TASVideos

Tool-assisted game movies
When human skills are just not enough

Submission #5030: Flip's SNES Eek! The Cat in 16:52.75

Console: Super NES
Game name: Eek! The Cat
Game version: unknown
ROM filename:
Branch:
Emulator: (unknown)
Movie length: 16:52.75
FrameCount: 60865
Re-record count: 113704
Author's real name: Phil Hutchinson
Author's nickname: Flip
Submitter: Flip
Submitted at: 2016-03-01 20:25:19
Text last edited at: 2016-03-28 13:15:27
Text last edited by: Flip
Download: Download (26364 bytes)
Status: published
Click to view the actual publication
Submission instructions
Discuss this submission (also rating / voting)
List all submissions by this submitter
List pages on this site that refer to this submission
View submission text history
Back to the submission list
Author's comments and explanations:
This is an improvement of 52.4s (3134 frames) over the previous submission by Warhippy.


(Link to video)

Eek! The Cat is the lesser known version of the PC game Sleepwalker, where our faithful dog needed to escort his sleeping master through various locations and safely back to his bedroom. Of course, Sleepwalker itself was only released for PCs, but the license was passed on and got redesigned into Eek! The Cat for SNES. The levels and locations remain highly similar, but now have our main character Eek escorting various characters from the TV show through all the pitfalls instead.

Game objectives

  • Emulator used: BizHawk 1.11.4
  • Plays on Easiest Difficulty
  • Takes damage to save time

Improvements

Improvements were found on all levels, however many of them get significantly reduced, or even totally erased due to environmental restrictions. All moving platforms/lifts are stuck on predetermined cycles, and so minor optimisations to movements here and there quite often make no difference whatsoever in the long run.

Level Flip Warhippy Difference
0101 1302 1306 -4
0102 2966 2975 -9
0103 3118 3133 -15
0201 3452 3477 -25
0202 3345 3395 -50
0203 4920 5072 -152
0301 1576 1615 -39
0302 5007 5875 -868
0303 4099 4334 -235
0304 2228 2262 -34
0401 2559 2575 -16
0402 3578 3584 -6
0403 5162 5890 -728
0501 3056 3139 -83
0502 4056 4121 -65
0503 3927 4439 -512
0504 5928 6230 -302

Stage by stage comments

Misereek

01-Kicking granny into the swing at a later time, gives granny less distance to fall onto the end goal.

02-Better optimisation, and attacking the porcupine in a single jump.

03-Using time at the beginning for a large run up to granny saves a nice bit, as well as minor optimisations afterwards saves us 69 frames total on the way up. However, due to the periodic nature of barrel throws at the end, getting up here earlier means a barrel will be too far away to clear in a single kick. This means an extra kick would be needed, killing our momentum and giving us a potentially worse final time. This may mean all improvements up to this point have to be totally negated, but not quite. Although we are in fact forced to wait for the barrel to roll closer, the extra frames saved up to this point gives us time to get a big run up. This grants a greater speed than the previous version, allowing us to clear the barrel slightly earlier than before. Since this speed would not be possible without the preceding improvements, it makes all of them still 100% necessary in achieving our goal. Overall, 15 frames ended up being saved; a fine example of how the environment gives a significant reduction, but not a total negation to our improvements.

Eek! vs. The Flying Saucers

01-Better optimisation, but once again reduced by periodic nature of enemies near the end section.

02-Better optimisation on first bit, and a helping kick gets Annabelle up before the laser cycle. A slight delay due to the bridge cycle still results in 27 frames saved up to this point. On the next section, kicking Annabelle gives less distance for her to transverse alone, and better bubble optimisation gets us finished 50 frames faster than before.

03-Dropping off Annabelle at the proper location, and better bubble movement.

Halloweek!

01-Normally getting hit by a falling stalactite will stun you for quite a while, but not if being hit during Eek's delayed fall animation. This now means our temporary invincibility gives us just enough time to get past the lower tier without having to flush out the 2 other stalactites first.

02-Better speed control in the downstairs basin gets us upstairs sooner, but once again gets reduced due to periodic enemy in our way. Afterwards, doing a mad dash straight for the switch will trigger it sooner, as opposed to decelerating to turn around asap. This means the gate will open early enough to allow our partner to pass straight through, so no need for a detour downstairs. Finally, rather than trapping our partner in a small recess, putting him on a path upstairs gets him to where we need him to be, so no huge backtrack needed at the end. This results in the greatest saving of the run, of 868 total frames

03-Just pausing for a moment to take the egg with us saves nearly 200 frames on first section. That of course, creates the usual problem with every periodic fireball blast in the next area being completely out of sync; causing a multitude of potential pathways to test which consumed quite a significant amount of time.

04-A few of the usual minor optimisations, the only significant one being avoiding the moving spark.

Bearz 'n The Hood

01-The first half is as fast as it ever will be, since the elevator granting access to the second half of the level will forever be on the same cycle each time. But at least in this version, the waiting time was used more wisely, utilising a secret corridor for a longer acceleration burst. But once again, the periodic behavior of the final trashcan enemy requires yet another waiting period. As always, the only thing we can do is have a set up a larger run. Another example of a significant reduction, but not total negation of improvements thus far.

02-The first half gets completely dictated by the lift, the second half by the moving barrel. Disappointingly the only room for improvement on this level is at the very end, with proper accelerated on the final shove. Only 6 frames saved.

03-Finally a periodic cycle which can actually be beaten. The way up is of course somewhat meaningless, especially when closed manholes will just reopen themselves later. But proper routing upstairs enabled us to beat the moving barrel 1 cycle earlier, and proper swimming got us a nice time saver too.

It's A Wonderful Nine Lives

01-Previously our partner was kicked onto the washing line from a nearby ledge, which results in 2 bounces needed to successfully land on the left side rooftop. But since this required an additional kick to get up there in the first place, it turns out just starting from the ground floor and using 3 bounces can save us 12 frames getting across. Next, using the downtime to set off future traps allows a more smooth crossing saving another 58 frames; coupled with other small things giving us 83 frames saved total.

02-Even without lugging a huge present around, there's not much time to get to the lift. Even less so with all the annoying kicking and pushing, so don't expect any improvement to the first cycle.

03-Pushing your partner into a fan will give the same height boost each time. Kicking your partner into it will give height proportional to your current speed. This means that with a quick run up, we can accumulate enough speed and thus height to land on the upper tier, skipping the middle one. 512 frames saved this level.

04-Icebergs galore dictate first section, so then the largest saving comes from utilising the lower washing line properly to access the higher one. 288 frames doing that results in an annoying lightning storm now being smack right in the middle of our way, causing sheer agony for routing. The lightning itself is supposed to be a solid obstacle, yet it seems its hit box can be glitched into letting Eek fly straight through it. This too consumed a large portion of the run, as there's quite a few variables to consider. We need to test all possibilities on where to start jumping, when to start jumping, and how high each individual bounce must be. And of course, then there's when we start to accelerate forwards; all of which need to be carefully checked that our use of the line doesn't make either one of us fall straight through it (Only 1 user at a time), but also syncs up a perfect rhythm which glitches our partner into bouncing higher than normal and actually lands them on the platform. Then remember that we also need to find a working path which actually gets us through the storm cloud without getting zapped, and then makes us land BEHIND our partner where we're in a position to start pushing. Needless to say it's tedious work, and the best efforts found all working pathways converge on the same time, of only 1 frame lost to navigate through. 302 frames total saved.


Samsara: Eek, two submissions in two minutes. Judging.

Samsara: Accepting as an improvement to the published run.

feos: Pub.


Similar submissions (by title and categories where applicable):