Actions
Engagement
  • Author Author: makerkaren
  • Date Created: Date Created
  • Last Updated Last Updated:
  • Views 901 views
  • Likes 5 likes
  • Comments 4 comments
Related
Recommended

Atari Dragster Speed Test Record - Man vs Computer

A while back, Ben received an email from a gentleman about this old speedrun record for the game Dragster on the Atari 2600. I had never heard of this particular record, so I asked Jace Hall, CEO of Twin Galaxies, for the back story.  From Jace:

 

On 1 September 1982, Todd Rogers (USA) set a time of 5.51 sec on the Atari 2600 version of Activision’s Dragster (1980), as verified by Activision and later acknowledged by Twin Galaxies. That early GWR gaming record remained unbeaten for 34 years 215 days, as of 4 April 2017.

 

This record has had the prestige and honor of being the longest recognized standing video game world record. It is also believed by many that it is the 1st official "Speedrun" record.

 

For more than 3 decades, no other person has been able to beat or even tie Todd's record time of 5.51 and this has resulted in some widespread skepticism that perhaps this record may be fraudulent. The record holder, Todd Rogers, maintains that his record is valid, and that he has accomplished the feat on at least 3 separate occasions.

 

Recently, some members of the Speedrun community decided to reverse-engineer and directly examine the Dragster game code to see what kind of joystick inputs and timing would be required to produce a time of 5.51 sec. The extensive software analysis was unable to determine how to accomplish the time, and concluded that achieving a time of 5.51 is not possible. The software analysis has suggested that the maximum achievable time is limited to 5.57 sec.

 

Based on the software analysis results, a formal Twin Galaxies dispute was filed by *** Moreland on 11-August-2017 calling for Todd Rogers' score to be removed. The software analysis evidence was compelling enough for Twin Galaxies to begin a thorough and comprehensive investigation into the matter, which not only includes looking at the software analysis results provided but also includes performing necessary and extensive Atari 2600 hardware execution tests as well.

 

As part of the Twin Galaxies investigation into the hardware execution analysis, we reached out to Ben as an independent 3rd party to investigate whether or not a 5.51 sec time can be actually achieved using an original Atari 2600 and a Dragster cartridge -- exactly the same tools that Todd Rogers used to achieve his time.

 

When Jace reached out to Ben, Ben was not much interested in the speed record.  However, Ben was fascinated by the idea of using new hardware to analyze the RAM of the 2600 – to copy the data being sent to the RAM from the CPU.  The idea is that data can be analyzed in another microcontroller to figure out how to send the most optimal Atari 2600 Dragster game inputs possible. Ben got to work creating a RAM emulator to program the computer to play the fastest game possible, not an easy task. Ben's been using the  MDO3104MDO3104 Oscilloscope from Tektronix to dig into the signals of the game to refine his program down to the frame.

 

Things are getting pretty technical here in the shop and Ben is getting closer...

 

 

imageimage

  • I love the idea of using new tech to read the Atari RAM addresses and data, but the results could be replicated with a TASbot like DwangoAC’s. Would’ve taken much less time, but then again where’s the challenge kn replicating Tech?

  • I am very curious where this goes next. I read a lot of comments about this being a false record. Let's have it proven by computers! Similar to those computer 'judges' in Buck Rogers.

    This series is already a classic in my opinion.

  • It should be noted there is a long history with Todd's records being 'suspicious' to say the least. Back when I used to volunteer as a ref for Twin Galaxies around 2002, I suspected one of Todd's records on another Atari 2600 game called "Barnstorming" of being physically impossible. I started raising concerns about the record when it became obvious nobody else in the world could even come close to matching it. I was met with scorn from Todd's loyal fan base, but nevertheless I pushed forward in my pursuit of the truth. Eventually I was able to make my case by hacking the code of the original game. What I did was remove all obstacles from the 'track' Todd's record was on, and then I started the game and flew the plane straight to the finish line. Now how the game works is you lose flight speed having to ascend and descend the plane in order to avoid wrecking into various obstacles like barns, birds and other structures. The less you have to move your plane up or down, the faster your time will be at the finish (assuming you don't hit something). In my hacked version of the game with no obstacles, my flight time came in a full half-second slower than Todd's 'official' record. At this point, the evidence could not be denied. The referee (one of Todd's friends) finally confessed he had never actually witnessed Todd do the record or even seen video documentation of it. Instead, he took the record from an Activision magazine that Todd submitted scores to, and claimed the record was incorrectly tracked due to a coffee stain smudging the entry in the magazine. Now keep in mind that until I PROVED the record was bogus, this information was not made public, and Todd did nothing to step forward and correct the entry, which had stood for decades. He was made fully aware of the situation, with his fans viciously attacking me for daring to question his record. "You just don't know how good he is.", "You'll be eating your hat when you see him do the record live", etc. were typical of the less offensive flame responses. I would often retort: "I will gladly eat my hat to see actual proof on a fair playing ground. If that's what it takes to have a fair and unbiased video game record database, pass me the ketchup!" 

    This was 15 years ago. I've never had to take one bite of my hat...


    In the case of the Dragster record, Todd has gone so far as to claim he "started the race" in 2nd gear. Subsequent investigation of the game code has shown that it does not allow for starting in 2nd gear. There's speculation that Todd may have been glitching the game code by physical means, either with the cart or introducing electricity into the difficulty switches. If any of that turns out to be the case, his record would instantly be disqualified.

  • I owned Dragster for the Atari 2600.

     

    Actually learned how to drive a manual transmission from this game at a very young age.

     

    Albeit this is not the best way to drive in real life, the reaction from parents is priceless.