Designing Smash Bros: The L-Canceling Advantage

[Alondite]

Fear not, this is not going to be another "manual vs. auto L-canceling" debate. Quite the opposite, actually.

Let me preface this by saying that I consider myself to be a video game design theorist of sorts. I enjoy looking at a game through an objective lense in an attempt to discern how it works, and what makes a game a "good" or "bad" game. I've learned a great deal about effective design methods and have a new appreciation for well-designed games; the things that Super Mario Bros. manages to accomplish using only gravity and the primary mechanic of jumping are simply incredible.

Smash Bros. is a game I've loved for a long time. It's very organically designed, and as a result, the gameplay is intuitive, deep, and incredibly dynamic. I've learned over time that a key component of design (as well as overall balance) is balancing advantage and disadvantage to give the player interesting decisions to make. After all, decision-making is a large part of what makes video games unique from passive media like books and movies; how enjoyable would video games be if there was always one obvious best answer or solution to a given situation? Ike's Fsmash has tremendous power and range, but is offset significant startup and endlag. It's up to the player to decide when and how to effectively use it based on those advantages/disadvantages.

Looking at all the design aspects of Smash Bros. eventually led me to L-canceling. I'm sure that most of Smash Boards understand's what L-canceling is: canceling the landing lag from an aerial by a properly-timed press of the shield button. The concept is simple enough. However, things started to become a bit troublesome when I started to weigh the advantages of successfully performing an L-cancel vs the disadvantage of missing one, or not attempting it at all.

Successfully performing an L-cancel significantly decreases the amount of lag suffered from an aerial, but missing it results in the same amount of lag if you hadn't attempted it at all. I trust now that you can see where the problem lies: L-canceling fails to provide an interesting decision for the player to make. There's no penalty for missing an L-cancel, only advantage to successfully performing it so there is no reason to at least not attempt it.

If not for creating more levels of offensive interplay, I'd say that L-canceling detracts from the overall design and "push-pull" of Smash Bros. Obviously L-canceling provides something positive to the gameplay experience and is worth keeping. It it's current state, however, it serves as little more than an arbitrary skill barrier. In this way it's easy to see where the argument for auto L-canceling comes from. And you know what? They are correct; automatic L-canceling is an objectively better solution than what is currently used. I don't believe, however, that it's the best solution because is takes the choice out of the player's hands entirely. Obviously there must be a better solution. I asked myself "how can we make L-canceling both provide a meaningful decision for the player to make as well as further rewarding skillful play even moreso than the current system of L-canceling?"

I think I've come up with a solution that will please both parties. It really wasn't that difficult once I looked at it from a design perspective; there simply is no disadvantage to screwing up an L-canceling, making it functionally identical to not attempting one at all. The solution? Simply add a disadvantage to L-canceling. I deemed the most effective way to do this would be to add in a lag penalty for missing an L-cancel, resulting in even greater lag than if you hadn't attempted one at all.

Let's look at this example. Fox's dair can be tough to L-cancel due to the multi-hit nature affecting the timing. If missing an L-cancel resulted in even more lag than not doing one at all, it would give players a decision to make. "Do I have faith enough in my skill to pull it off? Should I even attempt it?" I think you can see how this further rewards skillful play as well. Players who can readily pull it off will do so regularly, but players who might not be as adept with the timing may either screw it up, or not even attempt it, giving the advantage to the more skillful player. This also opens up a world of character design and balance options that I won't get into now, but that don't require much effort to see.

Obviously this presents some design challenges in itself. How do you balance the advantage vs disadvantage? If the lag penalty is too severe, players won't attempt it, and if it's not severe enough, or if players will easily have success then it once again fails to provide an interesting decision to make. In the former case, simply reducing the lag penalty works well enough, but in the second case, extending the landing lag won't matter if players are always hitting L-cancels, so simply decrease the timing window.

There are even more design elements to consider. For example, how can failure/success be communicated to the player so that they can make informed decisions instead of guessing? Project M already does this in a rather elegant way, which I consider to be one of, if not its single greatest design success: that little white flash upon successfully performing an L-cancel. Simply add a similar type of flash when the player whiffs an L-cancel. Yellow and red are already taken by various armor types, so to prevent overlap and player confusion something like blacking out the character for a moment and/or increasing the size of the dustcloud could provide a fitting alternative that intuitively communicates to the player: The black flash opposite to the white success flash, and the dustcloud simulating a harder landing.

I feel that such a mechanical change to L-canceling adds considerable depth while not altering the actual function and resulting options in any way, thus being all positive with no negative.

I fear, however, that codding issues and/or the political nature of Project M development (wanting what gives you an advantage rather than what is ultimately the best solution for everyone) could get in the way. I'm interested in hearing what you all think. If the feedback is positive, I'd be more than happy to contribute some other similar design changes that could be made without sacrificing anything or altering the way the game is played in a significant way.

 

[Kink-Link5]

I proposed something similar that could be done in a different environment.

L-canceling, I find, has problems BECAUSE of how easy it is to use, on top of the problems addressed with "why would you ever not do it." Lack of a fail window means L, R, and Z can all be mashed in quick succession to safely option select and have the L-cancel occur over any variety of timings. The 6 frame window is also relatively large, but that itself is fine enough. What would be better to encourage proper timing would be to include a 10-20 or so frame fail window where-in a shield input sets a flag. If a mash L-cancel is then attempted, the landing lag is actually increased.

If the flag is set but shield inputs are not mashed, the aerial just has its normal landing lag.

Couple this with higher hitlag (and appropriately higher shieldstun attributed on a per-move basis) and the proper window for an L-cancel becomes something to actively pay attention to while providing proper reward for attentive players and punishing those attempting to abuse the mechanic through mashed L-cancels. Along with attention put into seeing what out of shield options can punish a: Canceled aerial, normal aerial, and failed aerial, the entire dynamic of player interaction is increased. If it were so desired, there could even be a visual indication for what kind of input is done to assist the shielding opponent: a proper L-cancel flashes the executor white, and a mash input flashes red or some other deep "negative" colour.

What is the downside of this?​

Having it in a PM environment would, however not be conductive to the players that have come to utilize the option select method, and this discussion is very very prone to a lot of negative and otherwise ****ty arguments. L-canceling in general is not the kind of topic that breeds healthy conversation.

 

[Alondite]

I agree that success through mashing undermines L-canceling as "tech skill" (an ambiguous and generally terrible term); if you're able to pull it off by simply mashing, then the skill required diminishes greatly.

I think we are basically saying the same thing here, though my solution aims to eliminate the use of mashing altogether, rather than designing around it.

"Having it in a PM environment would, however not be conductive to the players that have come to utilize the option select method"
I take issue with this. Not your statement, but the fact that it's true. It is detrimental to the design if multiple inputs (especially if they have varying degrees of input difficulty) get you to the same result. It's also not meaningful decision-making because the end result is the same. Designing around design flaws just creates more issues.

 

[OSCA MIKE]

sometimes you just gotta bite the bullet and do whats best for the game

 

[Alondite]

I agree. Project M can be improved considerably without abandoning Melee's design core at all. If fact, it already does that in a number of ways! I just don't understand why some things can be changed, but others are too "sacred." Where is that arbitrary line drawn?

 

[BTmoney]

Hopefully this stays civil and reasonable. I agree with you Alondite as far as your intentions go however I do have my qualms with your suggestion.
How does this not create a larger barrier into high level play? It is obviously in your best interest to always do it.


Not being able to L-cancel consistently just makes bad players and non-elite players worse. It does not change the nature of L-canceling. At a high level you shouldn't ever miss one, at my level I shouldn't ever miss one.
I beleive you want to to create a more advantageous upside to not L-canceling.
Such as less damage or hitstun (yes you hit them, then L-cancel but you get it).

 

[Problem2]

Adding more landing lag for mistiming an L-cancel just adds a more arbitrary punishment to attempting it. Even with the knowledge that you could be worse off than before, the optimal answer is to just never miss one in the first place. The change would have a big effect on mid-level play, but would have less effect on high level play.

...But there are actually ways to punish a player for doing L-cancels anyways. When you press the the L or R button during stun outside the tech window, you're given a tech penalty. You're not allowed to tech for a certain number of frames. This actually happens a bit in competitive play. A player will begin a shffl'd aerial, only to be hit before he touches the ground and l-cancels, but he already hit the L button anyways. Because of this, he is not allowed to tech when he touches the ground shortly after, and the opponent can get a free combo starter.

In addition, the timing for pressing L or R after an aerial changes depending on if you land a clean hit, hit their shield, or miss their body entirely. Sometimes, a defending player can attempt to side-step the attack as a way to encourage missing the timing on the L-cancel, when the attacker is expecting to hit the shield.

If L-cancelling is too arbitrary in design, then the solution is to make not L-cancelling have some sort of purpose to where there are times when not L-cancelling is simply better than L-cancelling.

EDIT: Here's an idea to toss around. Make it so that L-cancelling only reduces stun if the aerial you are cancelling hits a shield or body. If the move whiffs, the attack has more lag. This does a few things:

1.) It most importantly provides decision and reason when to and when not to use it, and introduces a new layer of Yomi into the game. In obvious situations, l-cancelling is bad to use when the move is no where near hitting the opponent, but good to use if you see the move land. However, when both players are moving really fast, things can get a little bit iffy. Then, it becomes up to the player's instinct, gut, or knowledge of the opponent to make an educated guess. Will my move connect or hit a shield? Will my opponent try to spot dodge to punish my pressure? These are things you would have to take into consideration, which produces a more yomi kind of situation.

2.) It changes the way footsies are done in Smash a bit. In melee, it's possible to throw out an aerial as a poke or "gtfo" sort of move to try and control space (see Marth and C.Falcon's n-airs). Even if they whiff, they are accomplishing the job of control space for you. With the punishment, players will not l-cancel, unless they think it would hit. So in this case, throwing out aerials is a little less safe than before, so a player has to be more cautious of his attack choices.

3.) This changes the way shield pressure is handled. Attacking players have to be a bit more careful when going nuts on a shield. If the opponent side-steps during your mix-ups and you L-cancel, you'll be left vulnerable. Now, if the new l-cancel was implemented into smash, I believe it would also require tweaking to everyone's landing lag and the duration of spot dodges. Ideally, a player putting on the pressure should be allowed to punish a predicted spot dodge by not l-cancelling thier aerial. (Naturally, there are other ways to punish too such as a tomohawk grab)

 

[Alondite]

EDIT: Ah someone else posted before I finished. This was in response to Tactician.

Your point and concerns are both completely valid. I'll respond as clearly as I can.

It does create a larger barrier into higher level play, in a way (which I'll get to), but does not ultimately make L-canceling more difficult to perform. "Dexterity" is one of the pillars of video game skill, so a technique that stresses it isn't necessarily bad. I'm against complexity that doesn't lend itself to more depth. Requiring a double counter-clockwise circle movement before a f-tilt would be arbitrarily more difficult to execute, for example, and wouldn't contribute anything positive to the gameplay.

"Not being able to L-cancel consistently just makes bad players and non-elite players worse. It does not change the nature of L-canceling. At a high level you shouldn't ever miss one, at my level I shouldn't ever miss one."

I have issues with each part of that which I'll address separately. With the first, Kink-Link5 already pointed out the issue. People are able to successfully L-cancel by just fudging it. Should these players be rewarded with the same advantage as the players who've learned how to time it properly? Should the more skilled player not be rewarded for greater dexterity? My proposed mechanical changes stress the timing, thus preventing being able to just fudge your way through it. Is not the point of competitive play to determine the more skillful of players? Is not the whole supporting argument for manual L-canceling that it adds an element of "technical skill"? In effect, you can achieve success without the skill with the current mechanics. Does that make sense?

As for changing the nature of L-canceling, that's precisely what I wanted to avoid doing. My changes are meant to add depth and reward skill, not change the nature of the mechanic.

I feel as if the 3rd part directly contradicts your first. What defines "high level play"? If L-canceling should have a 100% success rate at "high level play," then what function does the arbitrary button-press serve? If everyone is hitting every L-cancel anyway, then why not just make it automatic, seeing as how the added complexity isn't adding anything to the game. Does it not serve to simply add an arbitrary skill barrier? Isn't that what you argued against in the first sentence of that line?

If any part of that sounds aggressive or insulting, I apologize. It wasn't my intent.

 

[Alondite]

Adding more landing lag for mistiming an L-cancel just adds a more arbitrary punishment to attempting it. Even with the knowledge that you could be worse off than before, the optimal answer is to just never miss one in the first place. The change would have a big effect on mid-level play, but would have less effect on high level play.

Is failing to perform the move properly an arbitrary reason for punishment? Obviously the optimal answer is to never miss one, but in practice it's not always so simple. At the theoretical skill-ceiling, one would never miss an L-cancel, but to reach such a level requires a honing of skill. Is development of skill arbitrary as well?

...But there are actually ways to punish a player for doing L-cancels anyways. When you press the the L or R button during stun outside the tech window, you're given a tech penalty. You're not allowed to tech for a certain number of frames. This actually happens a bit in competitive play. A player will begin a shffl'd aerial, only to be hit before he touches the ground and l-cancels, but he already hit the L button anyways. Because of this, he is not allowed to tech when he touches the ground shortly after, and the opponent can get a free combo starter.

I had already considered this when coming up with changes. It's not that he incorrectly timed the L-cancel, it's that the other player hit him. It becomes second-nature to L-cancel following a SHFF aerial, which requires a deal of commitment in itself. Is the disadvantage of being hit while attempting to L-cancel significant enough to offset the advantage and make it an interesting decision? I don't believe so. In fact, I'd argue that this actually helps my case, because attacking someone out of a SHFF aerial is one of the very few counters to such an offensive move. L-canceling is so powerful a technique that the entire metagame has been built around it. You don't need to look any farther than the prevalence of SHFFLing in competitive play to see what I mean.

In addition, the timing for pressing L or R after an aerial changes depending on if you land a clean hit, hit their shield, or miss their body entirely. Sometimes, a defending player can attempt to side-step the attack as a way to encourage missing the timing on the L-cancel, when the attacker is expecting to hit the shield.

I already addressed this in my original post. It simply changes the timing, it does not make it a meaningful decision. There is only advantage to gain from success, and no real disadvantage from failure.

If L-cancelling is too arbitrary in design, then the solution is to make not L-cancelling have some sort of purpose to where there are times when not L-cancelling is simply better than L-cancelling.

That would be the ideal solution, yes, but would require significant design changes. I'd considered that as well, but I felt that my changes yielded as similar result as possible without having to radically change the game.

Responses in bold.

 

[Alondite]

EDIT: Here's an idea to toss around. Make it so that L-cancelling only reduces stun if the aerial you are cancelling hits a shield or body. If the move whiffs, the attack has more lag. This does a few things:

1.) It most importantly provides decision and reason when to and when not to use it, and introduces a new layer of Yomi into the game. In obvious situations, l-cancelling is bad to use when the move is no where near hitting the opponent, but good to use if you see the move land. However, when both players are moving really fast, things can get a little bit iffy. Then, it becomes up to the player's instinct, gut, or knowledge of the opponent to make an educated guess. Will my move connect or hit a shield? Will my opponent try to spot dodge to punish my pressure? These are things you would have to take into consideration, which produces a more yomi kind of situation.

2.) It changes the way footsies are done in Smash a bit. In melee, it's possible to throw out an aerial as a poke or "gtfo" sort of move to try and control space (see Marth and C.Falcon's n-airs). Even if they whiff, they are accomplishing the job of control space for you. With the punishment, players will not l-cancel, unless they think it would hit. So in this case, throwing out aerials is a little less safe than before, so a player has to be more cautious of his attack choices.

3.) This changes the way shield pressure is handled. Attacking players have to be a bit more careful when going nuts on a shield. If the opponent side-steps during your mix-ups and you L-cancel, you'll be left vulnerable. Now, if the new l-cancel was implemented into smash, I believe it would also require tweaking to everyone's landing lag and the duration of spot dodges. Ideally, a player putting on the pressure should be allowed to punish a predicted spot dodge by not l-cancelling thier aerial. (Naturally, there are other ways to punish too such as a tomohawk grab)

Adding more landing lag for mistiming an L-cancel just adds a more arbitrary punishment to attempting it. Even with the knowledge that you could be worse off than before, the optimal answer is to just never miss one in the first place. The change would have a big effect on mid-level play, but would have less effect on high level play.

...But there are actually ways to punish a player for doing L-cancels anyways. When you press the the L or R button during stun outside the tech window, you're given a tech penalty. You're not allowed to tech for a certain number of frames. This actually happens a bit in competitive play. A player will begin a shffl'd aerial, only to be hit before he touches the ground and l-cancels, but he already hit the L button anyways. Because of this, he is not allowed to tech when he touches the ground shortly after, and the opponent can get a free combo starter.

In addition, the timing for pressing L or R after an aerial changes depending on if you land a clean hit, hit their shield, or miss their body entirely. Sometimes, a defending player can attempt to side-step the attack as a way to encourage missing the timing on the L-cancel, when the attacker is expecting to hit the shield.

If L-cancelling is too arbitrary in design, then the solution is to make not L-cancelling have some sort of purpose to where there are times when not L-cancelling is simply better than L-cancelling.

EDIT: Here's an idea to toss around. Make it so that L-cancelling only reduces stun if the aerial you are cancelling hits a shield or body. If the move whiffs, the attack has more lag. This does a few things:

1.) It most importantly provides decision and reason when to and when not to use it, and introduces a new layer of Yomi into the game. In obvious situations, l-cancelling is bad to use when the move is no where near hitting the opponent, but good to use if you see the move land. However, when both players are moving really fast, things can get a little bit iffy. Then, it becomes up to the player's instinct, gut, or knowledge of the opponent to make an educated guess. Will my move connect or hit a shield? Will my opponent try to spot dodge to punish my pressure? These are things you would have to take into consideration, which produces a more yomi kind of situation.

2.) It changes the way footsies are done in Smash a bit. In melee, it's possible to throw out an aerial as a poke or "gtfo" sort of move to try and control space (see Marth and C.Falcon's n-airs). Even if they whiff, they are accomplishing the job of control space for you. With the punishment, players will not l-cancel, unless they think it would hit. So in this case, throwing out aerials is a little less safe than before, so a player has to be more cautious of his attack choices.

3.) This changes the way shield pressure is handled. Attacking players have to be a bit more careful when going nuts on a shield. If the opponent side-steps during your mix-ups and you L-cancel, you'll be left vulnerable. Now, if the new l-cancel was implemented into smash, I believe it would also require tweaking to everyone's landing lag and the duration of spot dodges. Ideally, a player putting on the pressure should be allowed to punish a predicted spot dodge by not l-cancelling thier aerial. (Naturally, there are other ways to punish too such as a tomohawk grab)

I don't need to respond to this individually, because I think it has one major flaw. I understand what you're getting at here, and in theory it makes sense. I still prefer my solution, but I see where you're coming from. However, I believe that your second point is more detrimental that it is positive.

I use SHFFLed nairs as Falco for defensive purposes, a barrage of attacks with lingering hitboxes to force my opponent to be more deliberate in their approach. I don't often land these attacks because they're very telegraphed, but they serve their purpose nonetheless. If I were forced to not l-cancel them, else suffer a penalty, my opponent would likely be able to close in and attack me. I feel that the negatives of this outweigh the positives, and that in detracts from the overall number of viable options provided to the played, thus reducing the depth. The defensive options are interesting, though.

I believe, however, that since the timing of L-cancels changes based on hitting your opponent, etc., that my solution essentially has the same effect. Your opponent predicts you, moves in range and shields, thus throwing off your timing and causing you to whiff the L-cancel and suffer the extended lag.

"Now, if the new l-cancel was implemented into smash, I believe it would also require tweaking to everyone's landing lag"

I believe that it opens up the design potential, requiring stricter L-canceling windows for moves which benefit more from a successful L-cancel, but I don't think it would require it explicitly.

EDIT: I'd like to add: your solution doesn't address the issue when landing an attack (such as during a combo) is a sure thing. Even if the attack isn't a sure thing, it's going to be basic conditioning for a player to L-cancel during a combo. In this case, your solution punishes the player for whiffing during the combo rather than messing up the L-cancel, which causes a change in an established element of the game that doesn't need to be changed. And as you said youself, even missed moves have purpose, so we shouldn't punish players for using them.

 

[BTmoney]

I have issues with each part of that which I'll address separately. With the first, Kink-Link5 already pointed out the issue. People are able to successfully L-cancel by just fudging it. Should these players be rewarded with the same advantage as the players who've learned how to time it properly? Should the more skilled player not be rewarded for greater dexterity? My proposed mechanical changes stress the timing, thus preventing being able to just fudge your way through it. Is not the point of competitive play to determine the more skillful of players? Is not the whole supporting argument for manual L-canceling that it adds an element of "technical skill"? In effect, you can achieve success without the skill with the current mechanics. Does that make sense?

Nah, it's fine lol.

I feel like if I address this part, I can address your post more concisely without having to rant and stand for things that are not going to further the discussion.
If the idea is to reward more skillful play then here:

Hypothetical-L-Canceling:

• Say, 10 frame window while doing an aerial about to hit the ground (this number is arbitrary)
• Aerial X has Y frames of lag
• For every frame you get closer to perfection ( perfection in this case would be L-canceling on the last frame while you are still airborne before you are on the ground), subtract 1 frame of Landing Lag

Ex:
"Random Nair" has 15 frames of lag, if you L-cancel on the 1st frame of being able to L-cancel (meaning you are still fairly high in the air), you would subtract 1 frame of lag. If you L-cancel on the last (which is the 10th frame, or the 1st frame of the biggest benefit) frame of being able to L-cancel (being as close the ground as possible), subtract 10 frames of lag. The numbers are arbitrary but I think this is a good concept. You get a reward for being more frame perfect. It is easy to L-cancel somewhere between the hypothetical 1st and 7th frame but is it easy to get it in the 8th to 10th latest frames consistently? By simply playing the number game only the best players will hit those low frames. Think of this as a standard normal curve

The majority of players who can L-cancel will be getting their L-cancels in the 5th frame+/- window. This accounts for 68.2% of the player population. As you move a standard deviation upwards, 13.6% of players will be getting their L-cancels in the 4-2 frames window while 2.15% would be getting the 1 frame window. It is hard to explain stats if there are not derived from actual results and I don't have a standard deviation. The standard deviation is made up but that is just incase you are not familiar with Statistics (I have no idea lol, I am not trying to be condescending or anything). This is all hypothetical but in theory does work and is applicable.
That benefits those with more skill. This also requires temperance because if you push L twice, you get the earlier benefit and not the last frame/later frame benefits.
Does that make sense?

If you want to take this concept out of a test-tube where ever player is not perfectly bad/good/and has consistent consistency then you can apply the same curve this way. 68.2% of the time, you can expect to L-cancel somewhere in the middle frames. 13.6 percent of the time, you can expect to L-cancel somewhere in the uppermost frames and 2.15 percent of the you can expect to L-cancel in the latest and last frames possible. Of course practice will allow you to hit those later frames more often which this curve does not and cannot display. There is also a tradeoff of waiting too late trying to get the perfect L-cancel and not doing it at all.


Basically, naturally the better players will get their L-cancels closer to frame-perfect than average or bad players. Also naturally fewer people will be able to do frame-perfect L-cancels. However practice is reasonable and this does not create an insurmountable gap between players.

 

[DrinkingFood]

Bookmarking or w/e we should call it now so I can read it when I don't have an essay on bioethics to write. lol
Glad you made this thread though, alondite. If you could, when you make a game design thread or anything like that, link us to it so we can have more varied discussions.

But to throw in a couple of pennies, I support the idea of punishing missed L-canceling attempts.

 

[BTmoney]

Bookmarking or w/e we should call it now so I can read it when I don't have an essay on bioethics to write. lol
Glad you made this thread though, alondite. If you could, when you make a game design thread or anything like that, link us to it so we can have more varied discussions.

But to throw in a couple of pennies, I support the idea of punishing missed L-canceling attempts.

When you get done with that paper gimme some feedback
I am fairly sure you are not an idiot judging from what I've seen from you on this forum and I value quality opinions.

lol

edit: I seriously think the concept of what I posted is the most well rounded idea I've seen so far. It benefits various degrees skill and there are trade offs in the timing (middle frames are easiest etc.). The numbers are just for the sake of example.

 

[Alondite]

Ah, hello DF! Was wondering when you would show up.

Nah, it's fine lol.

I feel like if I address this part, I can address your post more concisely without having to rant and stand for things that are not going to further the discussion.
If the idea is to reward more skillful play then here:

Hypothetical-L-Canceling:

• Say, 10 frame window while doing an aerial about to hit the ground (this number is arbitrary)
• Aerial X has Y frames of lag
• For every frame you get closer to perfection ( perfection in this case would be L-canceling on the last frame while you are still airborne before you are on the ground), subtract 1 frame of Landing Lag

Ex:

"Random Nair" has 15 frames of lag, if you L-cancel on the 1st frame of being able to L-cancel (meaning you are still fairly high in the air), you would subtract 1 frame of lag. If you L-cancel on the last (which is the 10th frame, or the 1st frame of the biggest benefit) frame of being able to L-cancel (being as close the ground as possible), subtract 10 frames of lag. The numbers are arbitrary but I think this is a good concept. You get a reward for being more frame perfect. It is easy to L-cancel somewhere between the hypothetical 1st and 7th frame but is it easy to get it in the 8th to 10th latest frames consistently? By simply playing the number game only the best players will hit those low frames. Think of this as a standard normal curve

The majority of players who can L-cancel will be getting their L-cancels in the 5th frame+/- window. This accounts for 68.2% of the player population. As you move a standard deviation upwards, 13.6% of players will be getting their L-cancels in the 4-2 frames window while 2.15% would be getting the 1 frame window. It is hard to explain stats if there are not derived from actual results and I don't have a standard deviation. The standard deviation is made up but that is just incase you are not familiar with Statistics (I have no idea lol, I am not trying to be condescending or anything). This is all hypothetical but in theory does work and is applicable.
That benefits those with more skill. This also requires temperance because if you push L twice, you get the earlier benefit and not the last frame/later frame benefits.
Does that make sense?

If you want to take this concept out of a test-tube where ever player is not perfectly bad/good/and has consistent consistency then you can apply the same curve this way. 68.2% of the time, you can expect to L-cancel somewhere in the middle frames. 13.6 percent of the time, you can expect to L-cancel somewhere in the uppermost frames and 2.15 percent of the you can expect to L-cancel in the latest and last frames possible. Of course practice will allow you to hit those later frames more often which this curve does not and cannot display. There is also a tradeoff of waiting too late trying to get the perfect L-cancel and not doing it at all.


Basically, naturally the better players will get their L-cancels closer to frame-perfect than average or bad players. Also naturally fewer people will be able to do frame-perfect L-cancels. However practice is reasonable and this does not create an insurmountable gap between players.

I think this is actually a cool idea. However, I have a few issues:

1. I can imagine this being a nightmare to code. I think we should try to balance "ideal" and "practical."
2. While it does further reward skillful play and dexterity, it doesn't address the issue of "to L-cancel, or not to L-cancel," it merely adds varying degrees of advantage, but still presents no disadvantage.
3. Are the different degrees of success going to be significant? That is, are they going to have a real impact on gameplay?
4. This one stretches a little, but I think it's still valid. Could performing an early L-cancel, which would typically yield less of an advantage, actually take you out of the "no tech window" should you be hit out of the aerial? This presents a possible advantage to doing something sub-optimally, and blurs the line of whether a sub-optimal L-cancel would truly be less advantageous than an optimal one.

 

[-Ran]

There is a penalty to missing an L Cancel in Melee. When you're properly timing your L-Cancels, it opens up every aerial that your character has. For some characters, this means that they are able to Dair, Bair, or even Fair on the stage when they other wise couldn't. Failure to L-Cancel the aerial results in a completely different level of play be it approaching or defensive. Your argument is, "There's no penalty for missing an L-cancel, only advantage to successfully performing it so there is no reason to at least not attempt it."

I disagree with your statement. L canceling shapes play styles and changes how you perceive and interact with the game. I submit a scenario to you:

Marth and Falco on FD. Fresh stock. Marth decides to go for a quick kill early on when he finds himself above Falco. Marth Dairs.

a) If Marth dairs and L Cancels, he is able to grab Falco. The added percent from the Dair, makes it very manageable to follow the DI of Falco during the chain grab. Due to this, it's very likely that Marth will not drop this grab, resulting in at edge guard opportunity. Dair is a great option.

b) If Marth fails to L Cancel, he has just used his most laggy move against Falco. Even with a hit on Falco, Falco is going to be able to retaliate before Marth can recover. Due to this gap in timing, Falco is able to shine. This shine at a fresh percent can [and should] lead to over half a stock worth of damage, or an edge guard pending stage location at the time. Dair is a bad option.

Yes, at peak levels, players are operating within the realm of essentially being mistake free. This means, the conversation of A vs B disappears completely; however, the players still had to get there. There was a time in the Marth player's life, when he could not reliably do A. This changed how he played the game, it made him choose when he should attempt to do the move. Eventually, the player reached a level of confidence where he could apply A to his game every time. This can be said about any implementation of L Canceling from Dair to Shine to aerial approaches with most of the cast.

By the time it reaches the top level, it is just a competitive tax. This tax of dexterity and timing could potentially add up when weighed with the other numerous buttons that will be pressed during any single tournament weekend; however it is unlikely it will hurt the seasoned player, but for young players, the ability to L Cancel perfectly is a real and tangible force in their matches. It is not something that needs to be altered, in my view, at the high or low level. The penalty for doing it wrong is a great one: the denial of options.

 

[DrinkingFood]

Man I want to respond to this stuff so badly but I don't have time... and since I'm pulling an all nighter to finish this, I'll probably sleep most of this afternoon. *sigh* first world problems...

 

[Alondite]

There is a penalty to missing an L Cancel in Melee. When you're properly timing your L-Cancels, it opens up every aerial that your character has. For some characters, this means that they are able to Dair, Bair, or even Fair on the stage when they other wise couldn't. Failure to L-Cancel the aerial results in a completely different level of play be it approaching or defensive. Your argument is, "There's no penalty for missing an L-cancel, only advantage to successfully performing it so there is no reason to at least not attempt it."

I disagree with your statement. L canceling shapes play styles and changes how you perceive and interact with the game. I submit a scenario to you:

Marth and Falco on FD. Fresh stock. Marth decides to go for a quick kill early on when he finds himself above Falco. Marth Dairs.

a) If Marth dairs and L Cancels, he is able to grab Falco. The added percent from the Dair, makes it very manageable to follow the DI of Falco during the chain grab. Due to this, it's very likely that Marth will not drop this grab, resulting in at edge guard opportunity. Dair is a great option.

b) If Marth fails to L Cancel, he has just used his most laggy move against Falco. Even with a hit on Falco, Falco is going to be able to retaliate before Marth can recover. Due to this gap in timing, Falco is able to shine. This shine at a fresh percent can [and should] lead to over half a stock worth of damage, or an edge guard pending stage location at the time. Dair is a bad option.

I have one question which completely deconstructs your argument: How does scenario B differ from not attempting an L-cancel? I'm perplexed as to why you offered this argument, even after quoting my statement. If scenaria B is so unfavorable, then why not L-cancel? And if you whiff the L-cancel, how does scenario B differ from not attempting one? Missing an L-cancel puts you into a standard state as if you hadn't L-canceled. This is not a disadvantage, this is the zero point. An advantage is when you are above the zero point, and a disadvantaged is below the zero point. Successfully L-canceling puts you at, say +1, while both whiffing and not attempting an L-cancel put you at 0. This is an unbalanced and poorly-designed maneuver.

Yes, at peak levels, players are operating within the realm of essentially being mistake free. This means, the conversation of A vs B disappears completely; however, the players still had to get there. There was a time in the Marth player's life, when he could not reliably do A. This changed how he played the game, it made him choose when he should attempt to do the move. Eventually, the player reached a level of confidence where he could apply A to his game every time. This can be said about any implementation of L Canceling from Dair to Shine to aerial approaches with most of the cast.

Unfortunately, this does not address the issue. It's simply a statement on the nature of skill development. Even when he can't perform an L-cancel, there's no reason not to try. The fact that he avoids using the dair is irrelevant.

By the time it reaches the top level, it is just a competitive tax. This tax of dexterity and timing could potentially add up when weighed with the other numerous buttons that will be pressed during any single tournament weekend; however it is unlikely it will hurt the seasoned player, but for young players, the ability to L Cancel perfectly is a real and tangible force in their matches. It is not something that needs to be altered, in my view, at the high or low level. The penalty for doing it wrong is a great one: the denial of options.

The penalty for doing it wrong is no more than for not doing it at all. I stated that several times and in several posts, yet you ignored it entirely in your response. If you're going to respond, please read the post carefully so that we avoid redundant topics. In its current state, L-canceling serves no purpose other than to put up arbitrary dexterity skill barriers. Functionally, it's no different than auto L-canceling. However, neither of them provide meaningful decision-making. Choosing not to use Marth's dair because you can't L-cancel is caused by the dair's landing lag; it's not about choosing to L-cancel or not, it's about choosing to use the dair or not. Again, that's why it was irrelevant to point out.

Responses in bold.

 

[-Ran]

Perhaps I wasn't clear. I thought I was, but perhaps it was my lack of sleep, or your desire to respond to each paragraph separately that reduced the effectiveness of my writing as a cohesive whole. We'll never know.

L canceling produces options. A player who cannot L cancel properly creates openings for their opponent when the fail. Every movement we do is weighed with risk vs potential for damage/kill. When one cannot L cancel, they opt out of the situation due to the reduced safety. A Marth that cannot L Cancel his Dair reliably, will never attempt A. This is applied to other methods of approaching/shield pressure/spacing in the game that rely on L Canceling. L Canceling is a gateway to new methods of play for a character, and when you cannot do it reliably, you will not employ those methods in your game.

To sum it up, the decision for the Marth player after having B happen to him, would be: C, don't dair Falco. The penalty becomes the removal of A being an option.

Sleep time.

 

[KayB]

I'm abnormally bad at L-cancelling. I can never get the timing down. Reading this thread, I can assume I am a terrible player lol.

 

[DrinkingFood]

Just practice it more, you'll get it down. Things like these often take that one click of the brain to register that it's perfected the timing, followed by sleep to help hardwire it into your brain. I had the same thing with PSing. Even if it takes you more time than others, persistence is a sign of your devotion to mastering the game, and understanding that builds confidence.

At least, that's how I learn things. I think.

 

[\Apples]

The idea is well-founded, but I think the oversight here is displayed in the argument presented by those who make the claim that adding a punishment to failing an L-cancel only increases the disparity of skill level between players ultimately voids the need for implementing a punishment.

I'm a game designer and developer as well, I've studied design formally and informally for a long time now, and one thing I've learned is that there really are no absolute formulas for good design. Good design is relative to the goal of the game's design. So, when designing a game, first we must ask ourselves 3 questions: "What is the player's goal?", "What tools am I providing the player to meet this goal?" and "How am I challenging the player to meet this goal, or in what ways do the tools I've provided to the player cause a hindrance to their success?"

In this case, the game is multiplayer competitive; the object of the game is to defeat your opponent (high level abstraction of the goal). So, how can the player do this? Well, we all know the tools available to us, movement, jumping, attacks, shielding, interaction with the stage (even the stage is a tool); the challenge comes from the fact that your opponent has access to the same or a similar set of tools, each of these tools has advantages/disadvantages which makes overcoming your opponent a tactical process, meaning the game has a prevalent metagame.

But we all already know that. What is not being addressed here is how accessible you (as a designer) want the game to be. The L-cancel mechanic is one which creates quite the disparity of skill level between players who are aware of its existence and use it and those who do not, those who do L-cancel and understand its advantages and put those advantages to use have a great advantage over those that don't, it's simple. What you're suggesting is that we change this, ultimately, by creating a downside to L-canceling.

Here's the trouble though, the method you use is likely to shift the largest average skill gap toward those players who are more consistent with the technique. In short, the only thing this change really affects fundamentally is the metagame.
This is basically how your change would affect the metagame:

So the question really is... "Do I want my game to be very difficult to become the best at via proper execution or do I want my game to be very difficult to become the best at via proper tactics?" Adding a downside to L-canceling only makes execution more important, it does not add depth to the tactical process at all. One is not better than the other, it all just depends on what kind of metagame you want you game to shape.

Personally, I prefer L-canceling the way it is. I think it's completely fine, it's already difficult enough to properly execute actions in this game. Just because it doesn't fall into the "push/pull" of your design theory does not make it fundamentally flawed, it just means your theory may be flawed.

 

[Mic_128]

All this seems to do is encourage less-good players to not L-cancel because unless perfect, is a much bigger disadvantage than not even trying.

Regardless, L-cancelling is to be done in the style of Melee's, something the P:M team have said over and over. And over.

They aren't changing it. If you wish to discuss using it for other mods, feel free to repost elsewhere.