Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
I am not getting penalty. I hit walls only in the safe area. It looks and works differently. No flashing stars or time penalty ) EDIT: I've thought I was going to sleep, but then I thought that I've slept for the whole day already, so I spent some more time playing Kururin. http://dehacked.2y.net/microstorage.php/info/3333/kuru1-3.vbm Grassland-1 : unchanged (0'05"78) Grassland-2 : redone. I found out that my joypad bind for rewinding adds empty frame, it did not affect first level, but this one I had to replay. New time is 0'10"66, that's 0'00'"7 faster then known possible from records page. Grassland-3 : 0'12"73! It's almost 2 seconds faster than record on records page, and half a second faster than current "known possible". I hit the wall three times during a short period of passing by healing square, it's a safe zone too, so no damage is taken. Without these three hits I would have to wait for ~second, and I didn't like it ^^ Please take a look and say your opinion ^^ P.S. I use his page: http://www.mzrg.com/games/records/KK/advnorm.shtml as a reference.
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
If anybody is interested '^^ http://dehacked.2y.net/microstorage.php/info/3357/kuru1-9.vbm - first 9 levels done 01: 382-748 (366) - 05"75 (vs 05"76 on records page) 02: 1115-1776 (661) - 10"66 (vs 10"83) 03: 2148-2831 (683) - 12"73 (vs 14"46) 04: 3372-3796 (424) - 05"50 (vs 05"33) - I REALLY don't get how that result was archieved unassisted =\ 05: 4169-4808 (639) - 10"36 (vs 11"86) - FODA had 10"33, I somehow can't get that result =( 06: 5180-6542 (1362) - 22"33 (vs 25"18) 07: 6988-7875 (887) - 14"33 (vs 16"76) 08: 8245-8978 (733) - 11"85 (vs 14"03) 09: 9352-10516 (1164) - 18"97 (vs 21"76)
Joined: 9/20/2006
Posts: 8
Location: Germany
It's nice to see progress on that game. It's one of the games where a TAS will look really cool because of the speed and damagelessness. And when you beat existing world records, it's even better. I wonder if it wouldn't be possible to let a programme calculate when the best moment for a take-off is to get the fastest possible time without hitting the edges? But maybe that would take too much time, even more than just rerecording the levels with begin moving at different propeller positions.
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
Grizzly wrote:
I wonder if it wouldn't be possible to let a programme calculate when the best moment for a take-off is to get the fastest possible time without hitting the edges? But maybe that would take too much time, even more than just rerecording the levels with begin moving at different propeller positions.
In fact it is decided by one or two first THIN passages in most cases...
Editor, Reviewer, Experienced player (979)
Joined: 4/17/2004
Posts: 3109
Location: Sweden
I enjoyed it, but i fear you are missing something on that fourth level if there is an unassisted record that's that much faster.
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
I can't find it :( If I had a video of that record... Spent few hours trying to figure out =( EDIT: I did it, my time is 05"33 now, that's equal to the know record, but I wonder how can that be reproduced unassisted... Not much progress, few levels are redone (8th is ~2 seconds faster), I have exams now, and it's not going well =( http://dehacked.2y.net/microstorage.php/info/3431/kuru1-10.vbm I thought to take two levels from Foda's WIP, as times were better there (almost 1 second on level 07), but frame count comparsion showed that my versions are still shorter (and frame length is my primary goal, ingame timer should not show time worse then unassisted record too). I hate waiting at start, but still if I have to wait, I wait in start zone. Current times: (level: frames - ingame timer (vs unassisted record)
01: 366  - 05"75 (vs 05"76)
02: 661  - 10"66 (vs 10"83)
03: 683  - 12"73 (vs 14"46)
04: 417  - 05"33 (vs 05"33)
05: 639  - 10"36 (vs 11"86)
06: 1362 - 22"33 (vs 25"18)
07: 887  - 14"33 (vs 16"76)
08: 655  - 10"56 (vs 14"03)
09: 1164 - 18"97 (vs 21"76)
10: 965  - 15"65 (vs 19"05)
I should agree with statement that some levels are just made for TASing ^^ EDIT: Now I am at level 12, and I have problem. there is one passage where I need just two more frames to pass, but I can't have them. taking damage there makes total time around 26"75(rough play after that passage) including 3 seconds penalty, while taking other way makes 27 seconds (that's 240 frames slower)... I think I still will not take damage here, but... It's subject for making another run. Two runs for this game are too much though.
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
http://dehacked.2y.net/microstorage.php/info/3446/kuru1-20.vbm
01: 05"75 (vs 05"76)
02: 10"66 (vs 10"83)
03: 12"73 (vs 14"46)

04: 05"33 (vs 05"33)
05: 10"36 (vs 11"86)
06: 22"33 (vs 25"18)

07: 14"33 (vs 16"76)
08: 10"56 (vs 14"03)
09: 18"97 (vs 21"76)

10: 15"65 (vs 19"05)
11: 33"68 (vs 44"56)
12: 27"71(vs 31"60)

13: 08"73 (vs 09"33)
14: 18"38 (vs 23"08)
15: 18"18 (vs 19"63)

16: 07"26 (vs 07"97)
17: 07"85 (vs 11"48)
18: 16"65 (vs 17"43)

19: 04"43 (vs 05"48)
20: 03"10 (vs 03"43)
Some levels may be redone (hex-editing works perfectly). Currently I am having problems with 21st level, I am ~1 second behind unassisted record and is looking for improvements... Levels are getting larger, more complex and interesting, this game is really worth TASing. I don't know about watching though ^^
Joined: 1/23/2006
Posts: 352
Location: Germany
Not collecting the birds prevents the credits from rolling. I think most people here want a TAS to end with some sort of ending sequence instead of being thrown back into the level selection screen.
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
=( Then I'll have to redo some levels :( EDIT: I took a look at it. and found that 2 levels will become much longer when redone with picking up missing bird. Also, Striving for fastest *completion* would include taking damage to save time. I won't complain if movie is not accepted because of unfinishedness...
Editor, Reviewer, Experienced player (979)
Joined: 4/17/2004
Posts: 3109
Location: Sweden
I think minimal completion time is better since there are records to beat. The sonic runs skip the emeralds for similar reasons.
Joined: 1/23/2006
Posts: 352
Location: Germany
Yes but the Sonic run just gets a different ending by skipping these, Kururin throws you back into the level selection. I think many people wouldn't like seeing the run end with either the "I forgot something" screen or just the level selection.
nesrocks
He/Him
Player (246)
Joined: 5/1/2004
Posts: 4096
Location: Rio, Brazil
I think this is the same as the mario64 times competition, where there are star times faster than the TAS (because it used cannon where it wasn't available before, etc). You're playing on different conditions than the record, so it's ok. And I really wouldn't go for fastest frames time, I would do the fastest in-game clock timer you can get. This is more logical and I think it's somewhere on this site's guidelenes too.
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
I think that there should be two runs of this game in fact, but this game is not too popular to have 2 runs on this site. One for fastest ingame times, and second for fastest completion, including taking damage to save time and collecting all birds.
nesrocks
He/Him
Player (246)
Joined: 5/1/2004
Posts: 4096
Location: Rio, Brazil
I don't see the point of finishing a game that has an in-game clock but ignoring it and going for least frames overall...
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
Taking damage involves different tactics and allows some way that are not usable in damageless run...
Joined: 1/23/2006
Posts: 352
Location: Germany
Umihara Kawase has two runs even though noone knows that game, I too think two Kururin runs would be justified.
Joined: 4/11/2006
Posts: 487
Location: North of Russia :[
I am even willing to do both. I really enjoy TASing it ^__^ I have finished level 21 at long last having saved few seconds, now next level is giving some troubles to me... Aww, I love this game :]
Active player (356)
Joined: 1/16/2008
Posts: 358
Location: The Netherlands
digging up an old thread, but thought I'd inform people about this youtube vid: KuruKuru Kururin Adventure mode TAS, aims for fastest real-world time, takes damage to save time (so unfortunately does not do bonus levels) http://www.youtube.com/watch?v=GJmR2APCidM edit: this guy has alot of TASes.... is he from around here? :P
TASes: [URL=http://tasvideos.org/Movies-298up-Obs.html]Mr. Nutz (SNES), Young Merlin 100% (SNES), Animaniacs 100% (SNES)[/URL]
Joined: 5/14/2007
Posts: 525
Location: Pisces-Cetus filament
I purchased this great game recently and have played it a lot. Ryuto, I have watched your awesome TAS of this game and wonder if you have considered TASing the challenge mode. I would love it.
AzumaK wrote: I swear my 1 year old daughter's favorite TASVideo is your R4MI run :3 xxNKxx wrote: ok thanks handsome feos :D Help improving TASVideos!
Skilled player (1543)
Joined: 5/31/2009
Posts: 86
Location: Japan
Actually I'm not going to TAS the challenge mode of Kuru Kuru Kururin. I have considered TASing the mode some times after I had TASed Kuru Kuru Kururin and Kururin Paradise. I thought Kururin Paradise' one would be more entertaining than Kuru Kuru Kururin's because its stages were harder.
mohoc
He/Him
Reviewer, Player (250)
Joined: 5/20/2018
Posts: 34
Location: France
Hey there. I am new here and I would appreciate if there would be anyone to help me investigate some OOB stuff in the GBA game Kuru Kuru Kururin. My main interest is to find an out-of-bounds goal zone that is close to the boundaries and could save time. OOBs are not new in this game but I have found an easy way for humans to go out-of-bounds and thus such a goal zone, if it existed, could be used in an actual speedrun ! I have already found a goal zone in Machine Land 3 but it takes 1m30s to reach it... If you are interested, the following video presents everything I have found so far by randomly messing around :) https://www.youtube.com/watch?v=NbN_AUI2Y6s Just that you know, I have no former experience in TASing. I was just messing around and found the goal in Machine Land 3 and the heart zone in Cave 2 by luck. Besides all the things to look for are probably invisible and off camera view.
Editor, Reviewer, Skilled player (1358)
Joined: 9/12/2016
Posts: 1646
Location: Italy
Hello there! I and mohoc are attempting to make a new TAS of this game, as there are a ton of possible improvements thanks to the OoB discovered by mohoc and the wall collisions abuse discovered by another person on YouTube: Link to video I've managed to reproduce manually the inputs for the first level: User movie #49439186419057236 However, I wonder if we could figure a way to beat these records, if possible... In any case, I'm trying to contact the author for proposing a collaboration. Edit: Oh, almost forgot to mention. I've written some lua scripts for this game: OoB viewer v1.1 is a script that displays a graphical representation of map while you're outside of the screen. It's a bit cpu-hungry, so emulation is gonna slow down a lot on older computers. There is also a VBA-rr version of this script, in case you have troubles using BizHawk. Speed hud v1.2 shows the player position and current speed in the lower left of the screen. Digits before the decimal point are the exact pixel amount, while digits after the decimal point are a representation of subpixels. The script does also show an arrow in the center of the screen, which will point the direction that the player is going to during the last frame, and will change length and colour according to the speed.
my personal page - my YouTube channel - my GitHub - my Discord: thunderaxe31 <Masterjun> if you look at the "NES" in a weird angle, it actually clearly says "GBA"
Editor, Expert player (2330)
Joined: 5/15/2007
Posts: 3933
Location: Germany
That video is nuts. Good luck on the project if you do decide to pick it up. I'm a bit confused by the Youtuber, he says "Cheat bot plays through the game" in the other video. Is that the name of the runner or is the runner making a joke? Is the runner using a script (he mentions a level solver in his other video). His other video (Kururin Paradise) also seems to beat the published run. If you do Kuru Kuru Kururin with the trick, I suggest that you skip the training levels to make it aim for fastest completion.
Player (138)
Joined: 8/27/2004
Posts: 165
Wow, crazy videos by Matt Shepcar. I can't believe I didn't hear about those before. He also has a "100%" run of this game (well, collecting all items with no damage in the main levels, then reaching the credits by beating the first Neo Land level, but he doesn't do the remaining Neo Land levels or formally achieve Master class). I think by "cheat bot" he means some kind of program that finds the fastest way through a level. Then he makes a TAS by putting those input sequences together, and records that. I found this half writeup he did: https://medium.com/message/building-a-cheat-bot-f848f199e76b He mentions writing up the rest but there's no other articles linked to that account. Although it's not really relevant to TASing, I'd be interested to see the fastest possible ingame times for all levels, and for the challenge levels of both games. If someone gets his solver maybe that could be computed as well.
Editor, Reviewer, Skilled player (1358)
Joined: 9/12/2016
Posts: 1646
Location: Italy
Making a new TAS is turning out much harder than I thought. After some testing, I got pretty sure that the zips showcased in the video by Matt Shepcar that I posted above can only be achieved by using a bot. Entering inside a wall (with some exceptions in some levels) requires a precise balance of subpixel position, Helirin angulation, knockback speed, and knockback angulation. I tried hard to reproduce it with trial and error in TAStudio, but I couldn't make it even once. Only the inputs that I copied from the video worked. Now, Matt Shepcar mentioned in his incomplete tutorial about the A* search algorithm, which is probably what he used for making his incomplete TAS, so I think that the best course of action would be to write a good brute-forcing lua script that makes use of this kind of algorithm. I already wrote a couple of brute forcing scripts for other things a year ago, but these were much simpler compared to what it's necessary for TASing this game, and right now I also lack the time and will to work on it... sorry.
my personal page - my YouTube channel - my GitHub - my Discord: thunderaxe31 <Masterjun> if you look at the "NES" in a weird angle, it actually clearly says "GBA"