Why Can't This World Record Be Beaten? "The Impossible Speedrun"
HTML-код
- Опубликовано: 27 сен 2024
- 35 years later and still nobody can beat Todd Rogers time, not even the game itself.
Follow Omnigamer:
/ theomnigamer
/ omnigamer
Atari 2600 Dragster Simulator spreadsheet:
docs.google.co....
Support us on Patreon: / speedrun
Subscribe to the Second Channel: / ezscapespeedruns
Subscribe to the VOD Channel: / @ezscapeclips
Join EZScape's Discord: / discord
Follow EZScape on Twitter: / ezscapeyt
Follow EZScape on Twitch: / ezscape
Videos Used:
www.youtube.co....
www.youtube.co....
www.youtube.co....
www.youtube.co....
www.youtube.co....
www.youtube.co....
www.youtube.co....
www.youtube.co....
www.youtube.co....
Outro:
www.youtube.co....
![DDG & G Herbo - Nosey [Official Video]](http://i.ytimg.com/vi/I9Ra0Zc2NlQ/mqdefault.jpg)
What do you guys think? Let me know here: strawpoll.com/5w7c11e
Hope you all enjoyed :)
EZScape took too long 0/10
EZScape Why did you not just use the RUclips poll function for this?
UsernameNotFound 404 never knew it existed, I'll use it next time. Thanks
EZScape spell icup
Good vid, but clearly this record should not stand. Why? NO CONSOLE VERIFICATION. SDA would never accept it.
I once got a 5.49 but i was working for the CIA at the time and the footage cant be released for matters of national security.
Insert POLYBIUS joke here
Why is link not working, pls help.
Lol noob I did 5.11 with a glitch but I was so excited I forgot to record it or take a screenshot
You were in 1982? :O “World Record: 5.1”
I have gotten a 2.34 once by just slamming the cartridge even tho i wasn’t alive when i got it but it’s 100% legit i swear
I got a time of 5:32 in 1997. My goldfish was there, and so was my dog. Both dead, and no video proof. Believe me pls
UPDATE: Holy shit, yesterday (May 7th, 2020) I replicated the record. I only needed to pop the clutch straight to third gear, then switched to second when the timer hit 0:57 and afterwards spam the clutch for insane speeds. My chinese action camera broke while recording, so no footage. Sorry.
Uncle Dave
Ladies and gentlemen, we have a new champion!
I used an ouija board, his dog and gold fish said that he did indeed get the time
Uncle Dave = Mr. Activision
Uuuuuuh no
You're the uncle of video games.
Mr. Activision is actually the worst nickname ever.
Tbf I think toddzilla pips it.
especially now, holy christ they are fucked...
RedVGFox Mr. Activision: the king of stealing people’s money
Mr Electronic arts sounds worse
It actually holds the Guinness world record for worst nickname ever.
There is some controversy about this, because it is a record that is impossible to beat or even tie.
maybe they thought the 7 was a 1
oh shit
They look nothing alike lol
That's reasonable depending on something written on paper as people have very weird ways to write a 7 but on a computer game (Atari) maybe not. They probably made up the record and left it as is without any intention to prove it
If you are referring this the commented numbers being nothing alike you should reconsider as there are things called fonts....
+PokéPersona
In the game you idiot.
You get my upvote for 1 simple reason:
You made "Dragster" interesting.
I had this game as a kid and it was fun for approximately 5.51 seconds.
lol
Shots fired lel
original :D
He got banned :D and all his records removed. lol
Heldraugr juiciness at it's best.
Even his guinnes record.
Niko Heinonen like any exist lil
Did you mean lol
Are you guys sure you're not mistaken him for Billy Mitchell, the king of kong guy?
I have a rather messy history with Todd. The most infamous of which was his Barnstorming record. At the time I was working for TG, I would play and review the games Todd had records for, and Barnstorming was the first one where I realized his record was simply impossible. I and others that suspected something was amiss, immediately got flamed and attacked for ever daring to question Todd's accomplishments. It was only after I hacked the code of the game and removed all obstacles that I was able to conclusively prove Todd's record was impossible. His time was FASTER than simply flying straight to the finish line, and when people saw this, we finally got some traction on investigating the record. However, we were still attacked with venom and vitriol whenever the question was raised, and often told "You don't know how good he is", "He'll do it live and you'll be eating your hat", etc. Well we were ready to eat our hats, but such demonstrations never happened. Ultimately, the referee finally had to admit he never actually saw Todd do the record, and merely recorded an "error caused by a coffee stain in a magazine". Keep in mind Todd was made aware of the debate over his record, yet never stepped up to say that indeed the record was misreported. He let it stand as is for decades, and his fans would violently defend his honor over it. As a result, I lost all trust in Todd, and indeed the friendship was broken. It still brings back very bitter memories for me as to how I was treated when I knew for a fact the record was bogus.
KarbuncleX woah
"The friendship was broken"
While this sound like a legitimate story, you going out of your way to get his record altered and evidently coming between your friendship makes it seem like you cared more about proving he wasn't that good than the friendship.
I trust you were right, but also question your choices, as if it were me in this situation, I'd keep a friendship over proving my friend a liar, something that would impact his livelihood, over something as insignificant as a record in a video game.
I hope it will be acknowledged some day.
I mean, it's not that it was "as insignificant as a record in a video game." TG is a place who's only mission is to verify old video games records. Damaging the reputation of a site like that over a friendship in my eyes is worse. I wouldn't lie or ignore something to try to be a good friend, and it's not like his life would be that impacted if he faked a video game record. He didn't do them for a living, so he'd be fine.
In essence, TG's authenticity is quite an important thing, and maintaining that kind of scrutiny is necessary to uphold that reputation.
@metobsessed Someone who would straight up lie to you like that is not someone who you would want to be your friend in the first place. I don't really understand the moral grandstanding here. If you *really* wanted to push it, perhaps one could say that a friend wouldn't lie to you without good reason, but something as trivial as this (i.e., not a life or death matter) certainly doesn't count as good reason.
Whether you think friends should ever lie to each other is certainly a contentious topic, but it's fairly uncontentious that if someone is willing to lie over such a relatively trivial thing, they have issues, and they are not friend material. Maintaining a friendship with someone like that is incredibly risky at best, utterly foolish at worst, especially if they're in a position of power.
Agreed my good sir
Had no idea Ron Jeremy was such a gamer.
patissanchez Ive never laughed so hard at a RUclips comment before. Thank you
patissanchez He Loves Sonic 'The Hedgehog' 😂😂😂
Ron Corcerean? He's serving 30 years for doing it to a little girl repeatedly.
I was able to get a 5.51 on my calculator, started with a 6 then used the - key with a 0.49.
Uqpaa Handy terrible spelling
Pranav Srinivasan Terriable Capitaliziation
bro if you had engaged the clutch on the countdown and gave your calculator gas, you could have made a sub 5.50...
@@TheEgglet
Also terrible capitalization.
@what in the world
Actually, yes it is.
It's literally impossible to get the time he did, people have dug through the code for the game
TacticalBBQSauce he was revoked as shit hallelujah
i saw something that said he started in second somehow i believe
Jacob Jensen Nah, the mechanic was confirmed as impossible. Again, the game's files would account for this, if it were true.
TacticalBBQSauce but did you factor in the human element?
D Sandoval go watch the Ben Heck video WITH Todd Rogers trying to explain why he can't get the 5.51 -- if you know anything about Ben Heck, I have nothing else to say
The Biggest Cheater in Gaming History?
PinkBlueUnicorn you dont understand the concept of a joke do you?
NallePu83 yep
NallePu83 yea
lrmclinn he is a cheater
He is not a cheater
I'm here after he was confirmed to be cheating...
Potato on his wife?
@@JihadBunnydick on everything.
Stupid fucking mistakes man
Tag-along.
I both
He got the 5.51 out of a supply drop.
Pethy 00 That's the joke my dude.
Damn. All I ever get is 7.62 :/
xuenilom thats his k/d in ww2 we can't replicate it yet cause its not out
3 times to! That lucky bastard!
The round is 5.56, not 5.51 (*triggered*)
If he was able to recreate it, it must be legitimate.
However, if he used his console and his cart, maybe not. Guess we'll never know?
The whole point was that some thought ig was impossible which is why he needed to recreate it multiple times, so they more than likely used a different console and cart each time.
Purge Gaming what kind of records do we have from those attempts though? He might have failed the attempts but still given credit for the record because people said he got it.
I'm skeptical that emulation of the game is perfect, but I imagine people have attempted this with the real hardware.
Nuclearboy
How do you believe ANY written record of history then? DUmb ass kid. EVERYTHING was in printed stories. News papers, magazines ect.... like it or not, 99.9% of history is by written record and word of mouth.
You're assuming he used a different set up each time. He would have been able to bring his own cart several times, as it is very unlikely that people would expect him to edit his own cart.
Not to mention he got the same result seemingly in 1 try each time.
If that is a record, there should be some variance, and not easy to perfectly recreate each time.
This creates a few questions, and assumptions only make things worse.
starbattles1 You're a fucking moron.
"You got the high score!"
Input your name
Todd: "hmm... T..O..D..D.. no wait..
"5 . 5 1 ..hehe yeah thats my name"
by jove, you've figured it out! :o
If Todd keeps starting on 2nd gear he's gonna fuck his clutch up.
oh, ok!
Jumbie Americans don't understand that shit
I hate weebs nice bait
Maybe there's something about the emulator that makes it different from how the actual console operates?
That's my belief. While emulation can accurately play games, they aren't perfect at actually emulating the console, since it's illegal to use the BIOS of a console for an emulator. This is why some emulators are banned in speedrunnning IIRC.
windowsforvista It could be similar to the WiiU virtual console and the actual NES.
Since the WiiU VC runs on slightly lower frames than the NES over the course of a speedrun you would lose time. In Super Mario Bros this comes down to about 2 seconds in roughly 4 minutes.
If something similar was the case here it could explain the missing 0.06 seconds.
However since apparently there have been calculations of theoretical execution of this which can not reproduce this time - this seems a bit unlikely.
Additionally noone was ever able to perform this "glitch" he claims to use to get this time, while he was apparently able to get it at least times to get this record. And not even code analysis even confirms the legitimacy of this.
Regardless of how accurate an emulator might be, the simulator I created (spreadsheet) was built solely from the game code. It is mathematical progression, pure and simple. The emulator simply backed up what the simulator gave out. As best as I can tell there was nothing else that could possibly influence the timing other than what I have in the simulator.
Omnigamer Unless you would be able to somehow perform the "glitch" he claims to use, which noone is able to replicate, and then not get a perfect run the record time is not possible.
And since the gamecode apparently does not allow for this "glitch" to happen the only way this could technically be possible is through some kind of memory corruption.
In my opinion this record is illegitimate unless someone is able to replicate "this glitch" without any sorts of manipulations.
He probably actually got a 5.61 which other people can execute as well. And the Activision Simulation of 5.54 never happened, but instead we are talking about the 5.64 they got with their "perfect run" - This sounds like the best and most logical explanation to me.
maybe a cart defect of some kind?
I guess Activision was not such greedy fuckfaces back in the old days.
CaptmagiKono Nice profile pic from the best film scene ever.
Dylan Higgins Nice profile pic from the best scene in the best videogame.
Virtal Nice profile pic from the best deviant art user ever
Alienrun Nice profile pic of the greatest ball ever.
AbstractNateruz Nice profile pic of the best weird green thingy
Darbian has just joined the 5.57 club
brad walker this should be proof enough the time was bogus. Darbian is half machine.
he's also a genius in computer science he's one guy who i probably can never beat in a video game for speedruning accept mabey 1 of the Harry potter games only because darbian dosnt do Harry potter but I don't know I don't even know if I will ever even beat any of toorothkings records in the hp games I can but it just may not happen
I think we’re missing a very vital nickname here: Todd Todgers
He got all the DLC upgrades
He brought the speed boost micro transactions
Todd got the version where only one person verifies his runs from twin galaxy who is now serving a prison sentence for repeatedly raping a child
I honestly believe he got 5:57 and took a photo of the screen at an angle such that the 7 was ambiguous/blurry/pixelated and wrote "5:51" under the picture and the record keepers believed it based on his reputation and that the "1" digit with the diagonal tail, looks a lot like 7
wow they still claim its a record, with no evidence he actually did it, times have changed man
I truly hope so. How fucking high off your own farts (TG) do you have to be to not see that everyone can see how full of bullshit you are?
Good news, Twin Galaxies has purged all of Todd's records and banned him for life
@Miles Noctis
Yes,they witnessed the record
or I must say,one referee witnessed the record
and that referee is Todd friend,how amazing is that?
They broke down the original games code, the time he got is literally impossible.
No, now times have changed because they finally removed this record after it being there for decades.
I think this was a publicity stunt, actually. Think about it, there's no proof that it happened outside of people at Activision, the time is literally impossible with how the game is set up, and this happened right around the time of the video game crash of 1983. Activision gets to promote their game, Todd gets fame as the "King of Video Games", and Atari gets to at the very least try to market to adults, now that adults can see another adult playing video games. This also ties in with the convenient facts that some of his runs turned out to be fake, and if you could actually start in second gear, you could get a time much lower than 5.51. So, either publicity stunt or console defect. Either is possible
Viewtiful Z or real
No. I don't think this was real. If it was real, then I assume it was a cart or console defect. The only way this could be real and this wasn't a defect, was if there was 2 printings of the cartridge, one having this glitch and the other not having this glitch. Seeing as how there was no recording of this happening, I doubt it
Viewtiful Z Its a valid point,it could well explain why its impossible and why its never been repeated in the years since
It would also explain why this glitch would get a much better time if it existed.
+Viewtiful Z Yeah, the whole part about being able to have less than a 4 if the supposed glitch is performed well kind of casts doubt on the entire story.
Guys I got a time of 0:00 in dragster. I have witnesses too including “Hitler but in an alternate universe where he did nothing wrong”, and “Disney’s good star wars trilogy”
They’ll both vouch for me I swear
Why don't they just have this guy do his trick now?
He probably can't do it anymore, considering he's a lot older and presumably out of practice. If he could replicate it again, that'd be awesome, but I doubt he will.
aye the last time he did it was 1982 lol
That's not the point. The point is, why is there no actual proof that he actually did it?
Viewtiful Z 1882...
what happened in 1882?
0:28 TheLegend27...
Garfield Yea kinda died within a month, just like most other memes
TITANSPIRIT?
I got 3.xx but unfortunately I smashed the vhs (with footage on it) and forgot my Dropbox password. I am the real King of the Video Games!
It could be the physical hardware of the console or even the tv.
I remember watching something about how some older consoles would actually run slower or faster depending on their power source.
Maybe the way he moves his controller affects the inside of the console slightly.
idk just some random ideas.
Apparently the game clock and speed were both handled in the same way by the processor, meaning that speedups or slowdowns affected them equally and wouldn't make them desync (according to omnigamer)
Doesn't change anything.
We are talking about a way older and simpler console generation there. they do not care about clockspeed at all. You can literally speed-hack them by changing a crystal and still the outcome would be exactly the same as there is no clock that measures the time.
Everything happens on a per frame basis. the care moves ahead a fixed amount of pixels, the time gets incremented by a fixed number. If you slowed down everything by a factor of 10 then it would take you 10 times longer to finish the race, yet the time shown in the game would be the same as if you played normally.
I have never seen the code, but the actual game speed would be dictated within the code not by the hardware. its possible some weird malfunction could skip a game tick however
Yeah, I think it is because of his console and game. Maybe there was some glitch that made it possible only on his console and game.
Memorable Name um dude thats not true if you ever played sonic as a kid then you'd know that depending on where you live either PAL or NTSC the game would run faster with higher pitched music
All these “beaten by Darbian” comments are hilarious…
Darbian’s Time, a 5.57, is tied for WR, because it’s a legitimately possible time. Todd’s “record” was a 5.51, which has no proof behind even being possible. Darbian did tie the WR, but he didn’t beat Todd’s supposed “time.”
Everyone either beat or tied Todd's time because Todd's time (and any time he's ever posted) has been removed from TG and he's been banned
Lol didn’t Darbian get 5.54?
Sam Yost No, nobody has ever gotten less than a 5.57. It’s literally impossible.
The Phenomenal Star THANK YOU
The Phenomenal Star yeah wtf is up with those comments. It's so dumb...
In 1706 I beat this record! At the time I was at my grandmas house showing my dragster skills on her television. Her dog, me, Her cat and her herself witnessed the attempt. It was a time of -69.69 by using a wrong warp in which I forgot how to replicate by the time video cameras were introduced. Because of this I have no footage and I haven't beat 5.57 with the normal method. Of course, my grandma, her dog and her cat are dead and I'm only alive due to being frozen in ice for 312 years. I know this because I was told on a piece of paper. I'll try to shift into second gear and get an untied wr!
YOU ARE LYING YOU ARE CLEARLY WRONG YOU PIECE OF GARBAGE HOW DO YOU THINK ANYONE WILL BELIEVE YOU@!?
/s
TV didn’t exist at that time
3 people in a row don't get the joke wtf
Triple (insertoutdatedsubreddit)
Guys *obviously* this is real he's *M a t t f r o m W i i S p o r t s*
4:18 when hes not running for his life, john wick enjoys playing video games
MrDammed1 Thats not John Wick, thats Garret Bobby Fergusson. (Giant Baby Face)
Karl Gerg too real
Emulation is well, emulation. I think if people could find a way to put perfect inputs on an actual Atari 2600, then we might get more information, if that hasn't been done before.
Honestly, the 2600 is such an old console running such simple code that I'm sure a 100% cycle-accurate emulator exists for it. Testing frame-perfect inputs on a good enough emulator should be indistinguishable from actual hardware, unless Todd was taking advantage of some really weird hardware quirks. I guess that theory would explain why no one else could match his time even with his instructions, though.
Emulators aside, the simulation (spreadsheet) I created was built from the game code itself. I also investigated other possibilities, such as sub-frame input switches, but the way the code is oriented that would not have allowed anything different. Even additional lag, or an artificially sped up console would not have affected the timing.
Is it possible, like Bootleg Jones said, for there to be some quirk of the 2600 hardware that could be abused to alter memory?
TBH, my money is on a dodgy cartridge, but the idea of the 2600 itself having some bizarre hardware "bug" is also exciting to me.
It's unlikely that anything else in the system could have affected the game state. The processor alone deals with manipulating any and all bits of data, and the only interactions with other components has to deal with display and audio. The most likely scenario is memory faults, but those still amount to hardware failure and nondeterministic influence.
Omnigamer Sir. you simply pull half the cart out while the countdown is starting. Don't Make me get a 3.11 world record using my amazing method.
He was stripped of his records this morning!
Volvary About time, too! 🤣😅😅
He now has a guiness record for most records revoked at once
5.51 is impossible. Its not even in the code.
Many people have known he cheated, he cant even do it again.
And now he's banned from Twin Galaxies and all of his records have been removed
Do you not now about DGR
I believe tod roggers got 5.51
Maybe it was an earlier version/revision of the game which would enable you to start in 2nd gear?
or he just cheated which is the simplest explanation
Ah this video has aged like milk. If only we knew about Todd back then with what we know now.
I got a time of 5.46 back in 2012. The way I did it was holding the A button as I entered the race before building up speed before the race began for 12 hours, letting me go to an alternate universe and, by a series of precise inputs, obtained the 5.46 in half an a press.
But an A press is an A press, you can't say it's only a half
I love these videos, they're so interesting!
oh hey, that's my post at 2:28. shoutouts to dropbox for screwing over my avatar by removing support for public folders. i guess i'll vamp about my experiences with TASing this.
the consensus at TASvideos back when i was working on it was that the emulation of BizHawk's Atari 2600 core was faulty in some way, something involving frame timing, but several people tested older revisions with a different frame timing without any luck there. a user named Alyosha then vastly improved the Atari 2600 core over several revisions, and each time there were changes to the core i'd go right back to Dragster and see whether or not i could push down the final time.
if i remember correctly, three people posted files of 5.57 times using slightly different strategies, and i eventually spent nearly an entire day working out the most optimal shifting method through pure trial and error only to find out that i could still only get 5.57 times. i don't honestly think it's possible to get below 5.57 even on console without some fluke that i presume Todd was able to take advantage of.
my memory's a bit faulty but i believe there's a thread on AtariAge somewhere where Todd posts a picture of a 5.5X time, i wanna say it was 5.57, and to my knowledge that's the only physical proof of a time below 5.61 outside of our TASes (which obviously don't count as human records). i couldn't tell you which thread, there are probably 5-10 threads there alone asking how to get good times and i read through each one several times trying to figure out what Todd did. that post of his shown in the video is burned into my memory with how many times i analyzed it.
a couple theories i have:
* Activision's simulation could have been programmed into the game directly and thus wouldn't have to deal with anything like input delay or human error, thus it may have been able to find a shifting pattern that only it could do to reach 5.54. my reasoning here is during my testing, i discovered a hacked ROM of the game on AtariAge called Dragrace, and someone had implemented an auto-shifter into that ROMhack, and after a few tests i was able to get a time below 5.50. i discounted it because of all the physical changes made to the ROM possibly affecting the timing (i could never find an auto-shifter for Dragster itself), but it could lend some credence to the theory
* TASes start from power on and all our attempts would have been read by the game as the very first attempt made on the game after power on. i presume during verifications that it took Todd dozens, perhaps hundreds of tries to recreate the record, which leads me to believe that it could be something having to do with multiple attempts, perhaps the frame timing matching up with a perfect shifting pattern (since the atari 2600 runs in 60 fps but only polls input every other frame, with the occasional frame dropped)
* on that same line of thought, maybe there's some faulty logic in the game where things get messed up after hundreds of attempts, allowing for a lower time than should be possible
either way, i spent way too long working on this game only to get nowhere, so it's awesome to see Omni (who's helped me out immensely with several TAS projects in the past, thanks again man) putting in the kind of technical effort i never could in order to figure this out. as of right now i'm convinced it's not possible under normal circumstances, but knowing Omni i feel like it's not gonna take too long to get to the truth behind this.
The "fluke" he was taking advantage of is called Photoshop.
Then write a program that can play the game over and over again without the need for a reset, let it run 1 million iterations then look over the results.
4:32 neglected info on the record... guys, it's just fake. If you can control every frame of the game manually and change its inputs and STILL don't get the 5:51, there's no way a human did it. Frame perfect means no fails, you can't go beyond it. There are enough reasons to just dequalify this record already, dammit.
Darbian got 5.57 like a month ago
GaleblazeXYZ the video was clearly made over 3 months ago
i peed myself. dont tell anyone twit you are replying to a comment 2 months ago you fucking idiot
Carrot,Brocolli,Cabbage,Vegitables Android lol your child-like rage is pretty funny
Carrot,Brocolli,Cabbage,Vegitables Android still
I love how chill Ben Heck was while utterly destroying Todd's legacy forever.
It was destroyed way before then.
i once had a 5.51 while sitting backwards with a pillow over my head blindfolded and eating pizza and drinking cola AND having a tattoo made on my arm while doing the eddy capo from tekken 3 movement while lifting my keyboard in one perfect frame and all this in 2nd gear!
I'd like to point one thing out, as a programmer: emulators are not a substitution for the real thing. Until a hardware TAS is attempted, there's no reason to doubt (or accept) the run.
Why? If you write an emulator, you quickly figure out the a "realness" problem.
First, the computer isn't just a big matrix of digital gates. You'd think every gate combination is specifically coded to do one thing, but isn't true. To save on transistors, certain pins drive certain basic operations, like pin 10 is "this loads a value into a register." Additional pins will signal from where and to where. The 6502 CPU (which runs the Atari 2600) has "unsupported" operation codes. For instance, the $AF opcode causes a value to be loaded into both the A and X registers simultaneously. It crosses the gates of LDA (load A) and LDX (load X), and just so happens everything runs ok. Thus, $AF is nicknamed LAX; although, it's not a "real" opcode. But, that's not the end of it.
Second, a computer is not so much a digital device as it is a pulse-oriented voltage regulator. As voltages are applied to the pins of the CPU, various gates open and close. It takes time for the voltages to settle. Machines often have to have "debounce" their inputs (sometimes hardware, sometimes software) to wait until the voltages are steady before accepting their values. Voltages fluctuating at just the right time means the machine can do very odd things, such as miss executing and instruction or corrupting memory. Super Mario 3 can be beaten two seconds by changing the controller values so fast that the game loop waits "forever" on the voltages settling. Due to the way the SMB3's game happened to be coded as a series of "fall throughs" for game state, the game falls through all the game modes to the last one, the game ending.
I would find it entirely reasonable that an exactly timed controller input change causes an unexpected opcode execution.
Thank you for the comment. Dragster does not make use of any undocumented opcodes. Its input handling is also very straightforward, with the input state simply latched immediately at the end of a frame to be processed during the next frame. I don't know the exact mechanisms for how the RIOT coordinates input toggles, but from the perspective of the 6507 it's just a memory address, and is timed accordingly.
That aside, the major issue with assuming a single (or even multiple) input reads caused the speed to deviate enough is just that the mathematical progression is insufficient. Specific occurrences, such as avoiding the speed freeze on a shift, would have to happen 7 or 8 times before a 5.51 is even in the realm of possibility. The most beneficial game state mistake would be starting in first gear during the countdown, which would allow forward motion early, but the input related to that is both registered and occurs on the negative edge. The code paths simply do not allow an advance into first gear while simultaneously avoiding the countdown check.
This. I work for a very large company that builds CPUs, and with the amount of weird quirks you get wouldn't put anything beyond the realm of possibility. As one example out of many, lower than designed voltages can result in some instructions taking longer to complete than they should, and if their result is read before it has completed than an invalid value (previous, partial, zero, one, random, whatever) may be used in future calculations (this has been abused by researches to break SSL by using a brown out to slow down division instructions in modern CPUs). Instructions may also have undocumented side effects than an emulator may not take into account even if the instruction itself is documented (e.g. certain NOP instructions have side effects on some PPC server processors that e.g. affect the SMT priority), an interrupt (even an unhandled interrupt) may change some state in the CPU that has a subtle effect somewhere (you say that this game polls the inputs, but did the inputs trigger an interrupt in the CPU anyway or in any other way change the state in the CPU or caused a currently executing instruction to be slightly delayed / restarted?), and the fact is that a lot of these known hardware issues end up being worked around in software and many times in the compiler (e.g. adding a NOP or other instruction in some circumstances to delay reading a register until the VHDL simulations tell us it should be correct) since it is very expensive to change hardware designs and the game hardware companies I've worked with are very accepting of weird quirky hardware bugs being left in production so long as they have some way to work around them in software, and the majority of these bugs are never publicly disclosed - this is the sort of thing that emulators will miss that in 99% of cases won't matter... until someone writes some assembly by hand and hits one of these that works perfectly on an emulator, but not quite so on real hardware. And that is even assuming that the bug is even discovered before release - it is not at all uncommon for bugs to be discovered much later, or not at all.
I understand what you're saying (I'm a hardware engineer myself), but the code surface here doesn't really support many hypotheses related to quirky hardware behavior. The 6507 is very straightforward as far as CPUs go, and has limited state that can be affected even by spurious activity. Beyond that, the code is oriented in relatively long cascading logic paths; most paths that would grant some beneficial effect also have negative penalties unless the memory state (incorrectly) lines up in just the right way. It's not a case of just one thing going wrong. Multiple consecutive things would have to go wrong at consistent places in the code to even approach the possibility of a lower time, on top of optimal play.
If it's related to undocumented quirks in the CPU instruction set, then it's true that it would only be possible on real hardware, but the instructions used in this game's code are as fundamental as it gets. If those have unspecified behavior that affects memory or conditional logic states, then it would almost certainly be documented elsewhere, as those instructions are ubiquitous and mission-critical. I don't see any way in the code that one or even two misbehaving instructions would allow for a 5.51, regardless.
Imagine spending all that time typing all that out to look smart to be completely wrong because you failed to realise how the in game timer actually works
a guy made a full TAS of the game on an original atari 2600, it performs the same time (5.57) as an emulator. there was no way found with ANY input that allowed you to do what Todd “did”
Never attribute to malice what could be attributed to stupidity.
It’s possible that Todd legitimately believes he got a 5.54 because of either a hardware error in the 2600, a cartridge tilt (I don’t know enough about cartridge tilts so that might be easily disproven), dirty screen, manufacturing error in the cartridge console or TV, or just misremembering. He was nervous people wouldn’t believe him, so he made up the 2nd gear thing, which is still him lying, but the 5.54 might not be. Either way, this is very unlikely, and I still believe he’s lying, but it’s worth looking into.
Except numerous other scores he has aren’t possible as well.
This score is currently being challenged now at Twingalaxies
I would look at exact versions of Atari and the game that we’re out during the time he completed the record. It’s likely that the current version available for emulators no longer have the glitch that lets you start in second gear. Games in those days went through many revisions but seldom advertised this on the cartridge. So one person might have a rev 2 and another a rev 4 but you wouldn’t know it easily without a lot of effort. Edit: Turns out this was dug into to the bottom and it was a false record and two of twin galaxy’s employees were faking these records for years. Board is cleaned up now though.
"Why Can't This World Record Be Beaten?" it should be "Why Can't This World Record Even Be Achieved?"
BANNED FROM TWIN GALAXIES LUL
Fun fact: You could have watched the real Dragster world record 71 times during this video.
Cool video, horrible ending song.
ShortCircuiit Yeah
Yeah, horrible. Generic annoying robotic nasal voice, generic computer sounds, generic arrangement. But I guess some people like it....
VideoTape wat
yall want smth ßoppin
soundcloud.com/kore-wa-hentai/uwu
Each to their own, that track is dope.
I feel like he lied but a few witnesses went with it for either better publicity or maybe they were even bribed
Regan King
Or they read a 7 as a 1, depending on the screen/font that would be an understandable mistake.
WAtch Ben heck he is covering this very subject it might change your mind
all of his speedruns have a pattern
it's much higher than second place
it's approved by "referee"
it most likely is impossible due to a mechanic in the game
EDIT: oh and they're all fake
so true
if a wr time is way too high above everything else , i mean , if i claim i got a 3 minute 59 seconds and 707 ms time in super mario bros ( 707 since it reads lol if flipped upside down) no one would believe it , as it should be especially if i cant replicate it , and dont have video proof ( if you dont know smb , the 5 best times all are 4 minutes 56 seconds and different amounts of ms
yea, this is very true aswell
thanks
Apollo legends video was the video that finally pushed twin galaxies to do something.
I got a 5.21 one time by shifting into 3rd before turning on the atari, hope this helps
Brb gonna go spend 300 hours and beat his record
For a game as simple as Dragster, it should be easy to reproduce. If no one is able to even meet a 5.51 even under TAS conditions, the record should be revoked.
Todd gave himself the nickname of "The king of video games" not the media.
There is actually no proof on this time. We believe the best possible time is 5.57 and there is no recording of this time which means that this possibly was faked.
CosmicX it's believed that it's possible to get lower using different strategies. it's even theorized that a 3.xx is possible
Except it's not? That was just a theory of if those things happened, which is impossible
Didnt he give himself "the king of videogames" title?
Yep
Give back the Chaos Emeralds
No.
CHAOS CONTROL!
I got a time of 0.00 because I haven’t played the game.
my new plant manager at unilever is named Todd Roger lol thats just funny
This record was probably faked.
No 'probably' about it. He made it all up, and _shame on those_ who covered for him and supported his lies for all the years...
DarkAce indeed, *it's fake*
DarkAce It was. In the code, the highest score was 5.54. His referee at TwinGalaxies was one of his friends, who is now arrested for "playing" with a kid.
You can check out Apollo Legendz video on it.
ruclips.net/video/8e-H4sEHB54/видео.html
“Probably” lmao
@@touchofthorn1841 That comment was 2 years ago before it was proven faked.
“Don’t think he intended to lie”
Yea we know that’s not true now.
He’s a pathological liar.
UP NEXT: The Biggest Cheater in Gaming History? (1700+ Scores w/ no Evidence)
Excellent video. I'm pleased and impressed that you were able to dig up this information, even though there are still some holes from this taking place in the days before internet documentation. My gut feeling does say that Todd's time is probably due to some cartridge error, though without more information about the exhibitions of the run, there's definitely some doubts that can't be quelled.
With that in mind, though, I have an alternative proposal: Is it possible that there is more than one publicly released build of Dragster, that are not well differentiated, ID'd, or documented? I know, probably not; but if it turned out to be the case, it may better satisfy the problems surrounding this record (especially if more details and verification of the exhibitions happens to come up).
An alternate build could have the exploit Todd describes, or other, minute differences. Even if it was only produced in small numbers or distributed to a limited area, circumstance may make it more likely that all three runs were performed on that particular (theoretical) build, than of three separate carts having the same, specific error (if that were ever confirmed to be the setup). And it could explain why Omnigamer simply couldn't get a 5:51 out of the numbers or code: they were simply using the wrong build. How many people are really publishing dumps from their personal vintage game carts, let alone comparing them against the only known version? That build could be out there. The evidence, the transactions and documents that might suggest it isn't are probably long gone; and proving that something doesn't exist is a fundamentally flawed endeavor. It could take just one person dumping their cart and confirming it doesn't match the known version.
I'm not at all saying this is the likely answer. Misreading a 6 or using the same, bugged cart for all three runs, or even that Todd didn't actually match the 5:51 during those exhibitions (but his claim deemed reputable anyway) are all more likely possibilities. I'm just offering it as a remote possibility that hasn't been mentioned. And certainly, if a previously-unknown build was identified, that's what I'd put my money on.
Todd was was proven to have photo-shopped images of his score
Ron Jeremy would never lie
Surely there must exist some footage somewhere of him doing this. You don't just stand by and watch someone do a world record by inviting him down to your company and not record it.
There is not a single piece of evidence that can back this up. No witness that can confirm a 5.51 either.
The only "proof" Todd Rogers provided was a World Record Certificate that he claimed to have recieved from Activision. When this was posted, it was quickly shown that it was fabricated.
But no, nothing that can claim that a 5.51 either happened or that it is even possible in the first place, it's against the game itself.
He's also had both the CEO of Activision and the creator of the game say that he did it. He did it several times in front of people when he submitted the record. Even this video says he was tested in front of people. Proof no longer exists because no one expected gaming to take off like it has. It's like the early Dr. Who episodes. BBC didn't think anyone would care, and they recorded over the tapes. He was also scheduling to redo the record at a live event this year. Not the kind of thing a fraud does.
All of his scores have been removed and he is now banned anyway.
And if you look past Dragster, there are LOADS of scores that are even more clear that they are faked, Dragster being the most realistic of them. Have fun looking into it.
Congratulations, you can read articles. Glad you sidestepped my point, or the fact that the video even says that Todd confirmed his record at CES for Activision and later in front of Twin Galaxies. Both times using different machines. Or you can sidestep why a major console and game manufacturer decided to run with a fake score from some unknown kid that didn't mean anything to them or they could have simply rejected his claim. And then another company that didn't have any stake in it decided to back it up. Or how no rival has contested his skill over the last thirty-five years. Interesting how, with all of his records, no one has ever come forward or said anything.
I've taken part of the discussions since I brough it up at Twin Galaxies in May, after Omnigamer's research, and I've followed the discussion every day since, reading every comment on both sides of the arguments that I could find. So I haven't just read some articles =).
Yes, he did do runs in front of Activision, with David Crane present. However, in interviews with Crane, he couldn't say for sure that a time of 5.51 was actually achieved. Though as it was such a long time ago, it's hard to remember. It was never done in front of Twin Galaxies however. TG added the time way later, with the data taken from other sources. There is not a single witness that can verify that Todd did a 5.51 on his second public performance. The only source for this is Todd himself.
One reason to why Activision could have played along with this is that they, as Todd, ended up recieving a lot of publicity stunt money. With Todd becoming a paid gamer to travel around the US. It has been talked about, though not 100% confirmed I believe, that he sometimes was playing on prototype. Which is understandable when he showcased some of Activision's new games. No such prototype or other version differences have been found for Dragster however.
There have been several people suspicious of Todd's records throughout the years, this is far the first time. Most of the arguments have just been "Todd is a great player and a nice guy, he is legit." without there being any evidence for anyone to take part of. Todd is also known to have added his own records on the leadeboard, which he could thanks to being a referee. It's unknown how many records he did this with, but he has confirmed that it is true.
If you look into the game's code, a time of 5.51 is simply not possible under normal circumstance. No irregularities have been found to this and it has been compared to lots of Dragster games, on both console and emulator.
And, as mentioned, there are several other scores that also have been proven to be impossible. Many of which there have been dispute threads Twin Galaxies where this has been looked into. Many which could be found here: www.twingalaxies.com/forumdisplay.php/468-Score-Dispute-Review
Evidence-wise, there is nothing that talks in the favor of Todd. No proof exists for any of his records.
Toddzilla more like cheatzilla
I love watching this video and then your later videos on Todd.
dude looks like ron jeremy .
Holy smoke! It’s a scavenger hunt!
IDK dude I like his mustache I think he's legit
the trick is now to LISTNEN. - Todd Rogers
That's why no one could beat the record, they just couldn't LISTNEN.
A streamer named Darbian just got the record 2 hours ago with 5.57
5.51 is the record
captainnemo nope
We beat his record after all. Every single one of them at once🙌🏻
Fuck, even I beat every one of his records, without ever even having played any of the games he did
Played the hell out of this game as a kid. Heaps of 5.67 and I’m sure the odd 5.64 was best I ever did.
You guys just don't believe in the human element *kappa*
What are the odds for the only cartridge of the game with a hardware error like that would be in the hands of a guy with the intent to use it and already has 100 of "records". He should have rather played lotto instead :P
It's either fake or there was a sort of malfunction. But considering it's Todd Rogers we're talking about, then it's most likely fake.
So basically there is no way to break the record because it is due to a hardware malfunction?
fangtimes who said it was a hardware malfunction?
4:39 OP
Both outcomes are the same either way.
I once promised to just "put the tip in" ..but glitched into third and stole home with a blown drive shaft.
Aaaand it's gone.
this guy is from the same town i grew up in crazy
Who come here after, Darbian (5.57) Dragster speedrun (Atari) *World Record*...lollll
Because this 5.51 is fake
Because it’s an actual WR. With proof. And is legitimately possible.
It would be funny if 40 years from now someone comes along and runs a 5.51 after we all assumed that Todd was just a cheat.
It can’t be beaten because the game can’t actually get a 5.51
;D
This just got debunked. What a bummer he is a fraud
So, we're supposed to have Both Todd Rogers and Billy Mitchell at a con here in 2 months to discuss all of what's going on. This should be interesting
Atleast he have achieved something in his life, unlike me.
the only unbeatable speedrun is club penguin btw.
There was a game called Grand Prix for the Atari 2600. There were only 4 levels on it and the 3rd level was fast but very short. I think I had the fastest possible time on it because every time I played it (after memorizing the level) I got the exact same time.
Darbian beat the record 😂
"The King of Videogames" is self-titled btw.
This is beaten by darbian
Flip the vehicle, add a handle with a trigger. Boom. You got the raygun from call of duty zombies
His nickname used to be Mr. Activision, but now people just call him Fraud Rogers.