(Link to video)
Submission Text Full Submission Page Publication Page

City Connection

City Connection is a side view driving game. The game was originally developed for the arcade by Jaleco in 1985, it later received a port on the NES in 1988. The story is fantastic, and oddly specific. You've broken into an exclusive paint store in New York City and are now on the run whilst carrying leaking 10 gallon cans of paint. You must inexplicibly drive over every mile of highway in order to make your escape.
Just another day in the city.

Game objectives

  • Emulator used: Bizhawk 2.2.2
  • Complete all six cities. True to the arcade style, the game loops indefinitely (the stage rolls from 255 back to 0). The backgrounds for the 6 cities rotate and do not reset when the stage counter rolls over. The platforms are shifted in later stages, but the discription of your world tour is in the game's manual and the progression of cutscene images between each stage in the North American version of this game suggests that the game should be considered complete after the sixth stage.
    • New York
    • London
    • Paris
    • Frankfurt
    • New Delhi
    • Tokyo

Game Mechanics

Controls

  • up - hold up when jumping to jump to a higher platform
  • left - turn left
  • right - turn right
  • down - makes the car fall down faster, doesn't work when jumping
  • B - shoot oil can
  • A - jump
  • select - select 1 or 2 players
  • start - pause

Items

  • Oil cans - You can collect and shoot these at police cars to take them out, but any extra left over are counted for extra points at the end of a stage, costing some time. So we want to end each level with no cans left (only acception is the last stage)
  • Balloons - collecting 3 balloons rewards you with a warp. The warp takes you to the next "stage" (increments the stage counter), but in the limited testing it did not seem to follow the normal progression of backgrounds. Additinally it is not used in this run since the warp animation is so long that it's faster to just finish the stage and sit through the cutscene.

Enemies

  • Police cars - you lose a life if you hit them, but you can take them out with oil cans
  • Cat - you lose a life if you run into a cat. Cats are impervious to oil cans
  • Spikes - these appear if you spend too long on the same level of platforms, but they're not seen in this run.

Strategies

The most difficult part of this TAS is the routing. The route is simple enough that stage by stage details probably aren't necessary, but the high level strategy is to keep turning and changing platform levels to a minimum. It turns out that the best way to achieve this is to work from top down, finishing each layer as you go. If platforms lined up in the right way it may be useful to change levels in some situations, but there were no cases like that in this run.
If you hold up and jump, you can jump towards the next layer up, but if there is a platform in your way you'll bonk against the ceiling. It turns out that you gain control to turn around 5 frames earlier when you bonk into a ceiling, so this is used as much as possible to save time.
You can gain a little extra hang-time during a jump by pausing and unpausing at the right time. This is used to avoid wasting time switching platform levels in New Delhi. Credit to L33Tz0rs (ツ ツ) on youtube who used this strategy in his good, but slightly slower TAS of this game.

Improvements

  • Route testing was by no means exhaustive, so I wouldn't rule out the possibility of saving a few frames with an improved route, but I think what we came up with is about as good as it's going to get.
  • Better enemy manipulation may result in the ability to end input slightly earlier, but should end up with the same real end time.

Thanks to:

  • EZGames69 for the collaboration and for suggesting this project.
  • LackAttack24 and other RTA runners for their interest in the project
  • my Twitch chat for the ideas and encouragement
  • Takanawa for his TAS and for showing us the down button strat.
  • The TASMania team

Screenshots:

372, 4204, 7280, 10862, 12619, 15117
Memory: Claiming for judging
Memory: Replacing file with switch to Japanese version that saves 5 frames with no changes in gameplay.
Memory: Replacing with improvement that reaches the final cutscene 304 frames faster.
Memory: After the implementation of pressing down to fall faster, this TAS seems fairly optimized. While the game does go on pretty much indefinitely, it does have a clear story ending which was reached in this TAS as indicated by the manual, completing Tokyo.
While the improvement seems to have received a little better reception than the original TAS, it still is on the bubble. The TAS is fairly repetitive outside of the one pause trick.
With the road this TAS has painted, I am accepting this to Vault.
feos: Oh hi. I've heard this movie's branch isn't obvious. Let's handle it as I publish.

Show more

Hide
1 2
Posted: 6/30/2018 2:40 AM
Quote
link_7777
Experienced player (853)
Joined: 11/15/2010
Posts: 267
The spawns are based on a spawn timer, and I didn't see any sort of difficulty increase or variance with the spawning. As for the warping I think that remains the same. The warp animation is so long that really the only thing that would be shorter in a warps run would be in game time. I personally like the idea of the faster real time warpless run, but I think there is definitely room for both if someone wanted to make a warps run.
Posted: 6/30/2018 3:25 AM
Quote
mtvf1
Skilled player (1151)
Joined: 5/11/2011
Posts: 425
Location: China
GBA TASer of 2013
link_7777 wrote:
The warp animation is so long that really the only thing that would be shorter in a warps run would be in game time. I personally like the idea of the faster real time warpless run, but I think there is definitely room for both if someone wanted to make a warps run.
But warps run is still faster than warpless in movie time. Though using the warp, it didn't skip any whole stage. Because of the long animation, in most stages, it still need be completed without warp. So I think warp is worth to obsolete warpless. In fact, in this game warp is too much more difficult than warpless, if do the 15 stage. It should find the best way to collect items, and this game has the different items in different city.
Posted: 6/30/2018 4:56 AM
Quote
link_7777
Experienced player (853)
Joined: 11/15/2010
Posts: 267
mtvf1 wrote:
link_7777 wrote:
The warp animation is so long that really the only thing that would be shorter in a warps run would be in game time. I personally like the idea of the faster real time warpless run, but I think there is definitely room for both if someone wanted to make a warps run.
But warps run is still faster than warpless in movie time. Though using the warp, it didn't skip any whole stage. Because of the long animation, in most stages, it still need be completed without warp. So I think warp is worth to obsolete warpless. In fact, in this game warp is too much more difficult than warpless, if do the 15 stage. It should find the best way to collect items, and this game has the different items in different city.
I'll have to keep an eye out on the later stages in case one gets out of hand, but based on my testing it is pretty unlikely that a warps run would be faster than a warpless run in movie time. If you were to warp from stage 0 it drops you in stage 1 not stage 2. The warp animation is so long that it would only be faster in two cases. First if a stage ran long and you could grab earlier balloons without wasting much time. Second if you go for in-game time and are not counting the time for the warp animation. I do agree that a warps run would be very different.
Posted: 6/30/2018 5:35 AM
Quote
mtvf1
Skilled player (1151)
Joined: 5/11/2011
Posts: 425
Location: China
GBA TASer of 2013
link_7777 wrote:
I'll have to keep an eye out on the later stages in case one gets out of hand, but based on my testing it is pretty unlikely that a warps run would be faster than a warpless run in movie time. If you were to warp from stage 0 it drops you in stage 1 not stage 2. The warp animation is so long that it would only be faster in two cases.
Compare the initial of 4 stages, Takanawa's wip(the 4th stage is Tokyo) is 1382 frames faster than published(the 4th stage is Frankfurt).
Posted: 6/30/2018 5:47 AM
Quote
link_7777
Experienced player (853)
Joined: 11/15/2010
Posts: 267
mtvf1 wrote:
link_7777 wrote:
I'll have to keep an eye out on the later stages in case one gets out of hand, but based on my testing it is pretty unlikely that a warps run would be faster than a warpless run in movie time. If you were to warp from stage 0 it drops you in stage 1 not stage 2. The warp animation is so long that it would only be faster in two cases.
Compare the initial of 4 stages, Takanawa's wip(the 4th stage is Tokyo) is 1382 frames faster than published(the 4th stage is Frankfurt).
I think you are confusing the stage number with the background number, but they are tracked independently (background in 0xA0, stage in 0xA1). Strangely warping increments the background counter by 2, and the stage counter by 1. So if you warp from stage 0 background 0 you end up in stage 1 with background 2. You're thinking that you would warp from New York to Paris skipping London, but but actually you end up playing what should be the London platforms with a Paris background. Since the ending criteria is about stages not backgrounds the warps are a bit less helpful than they appear.
Posted: 6/30/2018 5:57 AM
Quote
mtvf1
Skilled player (1151)
Joined: 5/11/2011
Posts: 425
Location: China
GBA TASer of 2013
link_7777 wrote:
mtvf1 wrote:
link_7777 wrote:
I'll have to keep an eye out on the later stages in case one gets out of hand, but based on my testing it is pretty unlikely that a warps run would be faster than a warpless run in movie time. If you were to warp from stage 0 it drops you in stage 1 not stage 2. The warp animation is so long that it would only be faster in two cases.
Compare the initial of 4 stages, Takanawa's wip(the 4th stage is Tokyo) is 1382 frames faster than published(the 4th stage is Frankfurt).
I think you are confusing the stage number with the background number, but they are tracked independently (background in 0xA0, stage in 0xA1). Strangely warping increments the background counter by 2, and the stage counter by 1. So if you warp from stage 0 background 0 you end up in stage 1 with background 2. You're thinking that you would warp from New York to Paris skipping London, but but actually you end up playing what should be the London platforms with a Paris background. Since the ending criteria is about stages not backgrounds the warps are a bit less helpful than they appear.
I know these RAM address, and I didn't make this mistake.
Posted: 6/30/2018 6:30 AM
Quote
link_7777
Experienced player (853)
Joined: 11/15/2010
Posts: 267
mtvf1 wrote:
link_7777 wrote:
mtvf1 wrote:
link_7777 wrote:
I'll have to keep an eye out on the later stages in case one gets out of hand, but based on my testing it is pretty unlikely that a warps run would be faster than a warpless run in movie time. If you were to warp from stage 0 it drops you in stage 1 not stage 2. The warp animation is so long that it would only be faster in two cases.
Compare the initial of 4 stages, Takanawa's wip(the 4th stage is Tokyo) is 1382 frames faster than published(the 4th stage is Frankfurt).
I think you are confusing the stage number with the background number, but they are tracked independently (background in 0xA0, stage in 0xA1). Strangely warping increments the background counter by 2, and the stage counter by 1. So if you warp from stage 0 background 0 you end up in stage 1 with background 2. You're thinking that you would warp from New York to Paris skipping London, but but actually you end up playing what should be the London platforms with a Paris background. Since the ending criteria is about stages not backgrounds the warps are a bit less helpful than they appear.
I know these RAM address, and I didn't make this mistake.
Ah, so I looked through that movie and it looks like I need to learn more about how the spawns work. The manual talks about getting an extra balloon spawn in Paris, which is what Takanawa does. He also ends up with an early spawn, and my statements about the warp being slower are based on the first normal balloon spawn when the spawn counter hits 0x700. I'll probably need to dig through the code more and see what else is possible
Posted: 6/30/2018 9:27 AM
Quote
feos
Site Admin, Skilled player (1237)
Joined: 4/17/2010
Posts: 11276
Location: RU
NES TAS of 2011
Yes, please redo this so that we don't pretend this movie is properly completed. One of the reasons it was accepted in the first place is that we trusted the author's research. And it appeared to be sloppy. I think we'll have to ignore this factor in the future.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Posted: 6/30/2018 11:40 AM
Quote
Alyosha
He/Him
Editor, Expert player (3537)
Joined: 11/30/2014
Posts: 2733
Location: US
GBA TASer of 2023 Homebrew TAS of 2023 Gameboy TASer of 2022 NES TASer of 2015
feos wrote:
One of the reasons it was accepted in the first place is that we trusted the author's research. And it appeared to be sloppy.
feos wrote:
Yeah, what link_7777 said. We did the due research.
(See here for second quote context.) Mistakes aren't a big deal, but it seems a bit unnecessary to throw the author under the bus after the fact. A simple 'we goofed' would have been sufficient.
Posted: 6/30/2018 11:44 AM
Quote
feos
Site Admin, Skilled player (1237)
Joined: 4/17/2010
Posts: 11276
Location: RU
NES TAS of 2011
We goofed by fully trusting the authors. This is not about blame, but about policies that weren't perfected, seeing the actual reasons of problems and trying to resolve them. I noticed one more thing. In a film about Deep Purple, the band said that managers operate in a certain way. They say "we succeeded" when the band succeeds with their help, but say "you failed" to the band when things go wrong. I guess I made the same mistake here: I said "we" when it was about how well the game was (supposedly) researched, and then I said "the author" when it turned out that it was sloppy. Sorry about this.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
1 2