Submission #4826: TehBerral's NES Wait and See! in 08:50.21

Console Nintendo Entertainment System Emulator BizHawk 1.11.1
Game Version unknown Frame Count 31865
ROM Filename Wait and See (Russia) (Unl).nes Frame Rate 60.0988138974405
Branch Rerecord Count 22619
Unknown Authors TehBerral
Game Wait and See!
Submitted by TehBerral on 9/7/2015 9:11:39 PM

Submission Comments
After some research, this game is based on Ну, погоди! a Russian cartoon and you play as The Hare (who is actually brown and looks nothing like Bugs bunny - in fact they got the colors of the two main characters switched around) who is being pursued by the Wolf for some reason (who we don't see except for some short scenes in between the levels - which I skip).
Game Objectives
  • Takes damage to save time
Movement/tricks used
  • The Hare moves at 4 p/f while running. (Usually, I’ll explain specifics later).
  • Spinning causes him to move at 3 p/f. As such, if you want to kill an enemy and taking damage is not an option, you’ll want to spin for as short a time as possible.
  • Jumping stops the Hare for quite a number of frames while in the air, but by spinning for only a couple frames this will put you into a “running” state in mid-air – allowing you to move at 4 p/f before hitting the ground. Abahbob actually mentioned this in the poll for the current submission. This is used throughout at the point in the air when you’ve already cleared an enemy and don’t need to move upward anymore
  • It should be noted that there’s a 4-frame window where the Hare is in a crouching position after a jump, spinning just before this cancels it out and can also be used to ledge-grab to some extent.
  • The spin doesn’t like to activate on odd-numbered frames for some reason - except when in the air, then it’s hard to spin on even-numbered frames.
  • The spin can actually be chained in mid-air to hover forward. This is done by spinning while holding forward, letting go of spin and hitting it again once the Hare has broken out of the spin. This isn’t a frame perfect maneuver but given the fact that spin durations change based on game’s blood sugar level that morning (or something just as arbitrary) turbo will not work for this this trick. This is good if you want to avoid a short jump to get onto a platform. This is also used in a few places to move through objects and slip between enemy attacks.
  • The game will allow you to double-jump. The game usually will only let you do this straight up into the air, but by doing a spin in between jumps it allows you to not only jump forward easier but you’ll also gain a bit of extra distance with your jumps.
  • It’s a bit difficult to explain the mechanics behind why you’d need to do this, but if you’re falling near enough towards the edge of a platform you can trick the game into thinking you’ve landed on the ground even when you’ve actually overshot the platform. You can think of this like jumping off the wall CWJ-style. There are very few instances in which this helps, but it does come up in this run – mostly in stage 2. Say you have to jump in a way that you have to end up overshooting a platform, but still need to recover your double jump by landing. It would be too slow to stop in the air to actually land on the ground and jump again, but you can delay forward movement enough to cause this false assumption on the game’s part (the assumption being that you’ve hit the ground when you haven’t).
Tl;dr version: wall-kicky thingy makes movement less suboptimal. - Additionally, this can be done on the hippos and alligators in the second stage as well.
Rather than commentate on every little movement I’m going to just be addressing improvements.
Stage 1 (Forest?)
The biggest improvement in this stage comes from spinning for a much shorter time to kill enemies. In fact this stage doesn’t have too many “tricks” per se, or any real strat changes but rather general movement - as mentioned in the bullet points - was improved.
One thing that I figured I’d mention now. The obvious thing with this and the last movie is the horribly choppy camera movement that happens whenever the bunny is in the air.
One thing I did to lessen this is to jump in a way that doesn’t move the camera as much – like in this stage I let the Hare drop into the pits as much as I can before jumping; this keeps him closer to the ground without the camera having to move. This actually allows the Hare to fall forward at his normal running speed and allows him to clip onto ledges quicker. Spin-canceling out of the jumps also helps this. So it actually didn’t end up costing me any time thankfully.
There are some damage boost strats I could employ, but keep in mind, the health doesn’t regenerate after each stage, there are no pickups anywhere and dying at any point would negate any possible time save. Overall in this stage it only costs me 15 frames, which I more than make up for when I actually do decide to take damage in the next stage.
299 frames (roughly 5 seconds) were saved in this stage.
Stage 2 (Grapefruit Falls)
This stage was by far the worst one to TAS. All of the jumping was painful to optimize not only due to the horrid controls, the janky camera, but most importantly, the dodgy enemy placement. The spear mice are the worst and this is where both I and the original publication decide to take damage to save time. The hit boxes for that enemy and its attacks in particular are incredibly hard to dodge, even in a TAS environment. Hover strats were employed a fair bit to either stay above the enemy to attack from there, or to skip the enemy altogether, or to reach the enemy a few frames later when its pattern would make it viable to move through safely.
Another thing that wasn’t necessarily hard but was rather tedious was the ledge clipping. By jumping onto the platforms from a certain height, this allows the Hare to clip onto the platform without the crouch animation. The only reason why it was bad in this stage is due to the overabundance of jumping across large pits with Mario 2 logs rolling down.
82 frames (a little less than 1 and a half seconds) were saved in this stage.
Stage 3 (Ice Cap Zone)
Boring. Next!
65 frames (1 second) were saved in this stage.
Stage 4 (Debt Collector Level)
The Hare runs at 4 p/f normally but in this stage movement seems to be halved in that every other frame the Hare won’t move. In terms of movement there isn’t much I could do to improve the train sections seeing as it moves at a constant rate anyhow, but I was able to improve clipping onto ledges by falling into the pits a fair bit like the first stage. Besides that I was also able to improve the in between sections on foot through cleaner movement.
81 frames (a little less than 1 and a half seconds) were saved in this stage
Stage 5 (Piss-Easy Jet Moto)
The Jet Ski section is extra prone to loss of movement in comparison to the train level and the more accurate emulation doesn’t help that much. As such, a lot of time (relatively a lot) was lost in this stage. I was going to lift XTREMAL’s input and list him as a coauthor, but it didn’t sync well when I tried it due to the change in when movement drops happen. The best I could do was attempt to jump on frames that would not only make the jump over an object, but would also be less susceptible to lag.
To give you an idea of how much more lag this stage is prone to in BizHawk when I entered the Jet Ski section I was 572 frames ahead of the publication, and I ended this movie only 433 frames ahead.
Other Comments
This is the first NES game that I’ve TASed and the first improvement on someone’s work besides my own and despite how horrid the game is (feel free to be as creative as possible when expressing this in the poll thread) I had fun doing it and it gave me a bit of a break from some projects I’m a bit more serious about at the moment. I want to say that I have a new-found respect for XTREMAL93 for simply finishing the original TAS. This isn’t the easiest (read: most fun) game to TAS and doing it without the guide of his TAS would’ve made this near to impossible.

ars4326: Claimed for judging!
ars4326: Hi, TehBerral. Very good optimizations made here (most notably, the movement) on a game that's pretty clunky, to begin with. Compared with XTREMAL93's run, things are now cleaner and a bit more polished, overall (more accurate lag on the last stage, aside). Nice work!
Accepting as an improvement to the published run!
fsvgm777: Processing.

Last Edited by fsvgm777 on 9/11/2015 11:41 AM
Page History Latest diff List Referrers