Posted: 2/5/2009 11:55 PM
(Edited: 2/10/2009 5:34 AM)
Post subject: Chu Chu Rocket! Stage Challenge
Quote
applepieman
Former player
Joined: 10/6/2007
Posts: 86
I've recently been messing around TASing myself. So I started messing around with Chu Chu Rocket! (CCR). What I am planning is to go through all 25 levels of the Stage Challenge mode. Unlike the puzzle mode where you just set the pieces then start the stage, you have 30 seconds to accomplish the goal at hand. There are 5 types of goals, each set to a letter (A, B, C, D, E) A being where you have to get all the mice into the rockets. B being A, but with Cats on the stage C being feed the Cats (get all the mice into the cats) D being getting 100 mice into the rockets E being Cat Soccer (you vs a computer), similar to multiplayer mode but no mice At the beggining, I go through the option menu to change the controls from the standard (where you have to press the A or B button multiple times to spin it) to Pro mode where all you have to do is press A for Up, B for Down, L for Left, and R for right. Currently have A1, B1, and C1 finished. In this WIP I have D1 finished with 7.05 seconds, but using this guide on gamefaqs (http://www.gamefaqs.com/portable/gbadvance/file/467919/15240), for a point in reference, I can't seem to get 6.91. Of course this is where numerous re-records took place. The D levels need a lot of planning, testing, for the optimal time, which I've been attempting for a bit now (no luck as of yet). http://dehacked.2y.net/microstorage.php/info/117850449/Chu%20Chu%20Rocket%21.vbm I know its not much as of now, but tell me what you think? EDIT: Progress:
Posted: 2/6/2009 12:40 AM
Quote
Mr._Pwnage
Joined: 6/6/2004
Posts: 223
Just to point out, I wrote most of that guide 7 years ago and the strategies have superseded me in most stages. Here's the current record list: A1 05.43 B1 03.70 C1 03.46 D1 06.40 E1 11.10 A2 09.55 B2 06.08 C2 03.21 D2 09.11 E2 14.58 A3 04.13 B3 04.56 C3 06.81 D3 12.53 E3 20.28 A4 07.81 B4 08.46 C4 05.26 D4 08.68 E4 13.63 A5 15.61 B5 03.70 C5 09.25 D5 16.65 Fin 02.00 Furthermore, much of this game is completely deterministic, with the mice starting in known positions and moving one square every 13 frames, the cats moving at two-thirds that speed, and the cursor capable of moving at three-halves the speed of mice. Unless you can find the random generation algorithm and abuse it to pull off fantastic manipulation in the Ds and final, I don't think this is a very worthy TAS game. For Ds you'd obviously want to choose one or two hatches to harvest, get all 100 mice to come quickly from only those hatches, and have the cat be the only thing coming out of the rest (repeatedly if necessary). I'd think the fastest way to do final would be to have a gold mouse come out of the top right hatch at start (but delayed just long enough that you can get your cursor in there), followed immediately by a mystery mouse which gives Speed Up, and another gold mouse to run in at the faster speed. Anyway, if stuff like that isn't possible, any such TAS will probably end up as gruefood.
Posted: 2/6/2009 12:52 AM
Quote
applepieman
Former player
Joined: 10/6/2007
Posts: 86
I'd figured the times were outdated but I needed a game to wet my taste in TASing. So I chose something simplistic. I was just curious on some tips and what not. Cause if you chose Puzzle Mode, you'd just place the tiles where needed and watch the game play itself. Also thanks for listing the records. I'll probably go ahead and finish the run just for something to show for it on youtube. Not submit it or anything though.
Posted: 2/9/2009 9:35 PM
Quote
applepieman
Former player
Joined: 10/6/2007
Posts: 86
Well I've been working on it and definitely not going to break any records. I try each stage numerous times to see what kind of times I can get, if I seem to hit one time over and over again, I just stick with that and move on since this isn't going to be a submitted run. At the beginning of D3 http://dehacked.2y.net/microstorage.php/info/1434853138/Chu%20Chu%20Rocket%21.vbm At D2 I exit the stage and re-enter it to manipulate where the cat spawns, if you go frame by frame, I waited about 40 frames at the start of D2 (before the pause to restart it) and that determined where the cat spawned. It seems 10 frame intervals or so determine where the cat starts. Not to sure though. (If true, I believe using this tactic on D1 might help a bit as well) Times in the run A1 6.08 B1 3.91 C1 5.78 D1 7.05 E1 12.70 A2 12.58 B2 10.20 C2 4.73 D2 11.16 E2 18.70 A3 5.00 B3 5.86 C3 8.35 EDIT: Just got this time for D3 so whoopy. (Not included in the .vbm) D3 18.61
Posted: 2/10/2009 6:45 AM
Quote
Mr._Pwnage
Joined: 6/6/2004
Posts: 223
D3 was one of the levels I found a better route for while I was still playing, but after the most recent guide edit. Basically you want to have the cat run harmlessly across row 4 while you harvest mice; manipulate the cat to come out of the row 3 hatch first thing, then place K3-D, J1-D, J2-U, and then J3-U after the cat passes by. Once 100 mice get on the field and they stop spawning, drop K1-D (assuming the cat's not in the way), then J4-D and the rest is obvious. That should get to the 13s easy (assuming you get a fast row 3 cat), and 12s with enough effort and finding a way to make mice spawn quickly. Also, a TAS strat for D5 will involve the cat spawning multiple times and you'll want them all on the same hatch. How might your start delay cope with such a thing?
Posted: 2/11/2009 2:38 AM
Quote
applepieman
Former player
Joined: 10/6/2007
Posts: 86
Thanks for the tip. I'll go back and try it out later to see if I can improve that time. Until then these are times I've got for E3-B4 E3 25.98 A4 7.81 B4 9.11 Check it out here
Posted: 2/11/2009 10:33 PM
Quote
applepieman
Former player
Joined: 10/6/2007
Posts: 86
Using your strategy, I've improved the time to 16.90 for D3. I have it to where the Cat spawns on the 3rd row immediately. I suppose I'll mess with the timing some more to see if I can control the mice or something. .vbm streamed video Also what's the best .avi recording setting to have decent quality?
Posted: 2/12/2009 4:26 AM
Quote
Mr._Pwnage
Joined: 6/6/2004
Posts: 223
More so than timing, it'd help to actually guide the mice down the lower area quickly! Placing an arrow at K6 is a tremendous waste of time because the mice will wrap all the way across row 6 before they get there, a total of 11 squares. Better is to go G6-D, G7-R, then I7-D and H8-D as usual. H9-R also saves a square compared to having the mice go left to the rocket. Even with the mice coming out relatively slowly like they do there, just guiding them through a faster route takes you down to 14.74.
Posted: 2/12/2009 8:15 PM
Quote
applepieman
Former player
Joined: 10/6/2007
Posts: 86
Can't believe I didn't put the right at H9. Didn't even cross my mine. Well using your exact strategy for the the later half I got 14.73. I'll mess with the timing of the mice here shortly. EDIT: Waiting an additional 90 frames before restarting D3 the cat spawns on the 3rd row immediately again, different mouse layout but not any better (as of now) Can't seem to break the 15 second mark
Posted: 2/7/2019 7:51 PM
Quote
t3h_Icy
Former player
Joined: 8/31/2009
Posts: 236
Hope nobody minds this 5-days-short-of-a-decade bump. C1 - 3.45 Fin - 1.88 I spent some time TASing the stage challenges and the non-TAS world records seem to be very strong. I tied a lot of them, and the others I either couldn't solve to match the record, or couldn't get the necessary RNG (D1-D5), so I've only made videos of ones that are actually faster. The RNG also seems to be very rarely called by anything, so a quickly designed custom stage that both creates and clears our mice/cats is the best method I found. Most of the values seem to give "clumpy" results as well, that is, running the custom stage across 10 frames may yield identical results for each of them. This isn't always true, but the patterns are very slow to change. I definitely recommend trying out this game for any puzzle-savvy TASers, or those who are interested in bot work for the 100 Mice /Final Battle challenges, given how deterministic the game is. It seems to be pretty maxed out as is though.
Posted: 3/7/2023 2:22 AM
Quote
t3h_Icy
Former player
Joined: 8/31/2009
Posts: 236
Here's to another 4 year bump. I improved my TAS of Fin to 1.40 using the same methodology and going for much longer. I would love to see what a well-coded bot could come up with!