Jump to content

Photo

Microsoft's AI Just Shattered the Ms. Pac-Man High Score


19 replies to this topic

#1 Flojomojo OFFLINE  

Flojomojo

    I say boom boom boom

  • 9,663 posts
  • You say bam bam bam

Posted Wed Jun 14, 2017 12:35 PM

http://sploid.gizmod...core-1796091352

 

The researchers chose Ms. Pac-Man as it was written to be far less predictable than the original version of the game, which will help the AI they developed better deal with real-world situations that are equally unpredictable. Because as impressive as this feat is, few companies are willing to pay for something that just sits around and plays video games all day.

Update, 2:14 p.m. EST/EDT: Microsoft has reached out to us to clarify that the high score for the Atari 2600 version of Ms. Pac-Man is 266,360 points.

 

 



#2 hepcat OFFLINE  

hepcat

    Dragonstomper

  • 510 posts
  • Location:Washington, DC

Posted Wed Jun 14, 2017 1:01 PM

Kind of cool, I guess, but I doubt the all time human high score on 4 ghost 2600 Ms. Pac-man is a mere 266K. The game is not that hard; I'm pretty sure I broke 300K on it back in high school. Rolling it would be tough with only 4 lives, but I bet someone has done it before Microsoft's AI.

 

Edit: Yup, I just checked the 2600 high score club results and the high score on Ms. Pac-man is a million... 


Edited by hepcat, Wed Jun 14, 2017 1:06 PM.


#3 Flojomojo OFFLINE  

Flojomojo

    I say boom boom boom

  • Topic Starter
  • 9,663 posts
  • You say bam bam bam

Posted Wed Jun 14, 2017 1:11 PM

Gizmodo botched the story, was fundamentally confused about the differences between arcade Pac-Man, Atari 2600 versions, and Ms. Pac-Man. 

 

Here's a better story

https://www.wired.co...-ai-ms-pac-man/

 

and here's the source.

https://blogs.micros...pzrfk1b5haiqep0

 

If the maximum score is a million, all they can do is hit it. Shattering won't happen. 



#4 NE146 OFFLINE  

NE146

    Dumbass Atari Fan

  • 14,762 posts
  • Location:Seattle, WA

Posted Wed Jun 14, 2017 1:38 PM

Lol I don't know if this is the case but it'd be pretty sad if you don't conceptually already understand that arcade ms pac and our beloved VCS port are pretty much entirely different games. It's almost like they grew up in a different era or something!! :lol:

Regardless though both games get stingy with the scoring once you hit the 300k mark or so.. you can't get any points from ghosts, but also unlike Pac-man you don't get 5000 pt bonuses consistently. Although I don't think the vcs ever gets to a point where the ghosts don't turn blue... do they?

Edited by NE146, Wed Jun 14, 2017 1:40 PM.


#5 Nathan Strum OFFLINE  

Nathan Strum

    Quadrunner

  • 7,269 posts
  • Enjoying a sandwich
  • Location:Newhall, CA

Posted Wed Jun 14, 2017 2:24 PM

Shouldn't Microsoft be spending their time fixing their own software instead?



#6 hepcat OFFLINE  

hepcat

    Dragonstomper

  • 510 posts
  • Location:Washington, DC

Posted Wed Jun 14, 2017 3:35 PM

Lol I don't know if this is the case but it'd be pretty sad if you don't conceptually already understand that arcade ms pac and our beloved VCS port are pretty much entirely different games. It's almost like they grew up in a different era or something!! :lol:

Regardless though both games get stingy with the scoring once you hit the 300k mark or so.. you can't get any points from ghosts, but also unlike Pac-man you don't get 5000 pt bonuses consistently. Although I don't think the vcs ever gets to a point where the ghosts don't turn blue... do they?

 

If I remember correctly, on the 2600 version you get a one second blue time on every board 8 and higher. It never drops to zero. Also, the 2600 Ms. Pac seems less stingy with bananas on the random fruit levels.



#7 negative1 OFFLINE  

negative1

    Chopper Commander

  • 125 posts

Posted Wed Jun 14, 2017 6:16 PM

i'm impressed. after playing the 2600 version, it definitely has a difficulty curve (not as hard as the arcade though).

 

would like to see more video of the gameplay though.

 

later

-1



#8 fiddlepaddle OFFLINE  

fiddlepaddle

    River Patroller

  • 2,286 posts

Posted Wed Jun 14, 2017 6:37 PM

Hey! Maybe driverless Ubers will spend their down time playing video games and re-popularize the arcade!

#9 xucaen OFFLINE  

xucaen

    Stargunner

  • 1,001 posts
  • Moving and enrolled in MCSD certification classes.
  • Location:Massachusetts

Posted Thu Jun 15, 2017 7:12 AM

This means it now knows how to evade capture and hunt humans I mean ghosts... o_O

#10 Flojomojo OFFLINE  

Flojomojo

    I say boom boom boom

  • Topic Starter
  • 9,663 posts
  • You say bam bam bam

Posted Thu Jun 15, 2017 7:43 AM

Shouldn't Microsoft be spending their time fixing their own software instead?

 

Heh. That would imply they're one big happy family. 

microsoft-org-chart.jpg



#11 rdemming OFFLINE  

rdemming

    Stargunner

  • 1,202 posts
  • Location:The Netherlands, Amstelveen

Posted Thu Jun 15, 2017 11:38 PM

The complete article on the developer's website (Maluuba) website (follow the link on the bottom of the page to continue reading).

 

Also available is the scientific article and finally the video explaining the algorithm:

 

 



#12 Kosmic Stardust OFFLINE  

Kosmic Stardust

    Princess Rescuer

  • 15,123 posts
  • Location:Milky Way Galaxy

Posted Fri Jun 16, 2017 5:08 AM

This is really cool, thanks for staring. Perhaps someone invents a "Johny 5" type approach where the bot interfaces the game capturing light and sound from a CRT telivision and servo actuated CX-40 joystick performs the in-game inputs, using the same feedback loops a human would. That would be awesome!



#13 negative1 OFFLINE  

negative1

    Chopper Commander

  • 125 posts

Posted Mon Jun 19, 2017 4:46 PM

if there was more video, it might be possible to see if there are patterns for the 2600 version.

 

given that the ghosts don't go in random directions for the levels (dont remember if they do or dont)..

 

there is a 266k replay here: 

 

later

-1



#14 tschak909 OFFLINE  

tschak909

    Stargunner

  • 1,793 posts
  • Location:USA

Posted Mon Jun 19, 2017 6:01 PM

Am going to guess before reading, and say they used a fluid diffusion model (model the maze as a three dimensional model of fluid going through trenches, with certain parts of the maze weighted. Similar to what Norbert Kehrer did when creating an algorithm to automatically place the direction arrows for his automatic mazes...am guessing that a markov model is then built from this fluid model and traversed, as a base-line)

 

-Thom



#15 tschak909 OFFLINE  

tschak909

    Stargunner

  • 1,793 posts
  • Location:USA

Posted Mon Jun 19, 2017 6:09 PM

Ah close, Definitely a parallel Markov traversal like I thought, but with a significant amount of high level domain specific knowledge in each Markov thread, with extreme care to make each thread take care of a specific aspect of a given reward... The algorithm seems to have a high vector factor, and therefore scales quite nicely. 

 

The domain specific knowledge describing the reward is the key here, the rest is the grind-work to approach the reward.

 

-Thom

 

(p.s. forget about the fucking articles, they're written by morons, AND ACTUALLY READ THE PAPER!, where the actual data is...it's far more insightful)

 

https://arxiv.org/pdf/1706.04208.pdf



#16 Nathan Strum OFFLINE  

Nathan Strum

    Quadrunner

  • 7,269 posts
  • Enjoying a sandwich
  • Location:Newhall, CA

Posted Mon Jun 19, 2017 7:35 PM

Now let's see them do Jr. Pac-Man. :D



#17 Aquadump Entertainment OFFLINE  

Aquadump Entertainment

    Combat Commando

  • 3 posts

Posted Tue Jun 20, 2017 9:46 AM

Yeah, but can it solve the riddle of the Sphynx?



#18 negative1 OFFLINE  

negative1

    Chopper Commander

  • 125 posts

Posted Tue Jun 20, 2017 12:02 PM

jr pac man would also be better test, but the thing is, since they've got a solution for ms pac man,

 

it should theoretically also work for all pac-man type games.

 

and most maze games in general.

 

later

-1



#19 Kosmic Stardust OFFLINE  

Kosmic Stardust

    Princess Rescuer

  • 15,123 posts
  • Location:Milky Way Galaxy

Posted Wed Jun 21, 2017 12:50 AM

jr pac man would also be better test, but the thing is, since they've got a solution for ms pac man,
 
it should theoretically also work for all pac-man type games.
 
and most maze games in general.
 
later
-1

I'd like to see any human or AI crack one million on Jr Pacman 2600. The ghosts intelligence and speed are merciless on that game. Even on teddy bear mode (one ghost) I still can't unlock all the mazes... :dunce:

#20 aofl OFFLINE  

aofl

    Combat Commando

  • 4 posts

Posted Sun Jun 25, 2017 5:03 PM

I'd like to see them apply similar research to arcade Donkey Kong which IMHO would be much harder to make AI for since the goal isn't grinding through board after board but maximizing points until the kill screen. 

 

I wonder if an AI can beat the 1.2 million point human record.

 

A






0 user(s) are browsing this forum

0 members, 0 guests, 0 anonymous users