Circus Charlie (サーカスチャーリー Sākasu Chārī) is an arcade platform game originally published by Konami in 1984 in which the player controls a clown named Charlie. It was ported to the Nintendo Famicom in 1986 by Soft Pro.
This run plays Mode B, which is noticeably more difficult than Mode A on most stages, especially on Stage 3 and Stage 4.
Historically, "Mode A" obsoleted "Mode B" for this game here, as "it's harder to optimize Stage 1 in a TAS in Mode A".
The author of this submission considers that reasoning rather weak as:
After all, the movie rules don't care about how hard it is to make a TAS to be accepted and published here.
It's not that really hard. One can even make an optimized Stage 1 run at zero rerecording count.
Many TASes nowadays feature much more complicated techniques that can be more difficult to optimize.
As mentioned above, Mode B has undoubtedly harder Stage 3 and Stage 4 for a TAS, which require more planning than in Mode A.
Mode B is faster in time than Mode A when both are equally optimized, and arguably more entertaining.
Comments
Stage 1
The same trick to slightly adjust the positioning of the fire rings. Otherwise nothing special to mention.
Stage 2
Very simple and straight stage. No comments.
Stage 3
In Mode B the rolling balls are much more annoying to optimize. Yes they can be manipulated to some degree.
This seems to be the fastest solution achievable in this run.
Stage 4
In Mode A this stage is trivial, but in Mode B you have to carefully plan for the jumping patterns and even slow down once to survive.
Stage 5
The biggest improvemnt in and motivation for this run is here. Faster strategy and implementation.
Other comments
Thanks to all previous players for their amazing runs!
Additionally thanks to mtvf1 for the encode!
<klmz> it reminds me of that people used to keep quoting adelikat's IRC statements in the old good days
<adelikat> no doubt
<adelikat> klmz, they still do
Setting the level to anything at all for this mode doesn't break movie sync. However setting it to 99 in any level will reset it to 0 upon completion of that level.
I'm trying to find any trace of increasing difficulty after this movie ends, and so far there's none.
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.
Here's what I found.
Difficulty seems to be address $66. And it seems to affect gameplay differently for values below 5 (in mode B). Values from 4 to 0xFF don't break sync of my test movie.
Things I see at that difficulty:
Now, difficulty does overflow/reset after 0xFF. But the above looks very much like unique content to me. You can check how it plays after starting the game at mode B and cheating the address $66 to 04. All levels become fairly different, and also clearly harder.
Difficulty increments after each time you beat the game. When you start at mode A, difficulty starts at 1, and for mode B, it starts with 2.
So I think for this game, a movie should complete 3 loops of mode B, so difficulty goes 2, 3, 4, and then it freezes gameplay-wise, even though the value of the address keeps increasing after new loops.
Code that runs every frame to determine object spawns:
It checks if difficulty-1 is above 3, which indeed means difficulty caps out at 4.
NOTE:
If all the content of the difficulty 4 appears earlier, then it'd instead be the point to stop the movie after. I just haven't tested every difficulty thoroughly. But this video gives some idea, even though it's for mode A.
Link to video
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.
So I think for this game, a movie should complete 3 loops of mode B, so difficulty goes 2, 3, 4, and then it freezes gameplay-wise, even though the value of the address keeps increasing after new loops.
...
It checks if difficulty-1 is above 3, which indeed means difficulty caps out at 4.
NOTE:
If all the content of the difficulty 4 appears earlier, then it'd instead be the point to stop the movie after. I just haven't tested every difficulty thoroughly. But this video gives some idea, even though it's for mode A.
So does the above make this game an example of one where a future submission that completes all the proposed loops would obsolete the current publication even though the future run will be a longer time than the current?
Of course. We don't care how long it is if the relevant part is more optimal than the current run, and the rest of it is just as optimal.
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.
This movie has been published.
The posts before this message apply to the submission, and posts after this message apply to the published movie.
----
[5270] NES Circus Charlie by klmz in 03:21.35
Now, difficulty does overflow/reset after 0xFF. But the above looks very much like unique content to me. You can check how it plays after starting the game at mode B and cheating the address $66 to 04. All levels become fairly different, and also clearly harder.
Difficulty increments after each time you beat the game. When you start at mode A, difficulty starts at 1, and for mode B, it starts with 2.
So I think for this game, a movie should complete 3 loops of mode B, so difficulty goes 2, 3, 4, and then it freezes gameplay-wise, even though the value of the address keeps increasing after new loops.
Code that runs every frame to determine object spawns:
