This is my first time actually trying to contribute meaningfully to the community at large. A bit of background: I've been playing melee since 2001, and I'd played SSB for some time before that. I learned about smashboards in 04/05, and lurked heavily until I joined. Although I wouldn't call myself good at playing, I do better when it comes to knowledge of the general game mechanics/theory, as well as explaining things that *should* be common sense.
Anywho, I'm being boring
First Topic: Safety Against Someone Shielding
So pretty much brawl is a game where shield stun has been lowered, hitstun has been lowed, and the time it takes to drop out of your shield has been lowered. We've all known this since the game came out. However, as we've experimented with the game since, people have found attacks that autocancel, attacks with little lag on landing, and attacks that end before you touch the ground.
I know that most people think that all of these are equally "safe," meaning you can use them and not be punished. However, I'd like to challenge that misconception, and try and widen the scope of player's thinking.
First to discuss, why auto-canceled attacks differ from attacks that end before you land.
This may seems to be the same thing, meaning that they are equally safe. However, there is a key flaw in this assumption, and this is: taking into account the hitstun on your opponent. With a move that autocancels, you can hit the shield right as you are going to land, meaning their stun lasts until after you land, giving you a few frames of safety.
With a move that you have to time to end with no lag, this isn't the case. To hit them, AND have no landing lag, you generally have to hit them several frames BEFORE you land, meaning in some cases they have already recovered from the stun by the time you land.
Now, to look at other factors of being safe
It's a commonly held believe that the range of a move increases it's safety...but if that were the case, Ike would be top tier, so obviously there are other factors. Again, one such factor mentioned before was hitstun--if the move is over before the enemy can get out of stun, it's safe. But then why exactly is the range important? The answer is the distance between you and the opponent when the move finishes. In addition to them having to get out of stun, they then have to reach YOU. This is why some attacks are used to space, because they are quick enough, and have a long enough reach to fulfill these requirements.
Now I bolded the word quick for a reason here is why:
what makes a move quick?
A move being quick consists of a few elements: time to come out, the time the hitbox is out, and the lag after the move is used. In terms of what makes a move safe, the most important part is the cool down lag, not the start-up lag.
Conclusion
There is a bigger difference between autocanceling and ending before one lands, which can make a difference between an approach being successful and failing. Also, more factors than just range are involved in properly "spacing" an attack. For example, Ike could space his f-air perfectly, but still be punished, while he can be more sloppy with his n-air, and still attack before you can counter him.
Now for the tricky part
as I've begun to explain, there are sever factors that determine whether a move is safe. Now I'll try and piece it together, both in words, and using math. The reason for the latter will be explained later on.
So the basic idea behind a move being safe is that you can act before they can. This is on the basic level boiled down to two factors--the distance between you, and the time it takes to get to bridge the gap. However, as previously explained there are several things that cause these factors. The distance is affected not only by the range of your move, but also by how much your move pushes the opponent back (moves like charizard's rock smash, for example), and how far back you can DI while/after doing it, before you touch the ground.
The factors that affect the time are also plentiful:shield stun, time to release the shield, the time is takes to initiate a dash, walk, jump, and the speed at which the character can dash, walk, jump. In addition, the speed which they can then grab, or attack, at which range is also a factor--if they have a quick, long jab, they don't need to come all the way up to you to hit you, for example.
Tricky Part 2--The headache of calculations
Now this is what I need help with. Mathematically minded people (M2K?) try and follow, this should be simple.
As said previously, all the aspects boil down to two factors: distance and time. These are, in a video game represented by pixels (p) and frames (d). To be safe (S), the result of the f(S)> 0, where the f(S)= tR-tL,
where
tR=time to recover/punish, and
tL is the time of the lag.
Thus, if it took the opponent 30 frames to punish, and you were lagging for 20, your move was safe.
However, as previously mentioned, distance plays a role. But for the result to remain with one unit, somedthing needs to cancel out the distance unit. Luckily, anyone whose taken physics knows:
v=d/t, which can be rearranged to t=d/v.
A quick example:
a man is running 10 meters, at a pace of 5m/s. How long does it take?
v=5m/s
d=10m
t=?s
so t=10m/(5m/s). The m's cancell out, leaving you with 2 seconds=t.
So know we can make our formula:
f(S)=v/d-tL, where
v=the characters running/dashing/walking/jumping speed in d/f, and
d=pixels between the characters.
Now we have the basic idea down, but alas, more completity is in store. Namely, what makes up d? As mentioned earlier, the distance is comprised of range of the attack, and how far the enemy is knocked back, as well as how far you can drift back. So mathmatically, we get to assign variables to all of those. so now we have
d=(R+Kn+Di)p, which when plugged in makes the equasion:
f(S)=v/(R+Kn+Di)-tL
If you are with me so far, the next step should be a piece of cake. In addition to the lag that the move's user has, we need to account for the lag that the opponent is in. So this lag consists of (in a shielding opponent) Shieldstun, time to drop the shield, time to start the motion, and time to start your countermove.
so the basic formula should look like this:
(S)=[(v/d)+tOL]-tL
where
tOL is the time of the opponents lag, which consists of:
SS, shield stun
SD, frames to drop shield
DSt,the time it takes to initiate run, jump, walk, dash
CA, the time to use a counter attack (any move)
so,
tOL=SS+SD+DSt+CA, making the complex equasion:
f(S)={[v/(R+Kn+Di)]+[SS+SD+DSt+CA]}-tL
starting to hurt your mind yet? It gets a bit worse now. Because the counter attack (variable CA) has a range attached to it, the distance the unit has to travel to punish the opponent is lowered. So now, d goes from
d=R+Kn+Di to d=R+Kn+Di-Rca, which makes the complex equasion:
f(S)={[v/(R+Kn+Di-Rca)]+[SS+SD+DSt+CA]}-tL.
Okay, so what was the point of this?
The point of the first section was just to make people think, which is important when playing brawl/melee any fighting game, as the principle is the same. I hope challenges people to think about why they are doing what they are doing, and whether it's more/less effective than they first assumed.
Using this formula, plus frame date, we can come up with a list of safe/unsafe attacks for each character, against a shielding opponent. Actually, with regards to auto-cancelled attacks, this equasion becomes much simpler, as d=Range of your attack-range of there counter attack option.
And, well wouldn't that be useful information? more the most part it would translate into semi-obvious things, but in my opinion you can never have to much information AVAILABLE, even if you don't need/want to look at it.
My request(s)
Unfortunately, I do not have brawl, and I have no means of counting frames. I know most of you have brawl, and no clue whether there is some way to get frame data yet. But if there is, I'm thinking of using this thread as a means of organizing frame data, and using it to break down the game.
I'd like the community's help with this, but if it's impossible, at least the math was fun (in a I haven't taken a math course in 3 years now) sort of way.
Also, comments, advice, more general info/feedback[1] would be appreciated, as in addition to all this mathy stuff this is at heart a Common Sense Guide to Brawl, and as I play more/come up with more things I think need explaining I'll write up about it ^_^
[1] If you think I've missed something, am wrong about something, glossed over something, didn't explain something, thing I'm simply wasting time, then let me know, please, so I can either give up on this or continue/correct things
[Future potential topics I've mused over]
*What if they get hit by an attack?
*the importance of multiple hitboxes
*How perfect shielding works
*What makes good spam+ a look at all projectiles.
*"Comboing" in Brawl
Appendix of other useful threads:
Tyr_03's list of autocanceled aerials
Magus' fram data for Luigi
Anywho, I'm being boring
First Topic: Safety Against Someone Shielding
So pretty much brawl is a game where shield stun has been lowered, hitstun has been lowed, and the time it takes to drop out of your shield has been lowered. We've all known this since the game came out. However, as we've experimented with the game since, people have found attacks that autocancel, attacks with little lag on landing, and attacks that end before you touch the ground.
I know that most people think that all of these are equally "safe," meaning you can use them and not be punished. However, I'd like to challenge that misconception, and try and widen the scope of player's thinking.
First to discuss, why auto-canceled attacks differ from attacks that end before you land.
This may seems to be the same thing, meaning that they are equally safe. However, there is a key flaw in this assumption, and this is: taking into account the hitstun on your opponent. With a move that autocancels, you can hit the shield right as you are going to land, meaning their stun lasts until after you land, giving you a few frames of safety.
With a move that you have to time to end with no lag, this isn't the case. To hit them, AND have no landing lag, you generally have to hit them several frames BEFORE you land, meaning in some cases they have already recovered from the stun by the time you land.
Now, to look at other factors of being safe
It's a commonly held believe that the range of a move increases it's safety...but if that were the case, Ike would be top tier, so obviously there are other factors. Again, one such factor mentioned before was hitstun--if the move is over before the enemy can get out of stun, it's safe. But then why exactly is the range important? The answer is the distance between you and the opponent when the move finishes. In addition to them having to get out of stun, they then have to reach YOU. This is why some attacks are used to space, because they are quick enough, and have a long enough reach to fulfill these requirements.
Now I bolded the word quick for a reason here is why:
what makes a move quick?
A move being quick consists of a few elements: time to come out, the time the hitbox is out, and the lag after the move is used. In terms of what makes a move safe, the most important part is the cool down lag, not the start-up lag.
Conclusion
There is a bigger difference between autocanceling and ending before one lands, which can make a difference between an approach being successful and failing. Also, more factors than just range are involved in properly "spacing" an attack. For example, Ike could space his f-air perfectly, but still be punished, while he can be more sloppy with his n-air, and still attack before you can counter him.
Now for the tricky part
as I've begun to explain, there are sever factors that determine whether a move is safe. Now I'll try and piece it together, both in words, and using math. The reason for the latter will be explained later on.
So the basic idea behind a move being safe is that you can act before they can. This is on the basic level boiled down to two factors--the distance between you, and the time it takes to get to bridge the gap. However, as previously explained there are several things that cause these factors. The distance is affected not only by the range of your move, but also by how much your move pushes the opponent back (moves like charizard's rock smash, for example), and how far back you can DI while/after doing it, before you touch the ground.
The factors that affect the time are also plentiful:shield stun, time to release the shield, the time is takes to initiate a dash, walk, jump, and the speed at which the character can dash, walk, jump. In addition, the speed which they can then grab, or attack, at which range is also a factor--if they have a quick, long jab, they don't need to come all the way up to you to hit you, for example.
Tricky Part 2--The headache of calculations
Now this is what I need help with. Mathematically minded people (M2K?) try and follow, this should be simple.
As said previously, all the aspects boil down to two factors: distance and time. These are, in a video game represented by pixels (p) and frames (d). To be safe (S), the result of the f(S)> 0, where the f(S)= tR-tL,
where
tR=time to recover/punish, and
tL is the time of the lag.
Thus, if it took the opponent 30 frames to punish, and you were lagging for 20, your move was safe.
However, as previously mentioned, distance plays a role. But for the result to remain with one unit, somedthing needs to cancel out the distance unit. Luckily, anyone whose taken physics knows:
v=d/t, which can be rearranged to t=d/v.
A quick example:
a man is running 10 meters, at a pace of 5m/s. How long does it take?
v=5m/s
d=10m
t=?s
so t=10m/(5m/s). The m's cancell out, leaving you with 2 seconds=t.
So know we can make our formula:
f(S)=v/d-tL, where
v=the characters running/dashing/walking/jumping speed in d/f, and
d=pixels between the characters.
Now we have the basic idea down, but alas, more completity is in store. Namely, what makes up d? As mentioned earlier, the distance is comprised of range of the attack, and how far the enemy is knocked back, as well as how far you can drift back. So mathmatically, we get to assign variables to all of those. so now we have
d=(R+Kn+Di)p, which when plugged in makes the equasion:
f(S)=v/(R+Kn+Di)-tL
If you are with me so far, the next step should be a piece of cake. In addition to the lag that the move's user has, we need to account for the lag that the opponent is in. So this lag consists of (in a shielding opponent) Shieldstun, time to drop the shield, time to start the motion, and time to start your countermove.
so the basic formula should look like this:
(S)=[(v/d)+tOL]-tL
where
tOL is the time of the opponents lag, which consists of:
SS, shield stun
SD, frames to drop shield
DSt,the time it takes to initiate run, jump, walk, dash
CA, the time to use a counter attack (any move)
so,
tOL=SS+SD+DSt+CA, making the complex equasion:
f(S)={[v/(R+Kn+Di)]+[SS+SD+DSt+CA]}-tL
starting to hurt your mind yet? It gets a bit worse now. Because the counter attack (variable CA) has a range attached to it, the distance the unit has to travel to punish the opponent is lowered. So now, d goes from
d=R+Kn+Di to d=R+Kn+Di-Rca, which makes the complex equasion:
f(S)={[v/(R+Kn+Di-Rca)]+[SS+SD+DSt+CA]}-tL.
Okay, so what was the point of this?
The point of the first section was just to make people think, which is important when playing brawl/melee any fighting game, as the principle is the same. I hope challenges people to think about why they are doing what they are doing, and whether it's more/less effective than they first assumed.
Using this formula, plus frame date, we can come up with a list of safe/unsafe attacks for each character, against a shielding opponent. Actually, with regards to auto-cancelled attacks, this equasion becomes much simpler, as d=Range of your attack-range of there counter attack option.
And, well wouldn't that be useful information? more the most part it would translate into semi-obvious things, but in my opinion you can never have to much information AVAILABLE, even if you don't need/want to look at it.
My request(s)
Unfortunately, I do not have brawl, and I have no means of counting frames. I know most of you have brawl, and no clue whether there is some way to get frame data yet. But if there is, I'm thinking of using this thread as a means of organizing frame data, and using it to break down the game.
I'd like the community's help with this, but if it's impossible, at least the math was fun (in a I haven't taken a math course in 3 years now) sort of way.
Also, comments, advice, more general info/feedback[1] would be appreciated, as in addition to all this mathy stuff this is at heart a Common Sense Guide to Brawl, and as I play more/come up with more things I think need explaining I'll write up about it ^_^
[1] If you think I've missed something, am wrong about something, glossed over something, didn't explain something, thing I'm simply wasting time, then let me know, please, so I can either give up on this or continue/correct things
[Future potential topics I've mused over]
*What if they get hit by an attack?
*the importance of multiple hitboxes
*How perfect shielding works
*What makes good spam+ a look at all projectiles.
*"Comboing" in Brawl
Appendix of other useful threads:
Tyr_03's list of autocanceled aerials
Magus' fram data for Luigi