Submission #7495: ShesChardcore's A2600 Math Gran Prix in 00:49.93

(Link to video)
Atari 2600
baseline
BizHawk 2.8.0
2992
59.9227510135505
3799
PowerOn
Submitted by ShesChardcore on 5/20/2022 6:47:31 PM
Submission Comments
This submission is Math Gran Prix for the Atari 2600. You are tasked with solving math problems to advance your car down a track, in an effort to beat the other car. We do this on the hardest difficulty in under 50s.

Game objectives

  • Emulator used: Bizhawk 2.8
  • Aims for fastest time
  • Uses hardest difficulty
  • I got 141 and 2/3 chance of winnin

Comments

While this game looks simplistic, there are numerous manipulations happening throughout the TAS. Any "missing" frames or gaps are intentional to carve a specific path. We manipulate everything from answers to AI movements to the outcome of the "spinner." We use Game 9 which is the hardest difficulty and uses all 4 math equation types.
As a general rule, you must never fall behind the opposing car. It is very aggressive and will knock you off the track if it can, and if it remains in front of you it'll be impossible to catch up. You can move 2 or 3 squares per turn, but so can the AI. For this reason we need to build a big lead.
To do this, the AI can be manipulated to choose 2 for the first 3 turns based on what frame we choose our own move and what frame we execute it. However, this stops working after the third turn, probably due to some rubber banding logic with the AI. Regardless, we build enough of a lead on the first 3 turns to where we don't have to worry.
Equations can be manipulated as well, and we obviously want the outcomes to have as few inputs as possible. Equations are on a sort of a global timer window. Within each window there are two equations, alternating frames between them until the next window hits. There is also some "tech" you can use to save time in some spots.
There are weird mechanics when it comes to inputs. For instance, to lock in an answer you have to hold the button for X number of frames and it doesn't have to be consecutive, it can be cumulative. In some cases I can buffer button inputs so that the move executes right after I hit the desired number, saving some frames. We can also use buffering and skipping a frame here and there to get different equations or manip the AI car.
There are special tiles on the track. The garage/house thing protects you from being knocked off the track. The X gives you an extra turn if you land on it and the lowercase i assigns you a random bonus move between 1 and 4. Through our path we manip the AI to roll a 1 and ourselves to roll a 3 near the end which sets us up to win on the following turn.
This was more complex than I expected, but I'm happy with the results.

Samsara: Sick, I love math. Judging.
Samsara: You know they say all TASes are created equal, but you look at Prix and you look at G.I. Joe, and you can see that statement is not true! See normally if you go one-on-one with another TAS, you got a 50/50 chance of winning, but Prix's a genetic freak, and it's not normal! So you got a 25% at best, at beat Prix! And then you add Cobra Triangle to the mix? You- The chances of winning DRASTIC go down! See the 3 way, at TAS-rifice, you got a 33⅓ chance of winning! But Prix! Prix's got a 66⅔ chance of winning, 'cause Cobra Triangle KNOWS he can't beat Prix, and he's not even gonna try! So G.I. Joe, you take your 33⅓ chance, minus my 25% chance, and you got an 8⅓ chance of winning at TAS-rifice! But then you take Prix's 75 perchance chance of winning, if Prix was to go one-on-one, and then add 66⅔ cha- percents... Prix's got a 141⅔ chance of winning at TAS-rifice! Senior Joe! The numbers don't lie, and they spell acceptance for you, at TAS-rifice!

despoa: Processing...
Last Edited by despoa on 6/11/2022 9:14 AM
Page History Latest diff List referrers