Billy Mitchell's Direct Feed Recordings are MAME Generated

[RickyRicardo80]

Hi Everybody,

After following this story for quite a while now I wanted to chime in with some finding I had on my own. I'm a software developer and have experience with 6502 programming, and 8 bit logic in general ( mostly C64 and ZX ). While everybody was focusing on reproducing the finger girder, I took a different approach. Imho it is very hard if not impossible to prove something that isn't there, so by just analyzing the footage alone you will never EVER get a 100% assurance it can NEVER happen.

So, here is what I did. I know most 8 bit games draw their background as character map. This is way cheaper and has less memory limitation than using sprites, but it also add a little limitation of where the character can be placed, namely, only on a grid. The DK display is 224 x 256 pixels meaning it has a character grid of 28x32 characters ( 1byte characters of height 8 ). To test this I took a pixel perfect rendition of Mame and overlayed a 8x8 block pattern and sure enough the characters immediately stood out. For instance the score numbers, extra men etc all have a 1pixel blank to the left so the never touch, which is exactly as any developer would do this.

So far, so good. Then I took the image of the finger girder and overlayed it to my image, to find out where exactly the finger is, and of what pixels it is made of. Sure enough, the finger is exactly 1 character wide and only has one row of pixels. The missing part of the girder is exactly 7x1 characters.

I did some calculation, because I was now assuming a memcopy size to pop up somewhere. The section drawn in that cycle is 128x224 pixels = 3.5K minus 7x8 = 56bytes is a memcopy of 3.4453K. No programmer in his right mind would use such a size of a block of memory to copy. I thought a bit about it, and realised back in the day with slower memory, etc, it would be very ineffective to copy blocks of blank memory. So I did some counting....

The number of characters of the first cycle that end exactly with that girder is 128 characters. This is absolutely a number any programmer would use. This is EXACTLY 1K of pixel data.

As somebody stated earlier, the DK hardware has 2 separate processes to draw the image. The first process is the game logic that puts the girders in video memory. The second hardware process is the video routine that reads that memory and spits it out. These processes do not interact. Finally, the actually drawing is an analog process. There is absolutely NO way that these 3 separate processes line up perfectly to stop exactly at that girder, not once but many times in a single recording.

[Sock Master]

Attached is a visualization contrasting the rendering behavior of a genuine Donkey Kong cabinet and MAME. Original hardware performs a linear scan of video ram (upper left) writing the contents to screen (upper right). The Z80 CPU and video generation are on separate boards. Their behavior isn’t synchronized and the Z80 happily writes to video ram while the video board composes the screen. MAME took shortcuts and didn’t emulate this async interaction between the Z80 and the video generator. It reads the entire contents of video ram at once (lower left) and renders discrete frames (lower right).

This leads to pronounced differences in frame generation.

Here are the rest of the the DK render models.

Barrels, Pies, Springs and Rivets:

[serphintizer]

 

[Sqrlmonger]

This may well end up being my longest post, but hopefully it will be worth looking at.  Much of the work presented here is building off of what bh has done. I will get into detail about that as I go.

Methodology

First of all, I looked at the animations/videos bh posted closely and took images at key points. bh put numbers in the top right to align things and for lack of a better term I will call these Transition Frame Number or TFN (note these are not the same as DK gameplay frames in the slightest). 

The following are the TFN's I took images of and why I thought they would be most useful:
TFN 28: The arcade scan line has just completed a left to right pass and MAME just performed an output of VRAM at TFN 19 and won't do another until TFN 54. Thus 28 is a good point to grab the output values for comparison as we are likely to see aspects of this output in transition analysis.
TFN 60: As previously mentioned MAME output its VRAM at TFN 54 and TFN 60 marks the completion of another Arcade left to right pass.
TFN 93: Arcade completes a scan at TFN 92 but MAME performs an output on 92 (which covers the VRAM output) so I took the image at 93. I later decided to not bother analyzing the VRAM so 92 probably would work, but there is no difference between the two since this was the final arcade pass for these animations.

With these frames decided upon I created an image comprised of two rows. The top row would have Arcade output for the three TFN panels described above in ascending order from left to right and the bottom row would be the same for MAME output.  With those images in hand I was able to then split the Arcade and MAME rows into images, align them, and perform a difference on them  using a logical pixel by pixel difference check performed by diffchecker.com's Image Diff functionality.  This provided the output for the diff images which show bright pink pixels where differences exists between the two images.

Using the diff output as a guide my methodology was to locate points of difference and then re-examine the animations bh created to understand how these differences arise. Understanding how the differences arise is extremely important as it can indicate whether such differences are likely or unlikely to present themselves in live output of Arcade or MAME gameplay.  Importantly, we can also line up this method to see if other known transition differences emerge from the analysis as expected.

Evidence Examined

Barrel Stage




1st Panel: Classic 3 vs 5 girder transition difference is clearly visible.
2nd Panel: This panel indicates that with the exception of the ladders on the bottom girder if we see a capture that shows only the three ladders on the right side of the screen that this is indicative of an Arcade transition that should not be possible in MAME. By the time MAME outputs VRAM for a 2nd time all of the ladders have been drawn into VRAM, but for Arcade the bottom left ladder is only just drawn underneath the scan line at TFN 42. In live hardware I would not be surprised to see arcade play with and/or without this ladder but this analysis indicates we should expect it.  You can see an example of this from Wes at 4:22 but notice we get an extra ladder on the left side one girder up.  I think this demonstrates that this analysis does have some limitation and that verification is necessary.  To continue the point however this extra ladder is not always there as seen in the same video by wes  at 9:02.  But more to the point, this extra ladder and a friend appear to tag along in Billy's 1.04m video at 0:13 and 3:53 (literally the first two barrel transitions in the video).  So the point is that while the exact order of rendering may not be perfect, the analysis of the evidence is still revealing clear differences with strong connections to the animations bh has created.
3rd Panel: No appreciable difference.

Annotations


Barrel Stage Empirical Evidence

Wes 4:22	Wes 9:02
Billy 0:13	Billy 3:53

Pie/Factory Stage




1st Panel: This is perhaps one of the more telling panels in the entire analysis. In the arcade we expect to see these two ladders on the right, and for MAME we expect to see this one ladder on the left. You can see that here with Billy (left ladder) at 9:38 and Here with Wes (right ladders) at 25:50.
2nd Panel: The primary differences here are the girders in the lower left have yet to be drawn Arcade but they have in MAME. This difference however cannot help us identify MAME gameplay, but ONLY arcade gameplay because these missing girders will appear in Arcade but in MAME we won't know if they are drawn because we are on this panel or the next panel where both MAME and arcade display the girders.  Here is an example of this from Wes' video you can see it faintly at 11:33 and 17:40 but the best example is the 25:50 transition mentioned in the 1st panel. In this example we catch the DK Cab frames (the diagonal transition line), thus we can see this Arcade transition for the 2nd Panel as well as the arcade transition for the 1st panel all in one shot.
3rd Panel: No appreciable difference.

Annotations


Pie/Factory Stage Empirical Evidence

Billy 9:38	Wes 25:50
Wes 11:33	Wes 17:40

Elevator Stage




1st Panel: Here the main differences are the paired ladders show up as a pair in MAME but solo (right side only) in Arcade, but we also see that 4 specific smaller ladders show up on the right side of the screen in arcade but those 4 plus two more central ladders show up in this panel in MAME.  At 6:03 in Billy's video you can see double ladders and even though we are mid transition on this frame from "How high can you climb" screen (so we can't tell if we have the 4 or 6 ladders ) we do see we have one of the extra ladders so it looks like this is actually a 5 ladder transition.  Another example of how the animations are helpful, but not perfect. More work will be needed to verify that arcade can't produce this 5th ladder without the 6th but the ladder pairs are still quite clear when compared to Wes' video at 48:35 (it was hard to find a clean version of this indicating it may not be a good transition to identify Arcade)
2nd Panel: This is another example of some lower left girders not being fully drawn in on our 2nd panel and again it can only help us identify MAME gameplay because MAME panel 2/3 looks like Arcade Panel 3.  Additionally, we can see from Wes' video at 11:56 the bottom girder fading in and that the girder between the bottom of the ladder pair is NOT fading in. This indicates another difference between the animations and yet another reason why comparing to video footage for concrete facts is important.
3rd Panel: No appreciable difference.

Annotations


Elevator Stage Empirical Evidence

Billy 6:03
Wes 11:56	Wes 48:35

Rivet Stage




1st Panel: The bottom right two ladders render in arcade in the first panel while the left side ladders do not (see Wes' video at 7:53). For MAME however we see the leftmost three ladders render in the first panel while the bottom right two do not. So in arcade we should never see a transition were the leftmost three ladders are drawn and the right bottom two are not. I couldn't find a good clean example (though the hallmarks/indications are there, it's just not black/white with a clean transition) in Billy's first video so I started looking at the 1.05m video and found this at 3:08. Clear as day we have the left 3 ladders and are missing the bottom right two.
2nd Panel: At first it may appear that the only difference here is that arcade does not render a small piece of the bottom girder, but looking at the diff output highlight that the girder sections drawn WITH the ladders are the barrel stage style "truss girders" while the Rivet stage draws over top of these with the "oval girder" (You may need to click images above and look closely to see, but the visual difference is clear when you know what to look for). Unfortunately, the video quality on the videos I was looking at was not good enough to spot this truss vs oval girder difference, but perhaps the digital original video would reveal it.  For now I will leave this as an exercise for someone else to wrap up.
3rd Panel: No appreciable difference.

Annotations


Rivet Stage Empirical Evidence

Wes 7:53	Billy (1.05m) 3:08

If I were to do this again I would exclude the third panels, they produced nothing interesting.  Though I suppose it was best to be sure and demonstrating this fact is useful in its own right.

This was just some quick analysis on my part, I know a more thorough examination would likely reveal a lot more examples in Billy's performances but for now I wanted to show my results for the community to reproduce and allow for crowdsourcing any further efforts with my diff images from bh's work.

PS - I am going to edit this to include images to all of the still examples I pulled from the videos. References to those timestamps will remain for others to verify obviously.

edit: Still image examples have been added.  Please let me know if any errors, issues, questions, etc.. are spotted.

edit2: Went to dinner and when I got back decided some annotations might help. I've added them above.

[smf]

Here are the rest of the the DK render models.

What version of MAME is that? In 2008 MAME switched to scan line rendering, which still isn't perfect as instead of looking at vram once it looks at vram 224 times. I'm interested in how much that changes the output. 0.196 has the -norotate orientation matching a direct pcb recording, so that will make a difference too.

[Sock Master]

What version of MAME is that? In 2008 MAME switched to scan line rendering, which still isn't perfect as instead of looking at vram once it looks at vram 224 times. I'm interested in how much that changes the output. 0.196 has the -norotate orientation matching a direct pcb recording, so that will make a difference too.

The animations are computed models to visualize DK PCB rendering and the rendering of pre-V0.127 versions MAME.  They're just that, computed simulations to show how the two are functionally different from each other.  It's just an animation to show the *why* of the difference between MAME and Arcade.   A picture is worth a thousand words.  An animated picture is worth ten thousand words.

The current versions of MAME are much closer to Arcade rendering in theory, but MAME still gets it very wrong.  No current version of MAME will render like arcade because it scans VRAM in the wrong direction (it starts at the right and works towards the left instead of the other way around).   The -norotate option doesn't affect the order in which the framebuffer is composed so the output frames will not change in composition.

...In live hardware I would not be surprised to see arcade play with and/or without this ladder but this analysis indicates we should expect it.  You can see an example of this from Wes at 4:22 but notice we get an extra ladder on the left side one girder up.  I think this demonstrates that this analysis does have some limitation and that verification is necessary.  To continue the point however this extra ladder is not always there as seen in the same video by wes  at 9:02.  But more to the point, this extra ladder and a friend appear to tag along in Billy's 1.04m video at 0:13 and 3:53 (literally the first two barrel transitions in the video)...

There's something that needs to be taken into account when looking at game footage taken with a camera pointed at a screen.  Many video cameras have rolling shutters.  The shutters are also not necessarily synchronized with the screen and not necessarily open for 1/60th of a second.  The extra ladder you mention does not actually ever appear in the output from a DK PCB but is simply an artifact of the interaction between the camera's shutter and the scanning electron beam on the game CRT.  The still shots in your examples show multiple overlapped partial game frames blended together into one image.    The effect can usually be taken into account by looking for the 'cut' lines, usually diagonal lines with soft edges.   In the blue triangle in the bottom of the image below, one of the ladders is brighter than the other.  This indicates that the brighter ladder was present in both Frames 2 and 3, while the darker ladder was only present in one of the frames - we can deduce its presence in Frame 3 because in this context it's very unlikely that it was present in Frame 2 and then NOT present in Frame 3.

[smf]

No current version of MAME will render like arcade because it scans VRAM in the wrong direction (it starts at the right and works towards the left instead of the other way around).   The -norotate option doesn't affect the order in which the framebuffer is composed so the output frames will not change in composition.

I think you should rethink.

-norotate doesn't change the emulation, but the output does now match the direct pcb recording. When you use -norotate it ignores the game specific rotation and yoke swapping and assumes a horizontal monitor that scans from the top left to the bottom right.

As MAME is outputting the pixels in the same order as a direct pcb recording now, it follows that it's fetching the bytes in the same order.

Before 0.196, MAME had the equivalent of this hardware mod applied.

http://www.brasington.org/arcade/tech/flip/dk.shtml

0.196 and later has it the opposite way.

WRITE8_MEMBER(dkong_state::dkong_flipscreen_w)
 {
-   m_flip = ~data & 0x01;
+   m_flip = data & 0x01;
 }

MAME arguably should have it as a configuration option .

[Sqrlmonger]

...In live hardware I would not be surprised to see arcade play with and/or without this ladder but this analysis indicates we should expect it.  You can see an example of this from Wes at 4:22 but notice we get an extra ladder on the left side one girder up.  I think this demonstrates that this analysis does have some limitation and that verification is necessary.  To continue the point however this extra ladder is not always there as seen in the same video by wes  at 9:02.  But more to the point, this extra ladder and a friend appear to tag along in Billy's 1.04m video at 0:13 and 3:53 (literally the first two barrel transitions in the video)...

There's something that needs to be taken into account when looking at game footage taken with a camera pointed at a screen.  Many video cameras have rolling shutters.  The shutters are also not necessarily synchronized with the screen and not necessarily open for 1/60th of a second.  The extra ladder you mention does not actually ever appear in the output from a DK PCB but is simply an artifact of the interaction between the camera's shutter and the scanning electron beam on the game CRT.  The still shots in your examples show multiple overlapped partial game frames blended together into one image.    The effect can usually be taken into account by looking for the 'cut' lines, usually diagonal lines with soft edges.   In the blue triangle in the bottom of the image below, one of the ladders is brighter than the other.  This indicates that the brighter ladder was present in both Frames 2 and 3, while the darker ladder was only present in one of the frames - we can deduce its presence in Frame 3 because in this context it's very unlikely that it was present in Frame 2 and then NOT present in Frame 3.

I'm aware of the rolling shutter, the problem I have is that my understanding is that rolling shutter would be left/right or top/bottom depending on the camera.  It's sort of like a piece of paper with lines drawn from edge to edge, and each line adds at least 1 more region and each region can be a window into a different moment in the frame transitions.  So what I am trying to understand is, if the DK cab is creating a diagonal "line on the paper" why are both "lines on the paper" (which are basically just the borders between frames) roughly parallel to each other when the camera should be left/right or top/bottom?  What am I missing in this understanding?

I mean I definitely notice the difference in the phosphors brightness and in fact it's one of the reasons I chose the image (in spite of its flaws) is because you can see in the bottom left corner the way the girder is just lighting up as we would expect from the 2nd arcade panel on the barrel stage.

I'm starting to wonder if I should have chosen another cleaner image just for clarity but this may prove interesting and I'm always open to learn something new.

[Sock Master]

Before 0.196, MAME had the equivalent of this hardware mod applied.

http://www.brasington.org/arcade/tech/flip/dk.shtml

0.196 and later has it the opposite way.

Ah then yes, the latest versions of MAME might be a bit more arcade authentic.   I haven't checked to see if it really does look like arcade now.  Last I checked the emulated Z80 in MAME executed code considerably faster than it does on arcade hardware.  The emulation did not implement wait states when the Z80 reads or writes from DK's VRAM.

I'm aware of the rolling shutter, the problem I have is that my understanding is that rolling shutter would be left/right or top/bottom depending on the camera.  It's sort of like a piece of paper with lines drawn from edge to edge, and each line adds at least 1 more region and each region can be a window into a different moment in the frame transitions.  So what I am trying to understand is, if the DK cab is creating a diagonal "line on the paper" why are both "lines on the paper" (which are basically just the borders between frames) roughly parallel to each other when the camera should be left/right or top/bottom?  What am I missing in this understanding?

The diagonal cuts are the result of interaction between the CRT screen and the camera's rolling shutter.  The camera's shutter sweeps from top to bottom.  But the CRT (which is rotated sideways) paints its image from left to right.   So as the shutter rolls to the bottom of the picture area the CRT had time to refresh more of it's screen.   The bottom of the image captured by the camera is from a later point in time than the top and the CRT had refreshed more of it's image toward the right.

[smf]

Last I checked the emulated Z80 in MAME executed code considerably faster than it does on arcade hardware.  The emulation did not implement wait states when the Z80 reads or writes from DK's VRAM.

Is there any information available on what correct timing is? And any software that tests it?

[Flobeamer1922]

The attached image confirms that 0.196 corrects the VRAM scanning direction. However, it still doesn't look quite like arcade due to the scanning speed still being off, likely because the Z80 is running too fast, as Sock mentioned before.

[maximumsteve]

Greetings All. So I am relaying the summary and conclusion of Carlos Pineiro's testing and research. He is not an active member on the TG or DK forums and so he asked if I would share it. On a personal note, I want to thank Carlos for the time and effort he volunteered on this. I have gotten to know Carlos over the last month and a half or so by virtue of living in the same area, and it's amazing how a friendship can be stemmed through unique circumstances. And here it is.............

"By the time I saw the dispute, it was over 150 pages long, jam filled with information and allegations, beginning from what I read with the swipe vs. pop graphic, which was caused by the shutter effect.

Next was the converter and it’s operation. The converter produced a duplicate or lagged image was the claim. I demonstrated that it’s not so on a CRT, where the spare screen was ahead of the actual Arcade screen.

Next we had that the converter would not have been used because it doesn’t record color. This issue was cracked using a specific brand of VCR that was around before the creation of the tape.

Next we moved to the issue where the dispute claimed that after a lot of testing, Mame’s Version 0.116 through 0.121 mimic the stage transitions on Billy’s Tapes & previous versions didn’t. What I found was based from Mame development that those versions came out about 2 years after the creation of the 1.047 Billy Tape.

Next came that previous mame versions with modifications and tweaking on the settings, plus refresh rate modification, would mean that you can produce an image which looks close to the stage transitions.

With so many changes and modifications, plus page after page of suggestions, I turned my focus on testing the broad functions of the Genuine Nintendo DK Arcade PCB. This includes the way it was recorded using the devices in hand and what was available at the time of the creation of the tapes.
2 Genuine DK boards were used for testing, and 1 DK Double modified board, a 60 in 1 board, 16 in 1 board and 2 versions of Mame on a home PC Laptop. Most of these tests when performed were recorded and captured to share among the others which were also running tests. Most of these tests were not released publicly but are stored. Recording Mame from a computer, moving to a DVD, playing to a VCR recording, and recording from a capture was done to find differences and similarities.

Though the famous finger(Tail some call it) on the girder was a long focus in the research, after so many boards and finding many video plays from other DK players on YouTube... The patterns began to become much more clear. Similarities that could be seen on one type of DK vs a DK played on an alternative board.
After MANY hours and looking at frame by frame of the test recordings, others play and the recordings I produced in my personal testings for evaluations were taken into account.....

My Conclusion on the 1.047 & 1.050 game tapes is that they were NOT generated from a Genuine Nintendo Donkey Kong PCB. The signatures which are normally produced by those PCB's are NOT on the recordings.

Repeated testing and viewing of the game on those tapes do not demonstrate the signatures found on recordings coming out of Genuine Nintendo Donkey Kong Arcade boards." End Quote.

Again in closing I applaud Carlos for his unbiased and independent research, study and testing on this matter, and I hope regardless of which side of the dispute you are on, that from a technical standpoint, you can respect and appreciate the work put in just like both of us do for Jace and the TG Staff, and the DKF. Cheers.

[Sqrlmonger]

My Conclusion on the 1.047 & 1.050 game tapes is that they were NOT generated from a Genuine Nintendo Donkey Kong PCB. The signatures which are normally produced by those PCB's are NOT on the recordings.

I think this is the relevant bit KongTower.

I'm glad to see Carlos didn't let his personal relationship cloud his judgement. 

Having said that, I do find it interesting he did not address the 1.06 score. I realize we have a lot less footage to look at but if you can't produce the MAME transitions from the 1.04 and 1.05 tapes I'm not really sure how you can justify their presence from the same player on the 1.06 tape.

Edit: Moved to separate post.

[Sqrlmonger]

A color coded guide to the Joel West rant.

Commentary Irrelevant to the Dispute.
Chest puffing to sell his legal credentials.
Baseless legal threats/lame attempts to question the validity of the dispute.
DWAYNE DID IT!!!
Text relevant to a post on the dispute thread.

It has been 67 days since “Team Billy” has known of the Jeremy Young Dispute. Throughout this time many strange things have occurred. People who were once friends have become frothing accusers. Threats of all kinds, including threats of death, have been issued to Billy and his family. His life, income and peace has been disrupted, and still the mob of haters … whom suddenly ALL have become arcade AND MAME experts, continue their dispensing of personally-targeted hate. All of this inane activity, that even Junior High schoolers would be embarrassed by, is over a #12 video game score that has not been a record of any kind in over 5 years.

Not being a video game technical expert, or ever caring to be … I tried to use my training to help my friend of 35 years in an area that I could be of assistance. My training is in paralegal science. No, that is not an attorney, but anyone in the legal profession knows the paralegal is the worker ant while the attorney is just the ant queenJ. The paralegal is the steak, the attorney is the sizzle.
I am fortunate to know four good attorneys in diverse areas of practice including one who has successfully argued and won before the United States Supreme Court. I am diligent enough over the last 6 weeks to have spent over 130 hours in research citing case law and cases that can be relevant to this situation.

While I had originally written, with my cites and instances, 14 pages of information showing the misuse of the Dispute System within the membership system of Twin Galaxies PLUS the fact that there were over 30 instances of improper and discriminatory application … I decided the average reader would not read the 14 pages, nor care to understand its legal applications of which I could have easily added 4-5 more. PLUS, there is no need to give the opposition info that they should know and should have researched themselves. I do not work for them and there is no future in aiding the competition. So I decided to shorten the post. Besides, why show them what they will possibly be up against in court?

I will say this: there is no doubt that Twin Galaxies has definitely contributed to the defamation of Billy Mitchell.
* Twin Galaxies contributed to the defamation by their self-designed and very PUBLIC Dispute System (even Guinness World Records has a much more private system they have used while investigation of impropriety is on-going).
* Twin Galaxies (TG) owner Jace Hall has basically taken over Jeremy Young’s original dispute with his comments. Jace is not only the Judge deciding if the disputed score stays or goes but Jace is also the arbiter that decides if the evidence is inconclusive in PROVING the dispute (against someone who is ALREADY listed on the scoreboard). In the Young Dispute Jace Hall now has become the prosecutor in this essentially powerless jury system. Yes, members do get to vote, but TG officials (i.e. Jace Hall and possibly one other) make the FINAL judgement.
* The original dispute by Jeremy Young contended that Billy’s 1,062,800 was MAME. When TG was asked to simply reproduce the tape, THEY COULD NOT. TG ownership contended that they just “took over” 4 years ago and that they just weren’t responsible, they “didn’t even know” if Billy had ever actually given TG the tape (Yet that creates another allegation against Billy’s character and we have several witnesses to counterman that spurious accusation).
HOWEVER, I contend that TG has an on-going fiduciary responsibility to have “evidence in any dispute” BEFORE bringing a man’s legacy, reputation and good-standing character into question. Instead, Jace Hall has imposed his ultimate authority as head of a private company that provides a public service to members and non-members alike in allowing a process that is based upon a LACK OF EVIDENCE that TG should possess, but now wants to cry they may not have ever had it.
* Even though TG has had its new ownership for 4 years, it has WILLFULLY NEGLECTED its fiduciary duties to gather and archive the records it is responsible for on a timely and complete basis. There are several depositories of past TG tapes all over America. How do I know, I have one right in my house!
* TG also fails in its fiduciary responsibility to those whose scores it oversees (namely Bill Mitchell) and TG also fails in simple good common sense … to accept copies of tapes from a well-documented frothing hater of Billy Mitchell, Dwayne Richard. Richard has made a movie about Billy Mitchell, made phone and email threats to Billy, Billy’s family, and to Billy’s wife’s work. Dwayne Richard has been used by Jace Hall in the past when Jace was contemplating joining in with Jordan Adler of Denver in a competing “Kong-Off style event” several years ago. From the words of Dwayne’s own mouth “Jace gave me $500 to get out to Pittsburgh to check this out for him”. Collusion?
* This is the same Dwayne Richard who “made a movie” about Billy, whom Jace admitted supplied TG (Jace) with the tapes of Billy and Steve Weibe’s game play.
* The same Dwayne Richard who is seen in a YouTube clip with EDITING equipment upon his desk.
* This is the same Dwayne Richard who reportedly said to several different people, whom WILL TESTIFY in court, that Dwayne ask their help in pulling a “fast one” on Billy Mitchell with a “MAME fake” as early as 2009.

While the technical aspects are not my speciality:
1. I can look in someone eyes whom I known for 35 years and discern the truth or a lie.
2. I can remember seeing Billy do 1,000,000+ twice in all our thousands of miles of travels and 50+ appearances together (the scores were low into the millions and not world records).
3. I know that while Jace has all appearances of an honorable man, and has never done me personally wrong … like all dedicated and great gamers, he has an ego. That ego does not like to admit it is wrong.
4. Jace should have NEVER listened to someone with such a history as Dwayne Richard. That kind of obsessive and destructive behavior in “reputable consultants” is NOT the legacy that for which Walter Day intended for his legacy to be associated. I believe Jace has been DECEIVED by Dwayne Richard, as ALL the mob of haters has been so led.
IF YOU START WITH A FALSE TAPE, ALL “PROOFS” ARE WITHIN THE REALM OF THAT FALSE TAPE.
Richard has had 7 years+ to fabricate his AVOWED revenge against Billy for Richard’s inane feeling that HE should have been in the King of Kong movie. Jace (TG) has fallen right into a clever disguised ruse.
5. As far as the technical aspects, as more and more items were disproved by Carlos and David Race of Team Billy, then intelligently explained away … the TG Mob (which IS what they became) kept making more and more absurd accusations, they kept moving the “goal posts” (the essence of the dispute) and their frothing hate escalated. Meanwhile, Jace Hall watched … as his site hits kept climbing (controversy IS cash!!!). Jace allowed the focus to move to scores NOT EVEN ON THE TG SITE for 8+ years (1.050 and 1.047). These scores were allegedly put on YouTube by, can you guess … Dwayne Richard!
6. Jace has allowed the TG Mob to froth hate, defame Billy (as well as others of Team Billy) and do little to nothing to stop it. One can tell by Jace’s post as of late that he has been legally advised as how to write and appear vague, pseudo-fair and to appear not responsible for the hateful, defaming actions of others. That will have to be decided in court.
7. I propose to Jace Hall (TG) to SUSPEND this dispute now and do the following:

A. STOP allowing the persecution of the MOST Famous Classic Gamer in History (one of the TWO individuals who made your company, TG, SO valuable) … issue an apology to Billy AND his family PLUS direct your thread members to do likewise,

B. Get Dwayne Richard and Billy Mitchell in one location and get a certified independent polygraph operator to ask questions of BOTH while on camera and judge their answers. If you so desire, you could request Rob (board swap) Childs, Wes Copeland or anyone else you see as pertinent to this situation to have an invitation to be polygraphed.
You have been quoted as saying that polygraphs are just “gauges of who is the most nervous”. That is an absurd statement. The FBI isn’t in the habit of gauging one’s uneasiness. While the courts are skittish in allowing these machines, which would put some lawyers out of business, they are trustworthy enough for TG’s purposes. If trustworthiness of TG’s actions are in question … “moving the goalpost from 1.062 to 1.050 and 1.047. TG would not be talking about a “finger” then discounting it. TG would not be allowing the froth over the “board swap” to go on for weeks then discounting that video altogether when interest waned and internet hits slowed,

C. Prepare to negotiate a financial settlement if Billy is shown telling the truth OR if Dwayne is shown to have perpetrated this hoax.

There there is lack of integrity and there is breach of responsibility on TG’s part in this entire affair. TG DOES want the “Dwayne made-it-up” allegations answered, do they not? They do want to know if the MOST Famous player in classic gaming history is telling the truth, right? Or does, TG want to continue their flawed Dispute System where the accused are never notified by TG and well-crafted “made-up” allegations can destroy a man’s legacy, reputation, family and mental health?
Don’t just be a “businessman” Jace, be a MAN, and treat others like you would want to be treated. Keep this out of the court system, because that IS where this is headed. This is no longer about a score, but about the harm your public Dispute System and TG’s members have caused a man who has GIVEN himself to all your betterment and even tho you once said you “respected” the old scores … you have designed the system and fueled the froth to injure this man’s legacy, reputation and harm him and his family. You still have NO proof, just froth-spun innuendo. The “millions of dollars in new investment” TG now has will not be able to be used to make your company grow, if you continue to tear down an innocent man.
Joel West

[smf]

My Conclusion on the 1.047 & 1.050 game tapes is that they were NOT generated from a Genuine Nintendo Donkey Kong PCB. The signatures which are normally produced by those PCB's are NOT on the recordings.

Repeated testing and viewing of the game on those tapes do not demonstrate the signatures found on recordings coming out of Genuine Nintendo Donkey Kong Arcade boards." End Quote.

We've known all along what it isn't, what we don't know is what it is.

It could be:

 1. MAME
 2. another emulator
 3. a bootleg donkey kong board, which is only loosely based on the original
 4. a modified original donkey kong board

Billy repeatedly saying it's not MAME could be the truth, but that doesn't mean the score is valid.