Post Verdict Dispute Discussion: Jeremy Young - Arcade - Donkey Kong - Points [Hammer Allowed] - Player: Billy L Mitchell - Score: 1,062,800

  1. 10-13-2020, 05:07 PM
    Quote Originally Posted by Snowflake View Post
    i want to make sure i understand your suggestion. It sounds like your ruling out save states (which would be the main reason to use mame), and i dont think your claiming real hardware can produce the transistions, so are you suggesting he played a mame game legitimately and just misrepresented it as being arcade?
    I am not ruling out save states entirely. I am ruling out the use of save states as far as making an input file goes. Input files are used to replicate gameplay sessions. If you load a state during an input recording on MAME, the movie desyncs at the point you reload the state because all inputs are kept. Let's say you are on frame 1000 of your gameplay session, and you want to redo it from 600. You could reload to that point, but if you do it during a recording, the input file doesn't think "let's go back to 600". The 400 frames you think you are redoing actually become frames 1001-1400 of the gameplay session. Does that make sense?
    It is still possible to use save states and stitch together footage if you record the video. Tedious, but it'd line up with Apollo Legend's claim that the 1.05m was spliced. I think it's possible even though I can't seem to find the same proof he had.

    The only thing I can find that allows save states to work that way is a branch on TASvideos. The version it uses was originally from July 2010 though, so there's no way that could have been used.
    Last edited by Streetwize; 10-13-2020 at 05:09 PM.
    Likes ersatz_cats liked this post
  2. 10-13-2020, 05:32 PM
    Quote Originally Posted by Streetwize View Post
    I am not ruling out save states entirely. I am ruling out the use of save states as far as making an input file goes. Input files are used to replicate gameplay sessions. If you load a state during an input recording on MAME, the movie desyncs at the point you reload the state because all inputs are kept. Let's say you are on frame 1000 of your gameplay session, and you want to redo it from 600. You could reload to that point, but if you do it during a recording, the input file doesn't think "let's go back to 600". The 400 frames you think you are redoing actually become frames 1001-1400 of the gameplay session. Does that make sense?
    It is still possible to use save states and stitch together footage if you record the video. Tedious, but it'd line up with Apollo Legend's claim that the 1.05m was spliced. I think it's possible even though I can't seem to find the same proof he had.

    The only thing I can find that allows save states to work that way is a branch on TASvideos. The version it uses was originally from July 2010 though, so there's no way that could have been used.
    well if you're just interested in the "can it be done yes/no" as opposed to how, there was a thread "help marp is lying about me" for some reason it was easy to find with searches before but now cant. but long story short, someone had learned how to manipulate inps to cheat and even pass analysis, it was only because of extreme expert analysis that it was caught, but so as to not to help others cheat, the how was kept secret. basically marp caught the guy, some of the marp people that caught him are also tg people. so running to tg pleading his innocence to trick tg to fight for him just resulted in him being banned from both sites.

    in some ways its interesting to see a cheater so sure he can trick others, that he literally runs to another site, purposely puts a spotlight on himself and asks people to look it over -- he was that confident his forgery wsa that good. he actually wanted tg to scrutinize under the belief that if tg did so, then tg would come to his defense and say marp got it wrong.
    Lode Runner Champion
    Likes The Evener, ersatz_cats, Streetwize liked this post
  3. 10-13-2020, 07:06 PM
    Streetwize, here's the link that Snowflake referred to:

    https://www.twingalaxies.com/showthr...lying-about-me
    Thanks JJT_Defender thanked this post
    Likes ersatz_cats, Streetwize, JJT_Defender liked this post
  4. 10-13-2020, 07:13 PM
    Quote Originally Posted by The Evener View Post
    Streetwize, here's the link that Snowflake referred to:

    https://www.twingalaxies.com/showthr...lying-about-me
    thx evener, i'd love tips on how you found it, all my searches gave me obnoxiously off base results

    anyway look at just the fourth comments
    Name:  image.png
Views: 231
Size:  59.5 KB

    amazing, he declares he doesnt have any technical expertise and so isnt even capable of cheating
    he maintains he has no idea for the anomaly but assures you it wasnt cheating
    he claims he'll prove his innocence by playing again, as if a future performance somehow proves the past performances are legit

    remarkable. times like this i wish i understand human psychology better, but theres a case study in here for sure
    Lode Runner Champion
    Thanks JJT_Defender thanked this post
    Likes ersatz_cats, Streetwize, JJT_Defender liked this post
  5. 10-13-2020, 07:30 PM
    Quote Originally Posted by Streetwize View Post
    I was under the impression that only the 1.04m score lacked sound thanks to Apollo Legend. What I can find regarding the 1.05m score, at least from the two sources on this website, didn't have any sound. The evidence writeup by Ersatz further states that both had no sound. Is that really true? With Mr. Mitchell's attempts to remove evidence I have a hard time searching for other sources to fully verify this. The audacity sounds he showed were far different from mine, so I can't back up his evidence of splicing. I don't know if I'm missing anything or not. If it really did lack sound, there's still good chance he spliced it but we lack something definitive. For now, let's look at the other possible explanation: save states.
    TASvideos said MAME doesn't support re-recording, or using save states in the middle of recording to redo parts. I grabbed some older versions of MAME and, well. the results weren't good. It turns out, at least under normal settings, the game pretty much desyncs with the use of save states. I saved before the game started, let a barrel kill me, reloaded, and dodged the barrel. Upon playing the input file Mario still died at the start. This was tested on MAME32 to make running the games easier, though I highly doubt the native MAME would be much different. Interestingly though, it does fix DK's orientation to be proper on a computer at least, which might be why he didn't use it. I'll record a video of me doing the replay and replaying it later. Right now company is coming over.

    I believe in logic, but I also carry an open mind. I try to understand the other side's points so everyone is on the same page. People like him who use logic so poorly are not worth me engaging in a proper debate. I'd bring up examples, but I'd just be echoing what has been stated many times already. A lot of his technical evidence simply has no weight because little of it stays grounded. It's mostly just loose threads that barely connect to each other. Let's look at how he infers the girder finger can somehow appear in transitions because it shows up in the introduction. To put that logic into a different light, I'll use colors. Red and green make yellow when you raise the values on MS Paint. By the way he thinks, it should mean that when you mix the acrylics together for real painting they should also make yellow. Instead, it actually becomes brown. It's not the best example, but it's the best I can do.
    I don't like making this post because because I'm essentially giving evidence in defense of him. I still believe he cheated. Right now, there's just no definitive method. I refuse to support someone who has zero regard for the truth. Slandering those he once called friends in an effort to save himself is something I will never condone. To stoop to his level and withhold information because it wouldn't benefit us is not what I want to do either. I care for the truth, and, honestly, I want to be proven wrong about this. It is my wish that there is indeed something I'm missing here and get proven wrong.
    Howdy, Streetwize! These are all valid questions, and I'm glad you have the courage to investigate dispute assertions and to ask difficult questions. I am assured that it is indeed possible to fashion fraudulent inp files, and other cheaters have been caught by examining this. Unfortunately, real world considerations being what they are, the experts who know the topic are very reluctant to discuss details, since that could give others a playbook on how to cheat and better avoid detection. That testimony is something that could possibly come into play if the Billy case ever goes to trial, but just isn't being discussed now. But again, fraudulent inp files are definitely a thing.

    I will point out that, while cheaters have been caught submitting fraudulent inp files, Billy's tapes have an extra layer of insulation in that (assuming inp file manipulation was the method he used) the actual inp file was never submitted. It was produced through whatever methods Billy or his accomplice used, and once finalized, it was simply recorded as a replay, masquerading as live play. So many of the signatures investigators might have access to when examining a full inp file just weren't available in his case. In other words, Billy's inp file didn't have to be particularly sophisticated compared to those of cheaters today. It didn't even have to match the sound (as you discussed) or the opening rug pattern. All it had to do was look like Donkey Kong - and ultimately, it failed to do even that.
    Thanks JJT_Defender thanked this post
    Likes Streetwize, JJT_Defender liked this post
  6. 10-13-2020, 08:12 PM
    I am actually aware of those events. While not a member of MARP, someone I follow on Twitch is a highly-respected member there. I found out about it after everything unfolded, but it was mentioned a few times in his Discord server as well as in Twitch chat.
    What I know about WolfMAME is that is the basic stuff: it disables most basic ways to cheat, such as pausing, save states, and even the cheat function. I don't know the exact method of what PAC used, but I may have a solid idea given that there's only one way to go from there. For MARP's sake, however, I'm not going to disclose my theory publicly.
    It is within the realm of possibility for BM to use it considering how MAME was used back then, even though there it is a chance it wasn't easily accessible that long ago. Overall I don't think it's worth looking into just yet.

    Let's keep it simple for now and verify if Apollo's claim actually holds ground or he mislead millions, including myself, with his analysis of the second submission. This was publicly spread information. With EZscape's video there is reason to doubt the evidence he presents with his prejudiced conviction towards his opponents. Obviously we can't access his videos now, but I'm sure Apollo said the second submission had sound and the audio showed signs of splicing. The two videos I found of the 1.05m submission on the dispute thread only have static. Was he really lying or is there a video that actually has audio? If such exists please point me to where I could find it. I'll display the results here, promise.
    Once that's out of the way then I will focus on what has been discussed in the past few posts. I must admit though it's not exactly a field I have that much knowledge on.
    Thanks JJT_Defender thanked this post
    Likes JJT_Defender liked this post
  7. 10-13-2020, 08:22 PM
    i'm also reminded of this defense

    Lode Runner Champion
    Thanks Streetwize, JJT_Defender thanked this post
    Likes JJT_Defender liked this post
  8. 10-14-2020, 07:56 AM
    Without going into detail for obvious reasons, there are multiple ways to manipulate an INP with save states. It is also possible to remove the evidence of manipulation with the proper set of technical skills, which we discovered PAC did not have.
    Thanks Jace Hall, JJT_Defender, ersatz_cats thanked this post
    Likes JJT_Defender, ersatz_cats liked this post
  9. 10-14-2020, 10:56 AM
    Quote Originally Posted by WCopeland View Post
    Without going into detail for obvious reasons, there are multiple ways to manipulate an INP with save states. It is also possible to remove the evidence of manipulation with the proper set of technical skills, which we discovered PAC did not have.
    To follow up more on this, I originally thought requiring players to provide supplementary evidence on MAME INP submissions (showing hands, seeing multiple attempts over a live stream, etc) was a misstep and a bad idea.

    Knowing what I know now, I'd only consider the analysis program that's run on MAME submissions to be considered a sanity check, just like showing the PCB on an arcade submission. In the grand scheme of things, it doesn't prove much. If I was asked to bet my life on a submission's validity if given only an INP, I would not take that bet.
    Thanks JJT_Defender, ersatz_cats thanked this post
  10. 10-14-2020, 05:29 PM
    Well (for the third time now), I'd still like to know more about the 1.05m submission. Is the information being spread about that submission having sound a lie or not? I don't feel like explaining everything again.
Page 58 of 60 FirstFirst ... 8 48 56 57 58 59 60 LastLast
Results 571 to 580 of 595
Page 58 of 60 FirstFirst ... 8 48 56 57 58 59 60 LastLast
Join us