask? It's simple: We've descended so far back in time and intelligence that we now have people arguing that video footage is somehow just as legitimate (or even more legitimate than) demo files. As it stands, we simply cannot reach a consensus, therefore the entire speedrunning scene is to be quarantined and put on lockdown until further notice. Until certain individuals get their heads out of their assholes all runs can be proven authentic beyond any reasonable doubt whatsoever and demonstrated as entirely legitimate in a court of law, they are henceforth deemed invalid.

 

We apologize for any inconvenience this may cause and hope you continue to enjoy your stay in our lovely Doom community.

 

This sort of bitter, biting sarcasm (and presumed to be aimed my way) doesn't do anyone any good. I also get that you're throwing shit my way but it's whatever. All I'm doing is stating my thoughts and getting some rather good, thought out replies. Sarcasm contributes nothing. It's discussion and that's where it's been on my end from the start.

[Doomkid]

It contributes some much needed humor and levity to this appalling shitshow. No other demo recorder/record setter in Doom as been held to this absurdly unattainable level of scrutiny, not that I'm aware of at least. Proof is no longer proof, so what will be enough? I'm sure some other bullshit ""reasonable"" doubt can be conjured up even if he records his hands while recording his demo and simultaneously recording a video of it all. "Well suuure, the video matches up to the demo, which matches up to his hand movements, which all matches up perfectly to his stellar records set prior to this, verifiable by a multitude of outside parties, BUT where was he last Tuesday afternoon, huh? No answer, eh? Well there we have it, the run is invalid!"

 

It's getting fucking ridiculous and has passed the point of constructive discussion and entered the realm of absurdity. That's what I'm driving at here. Of course there's a bit of bitterness underneath, look at the fucking absurd lengths people are going to to justify their skepticism which should have well and truly been put to rest by now. that's my take on it and its no more or less valid than yours, nor are you the arbiter of what does and does not 'do anyone any good'. I want to say let's just leave it at that but I fear there are going to be pages to go after this post... For what reason, who knows.

[Altima]

When someone has given me a good answer, I've stopped pressing it as an issue. I personally found nothing funny about your post and feel it offered 0 to any discussion.

 

Also, let's not forget Winterfeldt's cheated run that was scrutinized. Maybe not to this extreme levels and maybe not to a legit run, but it did happen. Winterfeldt's was obviously cheated, but all I'm doing is looking for good, solid answers.

[Vorpal]

43 minutes ago, Doomkid said:

It's getting fucking ridiculous and has passed the point of constructive discussion

 

Actually it was pretty tame and constructive for a while ;-o

[Doomkid]

Well, I apologize then. I just don't like seeing people getting falsely accused and it seems like there is still some level of doubt, which I just don't get at this point. What I really want to know is, from this point forward, what will be considered acceptable proof for further speedrunning milestones/records/etc? (not sarcasm, just in case of any chance it comes off that way)

[Nine Inch Heels]

3 minutes ago, Doomkid said:

What I really want to know is, from this point forward, what will be considered acceptable proof for further speedrunning milestones/records/etc? (not sarcasm, just in case of any chance it comes off that way)

My take on it is that this discussion merely ended up confirming that the way speedruns were handled up to this point is not about to change, unless of course a new paradigm in terms of source ports establishes itself.

[Altima]

Unless someone has some sort of better idea or a more convenient solution, it probably won't. The speedrunning scene in Doom is too broad to really "enforce" camera shit but not big enough where there has to be some sort of "super secretive" council that evaluates every run with a fine-toothed comb. So realistically? I don't think anything will change.

[Killer5]

I have noticed the mention of heuristics in this thread. What is the implementation of these heuristics and how are they being used to detect legitimacy of demos (I am assuming this is their purpose).

 

Because all I have heard is that there is no way to detect cheating of demo recording. Period.

 

Is the speculation of someone watching who is a seasoned runner the only way we detect this?

 

Because this community is the only one (I participate in) where most of the top runners don’t stream all of their godlike runs. Examples would be Xelna for alltp and zoasty et al for Super Metroid.