• Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!

    You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!

Stages and stage hazards

mixa

Banned via Warnings
Joined
Jun 6, 2012
Messages
2,005
Location
Isle of ゆぅ
Hi everyone

This thread is about stages and everything related to them. (VS mode stages)

I will try to make this noob-friendly and good at the same time.

This isn't a wall of text, I made several videos and pictures to illustrate.
It's all hidden in the orange words~~~~

This is very far from done, and even the parts that are done could use peer reviewing.
The reason why I'm coming undone -- oh tricky idioms -- is because, mainly with the Stage Hazards, this could and should be a collective work, add bits here and there and together we'll rise.
And also this has been sitting here forever and I figure I'd just publish before I get BUSY.



Table of contents

Stages: general
Blastzones​
Starting slots​
Average time length​
Auto ledge cancels​
How to counterpick a stage (idea)
Stages Hazards: specific
Peach's Castle​
Congo Jungle
Saffron City (pretty much done)
Hyrule (only data)
DL (only data)
other stages (won't do)


note: when not talking about data, I'm mostly referring to the 4 competitive stages (DL, PC, Congo, Hyrule), since it is in my interest to make this applicable, example-based, useful stuff. so drawing comparisons to Saffron isn't the best idea. when measuring cycles, I'm using approximate values.



Stages: General info and comparisons



Blastzones

A stage blastzone is the area where a character will lose a stock upon touching it.
There is a hacked Yoshi's Island stage that has one so close to the stage that you can see it without any camera tricks: click here.
As far as I know, there isn't something like this blastline display for the original Smash, so I did some manual testing:


Vertical blastzones (ceiling): I tested out the minimum % in which Fox's fresh upsmash can kill Pikachu from the ground: results.

So the stages share pretty much the same ceiling heights. In practical terms, the stage with the lowest ceiling would be the one that it is easiest for your character to set up a vertical finisher on, say, a platform.

Even though I supplied you with numbers, they reflect a clean kill, but we must also take into account a few elements that hinder that clean vertical kill. The most common example is the wall DI in Hyrule (arrows show where such DI happens), Wall DI ex. 1. Another example is the hovering bumper in Peach's Castle.

As for the floor blastzone, it is clear that DL < Congo < Hyrule. So some gimps with Dair (Falcon, Samus) might not work on Congo and definitely won't work on Hyrule. (For the sake of simplicity, I am ignoring the incoming edgeguard). Take this Falcon on Pika combo, this definitely wouldn't work on Congo because it barely works on DL, Pikachu was just about to get out of the hit stun before dying.

On a related note, when Kirby swallow-suicides with the opponent, it is sometimes escapable, and if your character has enough vertical recovery, you can survive, but not on Hyrule, because the distance between the edge and the floor blastzone is so damn huge.


Horizontal blastzones (sides): I tested out the minimum % in which Falcon's non-angled fresh fsmash can kill Yoshi (here's how I positioned myself for this): results.
So, Congo Jungle < Dream Land < Hyrule < ... < Final Destination

This test is neither great nor legit though, because:
Even though I positioned the characters with rolls, it might be off by a % or two.
The position is just arbitrary and unlike being hit by an upsmash from the ground, it's DIable.
Yoshi was hit from the edge and not center stage, and it's much easier to be hit from the edge at the smaller stages than at Hyrule, so in practical terms the sides of a larger stage are even wider.

So just like the ceiling, but worse, how easy it is to kill to the sides can't really be explained in numbers.
There are other factors that will complicate this even more, such as tornados, DI, walls, the ramps at PC which, depending on the relative position of the players and the ramps themselves, will most likely interrupt some finishers, etc.



Starting Slots

Here's a video showing all the starting slots of the official VS mode stages. And here are the pictures used in the video.

Mainly with 1P vs 3P on Dream Land and Peach's, and with adjacent port slots in Hyrule, changing your starting slot to one of those is like gambling, or a test of skill/mind-games, it adds that rush in the beginning of a match.
This is most noticeable when Pikachus counterpick PC (whoever gets the first Jab-grab, gets 20+%), or in Falcon dittos on DL -- really, any character being 3P vs a 1P Falcon isn't in a very good spot.

Ex. 1: Pika (1P) vs Pika (3P) on PC
Ex. 2: Another Pika ditto on PC.
Ex. 3: Falcon (1P) vs Falcon (3P) on Dream Land.

It's also interesting how the videos of Japanese tournament matches always have 2P vs 4P, and sometimes in online matches, where it's 1P vs 2P by default, you'll find they'll artificially make 2P vs 4P: (example).

Such port choices on Dream Land give the players the most neutral position to start a game.

Port Priority: 1P > 2P > 3P > 4P


Average Time Length

Someone calculated the time length of more than 400 console tournament matches: it's all in here.



Auto Ledge Cancels

Video on some of the sweetspots in various stages that can yield a ledge cancel without much skill: here.



How to counterpick a stage

^ Sounds interesting. Any ideas? Not related enough? moving on...






Stage Hazards: Specific info


hazard, n (a)

1. A game at dice in which the chances are complicated by a number of arbitrary rules.
2. Chance, venture; a chance.
3. Risk of loss or harm; peril, jeopardy.
Peach's Castle: platform | bumper | flying ramps
Congo Jungle: platforms | barrel
Hyrule Castle: tornado
Planet Zebes: lava | platform
Yoshi's Story: clouds
Dream Land: wind
Sector Z: arwing
Saffron City: pokémon | platforms
Mushroom Kingdom: piranha plant | *pow* | scale | pipes
active hazards give damage and/or knockback. passive hazards, by themselves, don't.

Stage Analysis


Peach's Castle: platform, bumper, flying ramps



Instead of a description of the movement of each hazard, it's easier to just watch how it moves:



Cycle duration: 1:20

All movement is synchronized.

ex.: If the bumper is on the left side of its trajectory, regardless if it's moving left or right, the platform is either going towards left, or is already there.

But you can't tell the position of the bumper by looking at the platform mid-match, since the bumper skips a platform cycle. Also, the competitive scene has seen little of Peach's Castle and wouldn't be able to make hardcore advanced decisions regarding the position of the ramps and "whether my Fsmash will go clean above/below the ramps".

  • Platform
Not much to say. Would be cool to have a version where you can grab the edges.

  • Bumper
Has set knockback. Deals 1%.

It can be used to combo, or to be in the way of a finisher move and save the character who got hit.

Whenever a character receives an attack, he will receive invincibility frames (sparkle), and if a character is hit while in that invincible state, he will receive the damage of the new attack but retain the knockback of original attack, as it can be seen here (DK's Giant punch knockback is preserved even though Ness is later hit by the bumper). and here is the same situation but without the use of a stage element.

  • Flying Ramps / wedges(?)
Move up and down in a somewhat vertical way. They affect the game in two ways:

- Since they move, at one given time an attack might kill the opponent, at another, the same attack will have its knockback interrupted by the ramps.
- When a character in tumbling animation hits a ramp, it sparkles with intangibility even after the collision (as it can be seen here), making it harder for the opponent to connect his next attack -- sex kicks are advised.

The slanted surface of this stage allows for aerials to be quickly canceled, as this combo shows. (TAS version).





Congo Jungle: platforms, barrel



The platforms and the horizontal movement of the barrel are synchonized.

Stage Movement: Congo Jungle
Barrel-Platforms cycle duration: 0:08
However, in practical terms it's hard to make use of this sync knowledge. By just batting an eye at the platforms mid-match, you can't know where the barrel is and where it is going to:

If the platforms are vertically aligned, the barrel is directly below them, but you where is it going to?
If the platforms are horizontally aligned, the barrel is heading towards center, but where is it?

Unlike Planet Zebes, where the lava cycle is long, in Congo the barrel cycle is too short for the music to be useful. That is not to say that there is no pattern between music and stage movement, but is there a relevant one like in Zebes when the music picks up at 54s, the lava will rise? Probably, otherwise, how else could Europe do this?


  • Platforms
Constant counter-clockwise movement, in sync with the barrel.

  • Barrel: properties
1. Constant horizontal movement.

2. Spins once each 5.7s. A spin lasts 1.2s. It spins, though not alternately, both clockwise and counterclockwise.

3. A character, and only one character, can stay 3s inside the barrel before being thrown out automatically.

4. The following buttons trigger the barrel throw: A, B, R.

5. Miscellanea / Chasing the barrel throw


  • Barrel: usage (the numbers on the previous section are related to the ones in this section)
1.
Imagine this is Donkey Kong Country, you gotta know those barrel trajectories so you don't miss the barrel.​
Roughly 3 seconds is the time you have inside the barrel, here's how much you travel in those ~3s.​


2.
Spins: when (each 5.7s) it'll spin, and to what side (probably 50-50, but not alternately,) they'll spin is, in a match, as good as random.​
When you see the barrel, unless you see it finishing a spin, its past behavior is unknown. Therefore, by going inside it, anything could happen:​
Ex. 1: Barrel starts the spin right after the character entered it. There's no adapting to this, there's no guarantee that waiting, instead of getting out immediatelly, is a good option. On the other hand, there is also the chance that the player decides to wait, and is granted with a clockwise spin, giving him the chance to be thrown diagonally towards the stage.​
Ex. 2: Same as ex. 1. They way I got to: 1 spin/5.7s was using a match where none of the characters actually entered the barrel. Maybe there's a slight chance the the spins are more random than that?​
Ex. wat: Falcon's DI was not suitable/Congo has a weird platform. Falcon entered the barrel, wasn't lucky enough and got jacked. Barrel spin is as good as random.​

Ex. 4: To wait for a spin is risky, but when at high %s and against a competent Falcon (fast and with great aerial movement to chase down the barrel throw), it's a possibility. Never forgetting you have 3s to get out, and that we haven't enough reaction time to get the first diagonal angle of a spin. So in that case, the Pika would need a counterclockwise spin to try the diagonal throw.​
3.
The barrel hitbox isn't obvious. It's tricky if you're coming from below.​
Ex. 1: Fox going right through it.​
Barrel hog: stop the opponent from entering the barrel by doing it yourself. Mew2King messed that one up, make sure you don't.​
4.
The reason why I put the buttons that trigger the barrel throw is because the camera isn't perfect.​
Ex. 1: a barrel might come out of nowhere, and you're pressing B to recover, and now you're dead.​
Ex. 2: another case of unintended barrel entrance.​
5.
It's true that against competent players the barrel most likely won't do anything (in fact, they might style on you):​
Ex. 1: Pikachu manages to kill a 48% Fox out of the barrel.​
Counter-ex. 2: Fox was certainly dead from the gimp if there weren't a barrel to catch him.​
We must also keep in mind that slow characters, mainly the ones with poor vertical speed, have a harder time killing the opponent who has been barrel thrown.​


Chasing the Barrel throw:
Characters thrown vertically from the barrel will get out of hit stun a little bit after they reach the apex of their trajectory.​
This apex is different for each character. But as you'll see, all of them are high, so the rule of thumb is to stay on the top platforms while you wait for the opponent to be thrown out of the barrel.​
I've taken pictures of each apex. This is not TAS, but I've tried my best to get the first frame where the character will turn upsidedown.​

From high to low:​
1. Jigglypuff (142)​
2. Samus (137)​
3. Luigi (136)​
4. Mario (132)​
5. Kirby (126)​
6. Pikachu (118)​
7. Ness (116)​
8. DK (113)​
9. Yoshi (112)​
10. Fox (105)​
11. Link (102)​
12. Falcon (100)​
This stage is dark and you can barely see the characters, so if you wanna emulate a slideshow with the original size of the images, you can download all the pictures by clicking download album on the imgur page.​
-- edit: As pointed out by Madao, Jiggly is higher, but the pictures still give you what you need to know in a match, which is how high approximately the characters go.​
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
I think it would be a good idea to know the actual X and Y coordinates for blast zones.
So far I tested Dream Land.
After you know the the blast zone coordinates, you can just use the physics formulas to determine pretty much any scenario without having to do nearly as much manual testing.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
Well I ended up finding the variables for the blast zones.

BlastZone Data

The numbers posted are the limit, so anything past the limit = death.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
I fixed the link.

What emulator do you use? I have a giant cheat engine table of addresses for Nemu and Mupen rerecording, if you want it.

Maybe soon I can take a look at hazard data, to get damage and knockback values.
 

mixa

Banned via Warnings
Joined
Jun 6, 2012
Messages
2,005
Location
Isle of ゆぅ
I use Project64, but I could try and use Mupen or Nemu. Then again, I know nothing about cheats. I can only use them because of the nice cheat file I got.

And I still get open that link. I get 'error 403'.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
Lol sorry, I'm not used to using dropbox. You think I should use google docs instead?

Do you know how to use cheat engine? If so, then it's definitely worth it to use my table. Also if you decide to use Nemu or Mupen, let me know which one, so I can upload that specific table. The only issue with my table is that since its pretty long, it's kinda messy. You will be able to see things like Blast Zones, Player coordinates, Frame counter, Character data, etc.
 

mixa

Banned via Warnings
Joined
Jun 6, 2012
Messages
2,005
Location
Isle of ゆぅ
You think I should use google docs instead?
Yeah, or some other simple one like 2shared.

I'll tell you right now, I have no idea how to use a cheat engine. But if it's not that hard and you'll only need to upload the thing, I'll give it a try.

Any specific Mupen version I should use?
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
Cheat Engine takes a few mins to understand the basics. The harder part is actually finding data on your own, if you plan on doing things by yourself later on. All you will have to do is install Cheat Engine. After that, you run the CT file that I will upload shortly. Then once Cheat Engine is running, you click on the icon with the magnifying glass and computer and then select whatever exe you plan on running (mupen in this case). After you select the exe, a message will pop up asking you to keep the address list. Hit yes and then you're done.

Here is the mupen you should use http://code.google.com/p/mupen64-rr/

Should I leave in things like Z cancel timer? Idk if you guys would want that stuff posted in public. It was interesting seeing the cpu always Z cancel after i froze the timer. I also found the hitbox glitch (glitchy and not perfect), so you can do it on any level now if you want. i have 2 issues with it though, 1 is that i can only get 1 character to show hitbox, and 2, it's buggy.

I guess I'll try to change a few names in my table to make it more understandable. When you get the table and get it working, let me know if there's anything you don't understand. I'll explain the hitbox glitch once you get the table working as well.

I took a look at Planet Zebes and found the data for the Acid.
Address------Stat---------------Value
8022D6C0___Damage____________16
8022D6C4___Angle______________80
8022D6C8___Knockback Scaling__130
8022D6D0___Base Knockback_____30

If you wanted to edit these values via gameshark or w/e, I think the variables are 4 bytes and the value is located on the 4th byte. So you will see 0x10 at 8022D6C3.
 

mixa

Banned via Warnings
Joined
Jun 6, 2012
Messages
2,005
Location
Isle of ゆぅ
I'm feeling lost already... I have the Mupen installed, where do I get this cheat engine? I thought you had to upload something first? lol I'm so bad at this.


That lava data is really cool. I've always wanted to know the angle. :3
Though I thought it'd come out as higher than 90 (100, in this case) since the lava sends you the opposite way you're facing.


And I've just looked at the Blastzone data and I've noticed that even though we now have the values, they aren't relative to the stage position in the air. For instance,

Congo Jungle has Y = 8000
Dream Land has Y = 8300
Hyrule Castle has Y = 9000

But from my Fox on Pika test, we know that Congo has the same height as Dream Land at their lowest levels. And that at center stage Hyrule, the ceiling is lower than DL and Congo.

I feared this would happen, that's why I thought of doing comparative tests.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
Lol i forgot to work on my table. I was adding stuff earlier to my Nemu table, so I need to update my Mupen one.

Here's Cheat Engine.
http://members.upc.nl/cheatengine4/d1/CheatEngine63.exe

I'll upload my table soon. I might do one last set of tests, to see if I find anything new before uploading.
Also, what do you mean by "they aren't relative to the stage position in the air." ?
Well, when you get the table, you will be able to see your X and Y Coordinates, as well as a bunch of other stuff.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
I uploaded the table. You can choose either Nemu or Mupen.

If you have any questions, let me know. I hope you guys find this to be useful. Feel free to organize it however you want it, since sometimes its more convenient to switch the ordering around. Also, the character data like weight and stuff is for Player 1.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
Hyrule Tornado
Address------Stat-------------Value
8022D6C0___Damage___________14
8022D6C4___Angle_____________90
8022D6C8___Knockback Scaling__60
8022D6D0___Base Knockback___115

DK Barrel
Knockback Scaling = 0.0
Base Knockback = 180

I started looking at the random number generation for the amount of time tornado lasts and how long it takes to reappear. So far I know the time the tornado takes (by number of frames) to reappear is 1600 + random number.
I guess I could add these stage data addresses and the seed variable to the cheat engine table. I'll try to put in the staleness multipliers in the table too.

I made myself a calculator for knockback and stuff. It's almost finished. I just need to implement a few things like spiking, and fixed knockback formula. I also need to see if I can improve the floating point accuracy (very minor). Idk if I'll be able to add in DI as well though. For the most part, you can pretty much determine what can KO and what can't. Some things may be challenging though, like bouncing off of objects. Next thing I should prolly do is write down the coordinates of common flat areas in each stage. Like for example, in Dream Land, the Y Coordinate on the ground = 0. The X Coordinate of far left edge on ground = -2318. Middle = 0. Far Right edge of ground = 2318.

I wish I knew how to make spreadsheets, instead of making an exe file lol. Also being able to directly read moveset and character values would be nice.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
I decided to screencap Dreamland and write the coordinates of different parts of the stage. So now you can plug in the knockback and velocity formula to determine what can or can't KO.
 

mixa

Banned via Warnings
Joined
Jun 6, 2012
Messages
2,005
Location
Isle of ゆぅ
I'm pretty sure the right platform is higher than the left one. Not that it'll ever matter.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
I'm pretty sure the right platform is higher than the left one. Not that it'll ever matter.
Also for barrel throw height, Jigglypuff goes higher than Samus. Would you like me to take a screenshot of 1 or 2 more stages and write in the X and Y coordinates of specific areas?
 

mixa

Banned via Warnings
Joined
Jun 6, 2012
Messages
2,005
Location
Isle of ゆぅ
Jigglypuff goes higher? Nice job, I should've suspected Samus getting first.



Using your table, and the Player 1 Y Position thing, it returned:

left platform: Y = 904
right platform: Y = 907

Which explains this:


By the way, what do these numbers mean, 904 units of what?

Would you like me to take a screenshot of 1 or 2 more stages and write in the X and Y coordinates of specific areas?
Oh, don't worry about that. Now that I know of the player 1 X/Y position thing I could do that myself.

Including finding out every surface where: run -> shine cancel doesn't work because of the slight inclination.
 

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
Jigglypuff goes higher? Nice job, I should've suspected Samus getting first.

Using your table, and the Player 1 Y Position thing, it returned:

left platform: Y = 904
right platform: Y = 907

Which explains this:

By the way, what do these numbers mean, 904 units of what?

Oh, don't worry about that. Now that I know of the player 1 X/Y position thing I could do that myself.

Including finding out every surface where: run -> shine cancel doesn't work because of the slight inclination.
Good find! I have a tendency to make silly mistakes like that one.
I must have thought 904 because I was looking at the other player's Y coordinates LOL. I'll edit the pic and replace the 904 with 907.
Coordinates are in floating point. I think most 3d games use floating point to determine stuff like size and location. 2d games usually use integer for the data type of coordinates. It's usually one of the first things I look for in any game that I explore.

It's funny because earlier today, I was recording the data for the polygon characters and realised I was writing down player 1's data instead of player 2 -__-.
You guys notice I added new stuff like Cpu level and handicap level? I'm liking the idea of changing cpu level and handicap, anytime I want. I wonder why they made cpu only lvl 3 in training more. I'm starting to find more interesting data. I've also been tweaking my calculator adding more stuff to it and fixing errors. I may post it soon, if you guys are interested in doing knockback calculations.

Here's an approximate for the max Y Distance for Barrel going straight up.
1. JigglyPuff = 6201.810547
2. Samus = 5975.609863
3. Luigi = 5931.310059
4. Mario = 5753.609375
5. Kirby = 5514.909180
6. Pikachu = 5180.008789
7. Ness = 5066.308594
8. DK = 4955.408203
9. Yoshi = 4888.807129
10. Fox = 4595.507813
11. Link = 4470.407227
12. Falcon = 4373.406738
When I manually measured Jigglypuff, I also got 6201.810547, so it seems to be spot on.
 

Studstill

Smash Ace
Joined
Mar 16, 2014
Messages
807
Hi everyone

This thread is about stages and everything related to them. (VS mode stages)

I will try to make this noob-friendly and good at the same time.

This isn't a wall of text, I made several videos and pictures to illustrate.
It's all hidden in the orange words~~~~

This is very far from done, and even the parts that are done could use peer reviewing.
The reason why I'm coming undone -- oh tricky idioms -- is because, mainly with the Stage Hazards, this could and should be a collective work, add bits here and there and together we'll rise.
And also this has been sitting here forever and I figure I'd just publish before I get BUSY.



Table of contents

Stages: general

Blastzones
Starting slots
Average time length
Auto ledge cancels
How to counterpick a stage (idea)

Stages Hazards: specific

Peach's Castle
Congo Jungle
Saffron City (pretty much done)
Hyrule (only data)
DL (only data)
other stages (won't do)


note: when not talking about data, I'm mostly referring to the 4 competitive stages (DL, PC, Congo, Hyrule), since it is in my interest to make this applicable, example-based, useful stuff. so drawing comparisons to Saffron isn't the best idea. when measuring cycles, I'm using approximate values.



Stages: General info and comparisons



Blastzones

A stage blastzone is the area where a character will lose a stock upon touching it.
There is a hacked Yoshi's Island stage that has one so close to the stage that you can see it without any camera tricks: click here.
As far as I know, there isn't something like this blastline display for the original Smash, so I did some manual testing:


Vertical blastzones (ceiling): I tested out the minimum % in which Fox's fresh upsmash can kill Pikachu from the ground: results.

So the stages share pretty much the same ceiling heights. In practical terms, the stage with the lowest ceiling would be the one that it is easiest for your character to set up a vertical finisher on, say, a platform.

Even though I supplied you with numbers, they reflect a clean kill, but we must also take into account a few elements that hinder that clean vertical kill. The most common example is the wall DI in Hyrule (arrows show where such DI happens), Wall DI ex. 1. Another example is the hovering bumper in Peach's Castle.

As for the floor blastzone, it is clear that DL < Congo < Hyrule. So some gimps with Dair (Falcon, Samus) might not work on Congo and definitely won't work on Hyrule. (For the sake of simplicity, I am ignoring the incoming edgeguard). Take this Falcon on Pika combo, this definitely wouldn't work on Congo because it barely works on DL, Pikachu was just about to get out of the hit stun before dying.

On a related note, when Kirby swallow-suicides with the opponent, it is sometimes escapable, and if your character has enough vertical recovery, you can survive, but not on Hyrule, because the distance between the edge and the floor blastzone is so damn huge.


Horizontal blastzones (sides): I tested out the minimum % in which Falcon's non-angled fresh fsmash can kill Yoshi (here's how I positioned myself for this): results.
So, Congo Jungle < Dream Land < Hyrule < ... < Final Destination

This test is neither great nor legit though, because:
Even though I positioned the characters with rolls, it might be off by a % or two.
The position is just arbitrary and unlike being hit by an upsmash from the ground, it's DIable.
Yoshi was hit from the edge and not center stage, and it's much easier to be hit from the edge at the smaller stages than at Hyrule, so in practical terms the sides of a larger stage are even wider.

So just like the ceiling, but worse, how easy it is to kill to the sides can't really be explained in numbers.
There are other factors that will complicate this even more, such as tornados, DI, walls, the ramps at PC which, depending on the relative position of the players and the ramps themselves, will most likely interrupt some finishers, etc.



Starting Slots

Here's a video showing all the starting slots of the official VS mode stages. And here are the pictures used in the video.

Mainly with 1P vs 3P on Dream Land and Peach's, and with adjacent port slots in Hyrule, changing your starting slot to one of those is like gambling, or a test of skill/mind-games, it adds that rush in the beginning of a match.
This is most noticeable when Pikachus counterpick PC (whoever gets the first Jab-grab, gets 20+%), or in Falcon dittos on DL -- really, any character being 3P vs a 1P Falcon isn't in a very good spot.

Ex. 1: Pika (1P) vs Pika (3P) on PC
Ex. 2: Another Pika ditto on PC.
Ex. 3: Falcon (1P) vs Falcon (3P) on Dream Land.

It's also interesting how the videos of Japanese tournament matches always have 2P vs 4P, and sometimes in online matches, where it's 1P vs 2P by default, you'll find they'll artificially make 2P vs 4P: (example).

Such port choices on Dream Land give the players the most neutral position to start a game.

Port Priority: 1P > 2P > 3P > 4P


Average Time Length

Someone calculated the time length of more than 400 console tournament matches: it's all in here.



Auto Ledge Cancels

Video on some of the sweetspots in various stages that can yield a ledge cancel without much skill: here.



How to counterpick a stage

^ Sounds interesting. Any ideas? Not related enough? moving on...




▬▬▬▬

Stage Hazards: Specific info


hazard, n (a)

1. A game at dice in which the chances are complicated by a number of arbitrary rules.
2. Chance, venture; a chance.
3. Risk of loss or harm; peril, jeopardy.


Stage Analysis


Peach's Castle: platform, bumper, flying ramps



Instead of a description of the movement of each hazard, it's easier to just watch how it moves:



Cycle duration: 1:20

All movement is synchronized.

ex.: If the bumper is on the left side of its trajectory, regardless if it's moving left or right, the platform is either going towards left, or is already there.

But you can't tell the position of the bumper by looking at the platform mid-match, since the bumper skips a platform cycle. Also, the competitive scene has seen little of Peach's Castle and wouldn't be able to make hardcore advanced decisions regarding the position of the ramps and "whether my Fsmash will go clean above/below the ramps".

  • Platform
Not much to say. Would be cool to have a version where you can grab the edges.

  • Bumper
Has set knockback. Deals 1%.

It can be used to combo, or to be in the way of a finisher move and save the character who got hit.

Whenever a character receives an attack, he will receive invincibility frames (sparkle), and if a character is hit while in that invincible state, he will receive the damage of the new attack but retain the knockback of original attack, as it can be seen here (DK's Giant punch knockback is preserved even though Ness is later hit by the bumper). and here is the same situation but without the use of a stage element.

  • Flying Ramps / wedges(?)
Move up and down in a somewhat vertical way. They affect the game in two ways:

- Since they move, at one given time an attack might kill the opponent, at another, the same attack will have its knockback interrupted by the ramps.
- When a character in tumbling animation hits a ramp, it sparkles with intangibility even after the collision (as it can be seen here), making it harder for the opponent to connect his next attack -- sex kicks are advised.

The slanted surface of this stage allows for aerials to be quickly canceled, as this combo shows. (TAS version).





Congo Jungle: platforms, barrel



The platforms and the horizontal movement of the barrel are synchonized.

Stage Movement: Congo Jungle

Barrel-Platforms cycle duration: 0:08

However, in practical terms it's hard to make use of this sync knowledge. By just batting an eye at the platforms mid-match, you can't know where the barrel is and where it is going to:

If the platforms are vertically aligned, the barrel is directly below them, but you where is it going to?
If the platforms are horizontally aligned, the barrel is heading towards center, but where is it?

Unlike Planet Zebes, where the lava cycle is long, in Congo the barrel cycle is too short for the music to be useful. That is not to say that there is no pattern between music and stage movement, but is there a relevant one like in Zebes when the music picks up at 54s, the lava will rise? Probably, otherwise, how else could Europe do this?


  • Platforms
Constant counter-clockwise movement, in sync with the barrel.

  • Barrel: properties
1. Constant horizontal movement.

2. Spins once each 5.7s. A spin lasts 1.2s. It spins, though not alternately, both clockwise and counterclockwise.

3. A character, and only one character, can stay 3s inside the barrel before being thrown out automatically.

4. The following buttons trigger the barrel throw: A, B, R.

5. Miscellanea / Chasing the barrel throw


  • Barrel: usage (the numbers on the previous section are related to the ones in this section)
1.
Imagine this is Donkey Kong Country, you gotta know those barrel trajectories so you don't miss the barrel.

Roughly 3 seconds is the time you have inside the barrel, here's how much you travel in those ~3s.​


2.
Spins: when (each 5.7s) it'll spin, and to what side (probably 50-50, but not alternately,) they'll spin is, in a match, as good as random.

When you see the barrel, unless you see it finishing a spin, its past behavior is unknown. Therefore, by going inside it, anything could happen:

Ex. 1: Barrel starts the spin right after the character entered it. There's no adapting to this, there's no guarantee that waiting, instead of getting out immediatelly, is a good option. On the other hand, there is also the chance that the player decides to wait, and is granted with a clockwise spin, giving him the chance to be thrown diagonally towards the stage.

Ex. 2: Same as ex. 1. They way I got to: 1 spin/5.7s was using a match where none of the characters actually entered the barrel. Maybe there's a slight chance the the spins are more random than that?

Ex. wat: Falcon's DI was not suitable/Congo has a weird platform. Falcon entered the barrel, wasn't lucky enough and got jacked. Barrel spin is as good as random.​

Ex. 4: To wait for a spin is risky, but when at high %s and against a competent Falcon (fast and with great aerial movement to chase down the barrel throw), it's a possibility. Never forgetting you have 3s to get out, and that we haven't enough reaction time to get the first diagonal angle of a spin. So in that case, the Pika would need a counterclockwise spin to try the diagonal throw.

3.
The barrel hitbox isn't obvious. It's tricky if you're coming from below.

Ex. 1: Fox going right through it.

Barrel hog: stop the opponent from entering the barrel by doing it yourself. Mew2King messed that one up, make sure you don't.

4.
The reason why I put the buttons that trigger the barrel throw is because the camera isn't perfect.

Ex. 1: a barrel might come out of nowhere, and you're pressing B to recover, and now you're dead.

Ex. 2: another case of unintended barrel entrance.

5.
It's true that against competent players the barrel most likely won't do anything (in fact, they might style on you):

Ex. 1: Pikachu manages to kill a 48% Fox out of the barrel.

Counter-ex. 2: Fox was certainly dead from the gimp if there weren't a barrel to catch him.

We must also keep in mind that slow characters, mainly the ones with poor vertical speed, have a harder time killing the opponent who has been barrel thrown.​


Chasing the Barrel throw:

Characters thrown vertically from the barrel will get out of hit stun a little bit after they reach the apex of their trajectory.

This apex is different for each character. But as you'll see, all of them are high, so the rule of thumb is to stay on the top platforms while you wait for the opponent to be thrown out of the barrel.

I've taken pictures of each apex. This is not TAS, but I've tried my best to get the first frame where the character will turn upsidedown.

Barrel throw heights

From high to low:

1. Jigglypuff (142)
2. Samus (137)
3. Luigi (136)
4. Mario (132)
5. Kirby (126)
6. Pikachu (118)
7. Ness (116)
8. DK (113)
9. Yoshi (112)
10. Fox (105)
11. Link (102)
12. Falcon (100)

This stage is dark and you can barely see the characters, so if you wanna emulate a slideshow with the original size of the images, you can download all the pictures by clicking download album on the imgur page.
-- edit: As pointed out by Madao, Jiggly is higher, but the pictures still give you what you need to know in a match, which is how high approximately the characters go.​
Hey Man can I use some of this on my site?
 

rpotts

Smash Lord
Joined
Apr 6, 2009
Messages
1,121
Location
Lawrence, KS
Most worthwhile bump, great read. Thanks for your never-ending hard work, mixa.

You too, Madao.

<3
 

Studstill

Smash Ace
Joined
Mar 16, 2014
Messages
807
Yeah exactly, cheat engine advice was what I meant. So: Fox vs All, on DL and Hyrule? Awesome, I dig it. Any other chars? Those are the only two boards I care/anyone does, I think. But I`d rather go the cheat engine route, lock down the stage sizes/plat heights, adn just plug all the numbers in. I`ve seen a few cheat engine tables, but would probably really benefit from a brief chat about it, pro-tips and what not. PM me here or chris.studstill on Skype, if that`s something you`d be willing to do. Thanks!
 

B Link

Smash Lord
Joined
Sep 26, 2007
Messages
1,579
Location
Toronto, Ontario
Hey Madao does your Mupen CT or PJ64kVE CT still work for you? Only your Nemu one seems to be working for me right now. Don't know why.
 
Last edited:

Madao

Moderator
Moderator
Joined
Jun 27, 2013
Messages
873
Hey Madao does your Mupen CT or PJ64kVE CT still work for you? Only your Nemu one seems to be working for me right now. Don't know why.
If you mean that it's not working with the emulator, it might mean that you're using the wrong emulator. Pretty sure the Mupen table was for mupen64-rr, and the KVE was for the exe I modded in order to make the RAM address static.

If that's not what you meant, let me know.
 
Top Bottom