Billy Mitchell's Direct Feed Recordings are MAME Generated

[Arbee]

Hi, long-time MAMEdev and PCB hacker here.  I had a few comments.

1) What are you guys talking about with the screen rotation?  You mean he was doing videos in default non-"tate" mode rather than rotating the monitor?
2) Playing DK in current MAME on an Nvidia GSync or AMD FreeSync setup with the monitor physically rotated should be identical to a real machine aside from the PC inserting 1 frame of lag; this would make it harder, not easier.  Prior to the advent of GSync and FreeSync around 2012 there would've been even more frames of lag.
3) No current version of MAME has re-recording capabilities on INPs, nor can you load/save states during an INP recording and get anything useful to happen (the game would simply desync on playback if we allowed it).
4) Getting video out of any standard-res arcade PCB is trivial.  You just need a DK to JAMMA adapter and a Supergun.
5) There's no such thing as a "validated PCB".  Contrary to the unintentional hilarity in the movie, asking Nintendo to validate a PCB is roughly like calling up modern "all we make is Call of Duty" Activision about the Dragster record.  They'll probably make some noises about it out of politeness, but you'd be better off finding the surviving former employees of Ikegami.

So that's my objections to all of this.  I don't have a side in this fight, but I do want discussion of MAME kept in the realm of reality.

[xelnia]

Hi, long-time MAMEdev and PCB hacker here.  I had a few comments.

Welcome to DKF!

1) What are you guys talking about with the screen rotation?  You mean he was doing videos in default non-"tate" mode rather than rotating the monitor?

It seems likely the first two scores were recorded using -norotate, with the third being record with -ror. I have witnessed people physically rotate their PC monitors to play this way.

2) Playing DK in current MAME on an Nvidia GSync or AMD FreeSync setup with the monitor physically rotated should be identical to a real machine aside from the PC inserting 1 frame of lag; this would make it harder, not easier.  Prior to the advent of GSync and FreeSync around 2012 there would've been even more frames of lag.

Lag isn't really a factor here; the issue isn't about whether the actual gameplay is easier or harder on MAME vs. real hardware, but the other benefits granted by using MAME (see below).

3) No current version of MAME has re-recording capabilities on INPs, nor can you load/save states during an INP recording and get anything useful to happen (the game would simply desync on playback if we allowed it).

There might be confusion here, because people are saying things like "splicing", "stitching", "save-sates", etc. I don't want to spell out the exact process here (for obvious reasons), but you absolutely CAN craft an INP that is made up of a segmented performance. And in versions of MAME that were available during the time Billy was setting these scores, it would be absolutely undetectable.

4) Getting video out of any standard-res arcade PCB is trivial.  You just need a DK to JAMMA adapter and a Supergun.

And Billy has stated that he doesn't even know how to switch a PCB in and out of a cab...so...

5) There's no such thing as a "validated PCB".  Contrary to the unintentional hilarity in the movie, asking Nintendo to validate a PCB is roughly like calling up modern "all we make is Call of Duty" Activision about the Dragster record.  They'll probably make some noises about it out of politeness, but you'd be better off finding the surviving former employees of Ikegami.

This is probably referring to the claims that his board was "validated" for the 1,050,200 score (the Mortgage Brokers score, not King of Kong). You're right in that the likelihood or reliability of that scenario is laughable.

So that's my objections to all of this.  I don't have a side in this fight, but I do want discussion of MAME kept in the realm of reality.

All MAMEdev input is greatly appreciated, especially if you (or anyone) have direct knowledge or experience into the progression of DK emulation.

[xelnia]

It's not particularly relevant to the evidence at hand, but I'd like to share this post from over at CAGDC. In that post, Ben Jos Walbeehm recounts the pain and suffering he went through after he beat Billy's 1982 score using MAME. He was actually the first publicly known person to beat Billy at DK, pre-dating Tim Sczerby's arcade WR. It's heart-breaking to read how he was treated by the community, especially knowing what we know now about Billy's use of MAME.

Ben Jos was a DK champion, and he rarely gets the credit for it.

[ChrisP]

All this "Billy did it for the greater good!" is some of the twaddle I have read in a long time.

I think those of us who are interpreting Billy's actions this way are well aware that we're being charitable with that interpretation, and we're all conflicted about it.

However, the interpretation that Billy was working alone out of pure self-interest is demonstrably false. The situation is much grayer than that.

We know that, at the very least, there were more cooks in this kitchen than just Billy, since getting these scores produced and verified would have required coordination between numerous people, especially on the technical level. There was indeed a conspiracy here .

[Sock Master]

2) Playing DK in current MAME on an Nvidia GSync or AMD FreeSync setup with the monitor physically rotated should be identical to a real machine aside from the PC inserting 1 frame of lag; this would make it harder, not easier.  Prior to the advent of GSync and FreeSync around 2012 there would've been even more frames of lag.

Not so, and that is how it can be detected that Billy's tapes were not recorded from actual DK PCBs.

Older versions of MAME did not emulate the way DK's video generator circuitry scans VRAM while the monitor scans across the screen when composing/drawing the raster image.   Old versions of MAME simply composed a complete image every 1/60th of a second then displayed them on the monitor.

That means any time sensitive writes to VRAM do not produce the same result in MAME as they do on DK hardware.   The entirety of the frame in (older) MAME shows what the video memory contained at one moment in time.   The frame from DK hardware shows the contents of VRAM across a spread of time - the left side of the screen shows an earlier point in time than the right side the screen.

Newer and even current versions of MAME do try to emulate this scanning, but even still, it will not display the same contents in these frames.  The latest version of MAME happens to emulate this scan from right to left instead of from left to right, so the timing information is opposite.

[Sqrlmonger]

All this "Billy did it for the greater good!" is some of the twaddle I have read in a long time.

I think those of us who are interpreting Billy's actions this way are well aware that we're being charitable with that interpretation, and we're all conflicted about it.

However, the interpretation that Billy was working alone out of pure self-interest is demonstrably false. The situation is much grayer than that.

We know that, at the very least, there were more cooks in this kitchen than just Billy, since getting these scores produced and verified would have required coordination between numerous people, especially on the technical level. There was indeed a conspiracy here .

I appreciate this post quite a bit, and I could definitely sense the conflict but it's nice to see it acknowledged at least.

This might be a crazy first post, but I did sort of sign up cause I had something I wanted to say about this, so hopefully I don't come off as just a new account for creating drama, this is not my intent at all.

For background I am not a vested member of the community but I have always been a fan of what this site and others like it do.  I just love good-natured competition that drives people to be get better.  I am keenly aware that a lot of what was in the KoK documentary was "Hollywood" so I try to base as little on that as I can, though admittedly it was my first introduction to this stuff.

Anyways, what I really wanted to add was a perspective of someone from more outside the community looking in. From this perspective I think the "Billy did it for the greater good" reasoning is not a good look.  I would hasten to add that how things look isn't necessarily the most important thing, but it does speak to the community's ability to inspire and bring in new members.  I won't presume to speculate on how that calculation looks, in terms of what is best for this community that I've only just now joined, but I do think a "Billy may have thought it was for the greater good, but fraudulent scores can ONLY hurt the community in the long run" sentiment would accomplish the same thing while delivering a better overall message.  Though in fairness, I suspect you would all agree with this, I would just hope to see it more explicit so people know where the community stands in stark terms (this thread is going to get a LOT of attention and I do not want to see news articles creating drama from ambiguity).

Ultimately, what's important to me as a spectator/guest here is to see that integrity and fair play are the highest goal of the community. Based on how this was handled I think this has been demonstrated to great effect and this community stands to grow from taking an active role and a firm but fair stance once the facts were in (though I have serious concerns about TG, but that's a whole other topic).

The only other thing I would add is on the issue of Billy's other scores, and specifically the issue of what does the leaderboard actually represent? I think this is an opportunity for the community to be proactive and discuss this in more depth. What has already been touched on in this thread is a good start.  Knowing that the LB is a record of what has transpired or a privilege for competitors who operate in good faith or something else entirely, is in my view a very good thing to know.  Knowing what the LB represents and having it as a defined and public statement of the community's views will make it much easier to answer these questions going forward without having to wrestle with the implications of each decision as you make it.

If it was up to me I would probably aim more for a historical approach. But I would also like to see any proven frauds/cheats scores reviewed at a higher standard of proof once they are found to have cheated at all (with all scores being temporarily suspended pending the outcome of the inquiry). Any scores that are clearly legit would stand, anything questionable or not meeting the higher standard would be asterisked and possibly included in the list for posterity but not considered for ranking.  But again, I recognize that as a "newb" here my thoughts probably won't carry a lot of weight, just a suggestion to consider.

Anyways, thanks to anyone who read such a long first post, and best of luck to you folks. Please keep up the great work!

Cheers~

[francoisadt]

Hi, long-time MAMEdev and PCB hacker here.  I had a few comments.

1) What are you guys talking about with the screen rotation?  You mean he was doing videos in default non-"tate" mode rather than rotating the monitor?
2) Playing DK in current MAME on an Nvidia GSync or AMD FreeSync setup with the monitor physically rotated should be identical to a real machine aside from the PC inserting 1 frame of lag; this would make it harder, not easier.  Prior to the advent of GSync and FreeSync around 2012 there would've been even more frames of lag.
3) No current version of MAME has re-recording capabilities on INPs, nor can you load/save states during an INP recording and get anything useful to happen (the game would simply desync on playback if we allowed it).
4) Getting video out of any standard-res arcade PCB is trivial.  You just need a DK to JAMMA adapter and a Supergun.
5) There's no such thing as a "validated PCB".  Contrary to the unintentional hilarity in the movie, asking Nintendo to validate a PCB is roughly like calling up modern "all we make is Call of Duty" Activision about the Dragster record.  They'll probably make some noises about it out of politeness, but you'd be better off finding the surviving former employees of Ikegami.

So that's my objections to all of this.  I don't have a side in this fight, but I do want discussion of MAME kept in the realm of reality.

Confirm my understanding of the MAME play by Billy.
a) Some mention there were possible ways to record the .INP playback via another videorecorder and then inside the cabinet play this back to the monitor? Is this what Xelnia is suggesting here?

Another question :
B) why is it not possible Billy had play on MAME-cabinet for real and just did record the screen while he was playing?
If the case is the later then the MAME DK score is still a valid score but not an arcade PCB DK version of the game but a MAME versione played and recorded. If there are no facts to certainty that he did not play the game why not upheld the possibiolity that he did but just on an MAME enabled cabinet which is quite popular years later (2018) eg: ICADE ect..?

That time when Billy did play it is all possible that custom MAME cabinet could have been the case - it just would have been a PC and a PCB inside the cabinet for game play and feed the graphics to the monitor from the pc. Then afterwards only do show the PCb footage and not showing the PC?

C) If this game-play is a valid MAME score why not list this on MAME DK score list instead?

[dnickolas]

Mame scores have a different mechanism for validation. Without recording on the proper version and looking at the INP it could have been “spliced” in the “keep trying a level until you get good rng” sense. The play style suggests that that may have been exactly what happened.

[TheSunshineFund]

How long would putting together an .inp like this take in theory?

And are we looking at a situation similar to what PAC did?

[Josephjo]

How long would putting together an .inp like this take in theory?

  in conversation with George Riley (Homerwannabee) at The Kong Off 2 in Denver.

"I'll let you on a little secret. Most people think I broke the record in one day, but in reality it took me two weeks of practice doing it."

This was posted in August last year in The DKF Shoutbox long before Xelnia's findings.



I reckon it took Mitchell about 2 weeks 

[xelnia]

How long would putting together an .inp like this take in theory?

And are we looking at a situation similar to what PAC did?

Unlikely this is like PAC, since he was manipulating WolfMAME. Manipulating vanilla MAME would be far easier. I reckon you could do it in a day or two.

[konghusker]

This whole thing is soooo disappointing. I like Billy. I agree with everything Dean and Chris said. The entire saga is what created many new friendships and adventures for most of us here today. For that, we can thank Billy and Steve. I also agree that ANY score not properly verified needs removed. Sure wish I was able to go to ko this time. Should be interesting. Still, sad this is showing up, but it's best to be thorough. Still has awesome hot sauce!

[maximumsteve]

For those that don't know me, My name is Steven Kleisath and I am the undisputed Guinness and TG world record holder for High Scores of "Mario Bros." single player and doubles play. I mention this because a score being legitimate is as important to me, when I know the hard work and determination it takes to break through to take down a mammoth score. I am just catching up with this relentless scrutiny of Billy Mitchell's DK scores on here.

I am friends with Billy Mitchell and Robert Childs. I have known them for years, and interact with them regularly on our Retro Arcade Night event we put on throughout the year at Arcade Game Sales in Fort Lauderdale, FL.

I also consider myself a friend to some of the DK players that have launched this new wave of trying to disprove Bill's past scores on Donkey Kong.
It is to the point where I feel I need to give my 2 cents worth of what I know and what I believe to be true and clear.
First order of business is Boomers in Fort Lauderdale in 2010.....
Boomers, also known as Grand Prix Raceway,  which was considered the world's largest arcade at the time, ironically did not have a DK cabinet at that time, so a distributor brought to the venue the cabinet that Bill would play on. Todd Rogers and his wife were in attendance, Robert Childs ofcourse was in attendance, the distributor who brought the game was there, but even more importantly, Peter Bouvier, the owner of Twin Galaxies at the time and a family member, was 100% there in attendance at the time of the rollover of the game, along with others witnessing in the room. These Original Tapes will show not only the gameplay, but the room overview of everyone that was in attendance. Also the manager of Boomers and a technician working there, were present.
On a personal note, I have known these guys for years, and they have families and full time careers of their own, and don't have time to go to such lengths to fabricate game scores.
As for the Direct Feed/MAME board part of this....
There are screen shots from the T.V. post that claims it was running Mame. You can see the defect on the right hand side. There is no way you can emulate a monitor defect like this in Mame, so even if someone used Mame in a cabinet, there would be a significant, obvious wiring addition tongue video output from the PCB. But if it was a video capture, then it wouldn't have the monitor defect. I haven't read into the whole post explaining the allegations, but from the screen shots they are making the references too, it's 100% an image of a Sanyo EZ20 Monitor.
CGA to Vega converters were not available back then, and even if they were, I am not aware of any that can replicate a monitor defect or can adjust an inverted video signal. For example, if your TV screen had a bunch of wavy lines and you recorded the Super Bowl on your DVR, then moved that DVR to a different TV, you shouldn't see any wavy lines.
In closing for now, Billy Mitchell will begin addressing this whole situation tommorrow, Tuesday February 6th, between 7:00-8:30 PM at https://www.compoundmedia.com on the East Side Dave Show. MUCH will be revealed in Audio and Video form in the coming weeks, so tune in and stay tuned.

[YesAffinity]

There are screen shots from the T.V. post that claims it was running Mame. You can see the defect on the right hand side. There is no way you can emulate a monitor defect like this in Mame, so even if someone used Mame in a cabinet, there would be a significant, obvious wiring addition tongue video output from the PCB. But if it was a video capture, then it wouldn't have the monitor defect. I haven't read into the whole post explaining the allegations, but from the screen shots they are making the references too, it's 100% an image of a Sanyo EZ20 Monitor.
CGA to Vega converters were not available back then, and even if they were, I am not aware of any that can replicate a monitor defect or can adjust an inverted video signal.

The likely scenario is the other way around -> VGA mame output to CGA.  ArcadeVGA/JPAC is one such solution.  I had my current MAME ArcadeVGA/JPAC-driven cab running circa 2004/5.  When were the scores in question performed?  Sorry if I've missed that detail in all this, and don't have the history fully committed to memory (honest question).

That leaves the question of inverting for a Sanyo 20EZ.  CGA output to Sanyo inverter board?

And there was an extra TV at at least one of these performances?  DK board set -> Sanyo 20EZ is a complete inverted transmission is it not?  So, how did the signal get to the viewing TV uninverted, in an original hardware scenario?

Plenty of scenarios for overcoming the technological hurdles, but at the time, a MAME PC outputting through a ArcadeVGA/JPAC setup to a stock Nintendo inverter board and Sanyo 20EZ in a DK cab, while simultaneously outputting to a composite TV (the ArcadeVGA has a composite output on it as well as a VGA DB15 output) seems the path of least resistance, and could have supported MAME and use of  recorded .inp for the performance.

Alternatively, if the cab performed on was 100% original, how did they get non-inverted video out of the cab to the viewing TV?

I think understanding the technology used to produce a direct feed at that time is also important here.  Inexpensive NTSC to composite and s-video encoders existed at the time.  But that still doesn't account for getting inverted video in a viewable format to a standard TV.

If I've messed up any of the details that might affect the technological approach for the time period, please correct me.

I also have a family and work 50-60 hours a week on average at my day job, and still find the time necessary to be a score chaser and general arcade technology enthusiast.   Not being stand-offish, just saying, we are all passionate about our pursuits.

:edit: I need to test my Arcade VGA and see if a composite signal comes out of it while the cab is outputting to the arcade monitor.

[JCHarrist]

Todd Rogers and his wife were in attendance, Robert Childs ofcourse was in attendance, the distributor who brought the game was there, but even more importantly, Peter Bouvier, the owner of Twin Galaxies at the time and a family member, was 100% there in attendance at the time of the rollover of the game, along with others witnessing in the room.

At 1:53 in this video, the games are over and Billy says that they are "waiting for Pete."


http://youtu.be/bFLGF933tgc?t=111

So how did Pete witness any "rollover" when he wasn't there yet?