• 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!

Theory Craft and Mix up game: Seeing where "on paper" meets "in match" ~Essay

OverLade

Smash Hero
Joined
Jun 19, 2006
Messages
8,225
Location
Tampa, FL
I write this essay because oftentimes browsing the boards on this forum I see two types of players. There are people who have experience in tournaments and people who analyze the mechanics of the game. Both types of players have intelligent posts, helping advance the meta-game and bringing up relevant information. However oftentimes while reading different character boards I see a lot of what has come to be known as “theory craft”. Much can be learned from analyzing frame data and what’s possible, but when this is done it often ignores the reality of an actual match.

I have a unique perspective on this game because I spent roughly a whole year not going to tournaments, but watching videos, keeping track of results, and practicing with my brother. It wasn’t until I went to tournaments that I was able to understand what I’m writing now however, because often times what should happen on paper isn’t what happens in an actual match, and I’m here to explain why. Most of this discussion I’m going to use matchups I understand fluently, as well as matchups that people will see often in tournament and videos. Meta knight vs. Snake, Meta knight vs. Dedede, and Snake vs. Dedede will be heavily used because I understand these matchups at the highest competitive level from both sides. Just for credibility reasons here are some videos of me playing the matchup (as unfortunately some people on this board won’t believe anything you say unless you have a video/results of beating someone good :laugh:. I will cite videos of myself playing as well as videos of other generally regarded top players so that readers will see where theory meets reality. That’s the point of this thread really, to explain where theory meets reality and where it doesn’t. I’m going to try extremely hard to keep this focused, though I could write hundreds of pages on this and I’m likely to go off on tangents that go into more detail.

Now I’m going to talk about the WRONG kind of theory craft. I often see these kind of statements on the boards. (I’m not exactly sure if these numbers are correct but I read them somewhere, for purposes of this thread the numbers aren’t important):
MK vs. Snake on the ground.

Metaknight’s forward tilt comes out in 4 frames. Metaknights forward tilt slightly outranges the first hit of snakes forward tilt and comes out faster. MK’s forward tilt is disjointed, and can be slowed down. This means that it beats Snakes dash attack, dash grab, and sidestep. This means the only thing that should beat it is a perfect powershield of the first hit to a shieldgrab.

Snake’s forward tilt comes out in 5 frames. The second hit of Snakes forward tilt outranges all hits of MK’s forward tilt but takes 8 frames to come out.

Theorycraft: “Due to the above data, MK should beat snake on the ground.” Most of the time, theory craft isn’t even this good. A lot of the time you see even worse kind of theory, like saying Snake should never beat Dedede because Dedede’s walking grab beats all of snakes approach options. It beats all of Snakes approach options EXCEPT sidestep. But a theory crafter would tell you that you should never run up to Dedede and sidestep because they could read it. Of course they could read it but this game is about what your opponent is expecting at an individual mix up. In all matchups you have to take a risk in order to get reward. Meta knight is an easy character to use because the risk reward on most of his moves is very lopsided. Anyway back on track.

(Back to the above bolded) Now in the heat of an actual match, this kind of theory is irrelevant. Now some of you may ask, why? The simple answer to everything that doesn’t match up with theory that you may see in a match is this.
This game is entirely about expectation and speed of recognition, not about what “should” mechanically happen.
Theory craft is often based on what someone “should” be able to do. For example, Meta knight and Dedede are at a standing position, neither character moving. Meta knight jumps and tornados toward Dedede. Dedede has a few options to beat/avoid tornado. He can shield/roll backwawrds/Utilt. But often when watching a match, you see at Dedede player sidestep and then get sucked into the tornado afterward. This happened because the Dedede player was expecting something else, and simply reacted to Meta knight’s movement in anticipation of a dash grab/dash attack. To make this relevant here’s a video where I’ll point out another example of reacting, and guessing wrong.

Redhalberd (MK) vs. Seibrik (Dedede)
http://www.youtube.com/watch?v=-8bVjiJEaxM

:20 I tornado’d toward Seibrik. Understanding the matchup he shielded long before my tornado reached him. But he dropped his shield and dashed toward me. Why? This is becasue he saw me stop my tornado for a fraction of a second. In experience, he recognized this as me seeing him shield and deciding to retreat. He then dashed after me to punish a landing. But this is where theory ends and MINDGAMES begin. I stopped my tornado on purpose for this very reason. Understanding what your opponent is expecting and doing something that doesn’t necessarily make sense on paper is what happens at the highest level in this game.

:38 I try the exact same tactic with tornado. This time he doesn’t fall for it. He holds his shield then rolls around it. Luckily Tornado is a versatile enough move for me to get away because I recognized that I wasn’t going to hit with it.

:54 I shuttle loop. He dash shields toward me to beat the glide attack and grab me. Instead of glide attacking I cancel it and sidestep, then punish the after lag.
1:22 I shutte loop again. This time Seibrik is wary because I sidestepped last time. But I’m one step ahead, and I expected him to adapt. The logical thing to beat my sidestep would be to dash shield. Even if he wasn’t close enough to me to grab me out of it it’s still the most logical option based on my last choice. So I dashgrab. On paper this doesn’t make sense, which is exactly why it worked.

To make a long story short, these explanations are made to explain why what you should expect over the course of a match should change as your opponent adapts to you. Against people you play against regularly it’s likely you’ll be familiar with things they will or will not try. When watching high level players, you probably notice that the mixup game changes, because after being dash grabbed while dash shielding, a player may dash shield again, but bait it with a sidestep this time. This is how mix ups generally progress, and a players ability to predict and manipulate what his opponent will be expecting gives them control over the match.

There is no formula to making the right choice when this game is played at the highest level. It’s simply about adapting and feeling out your opponent. Sometimes you “just know” when your opponent is going to do something, based on patterns from earlier mix ups. The speed at which you recognize and react to patterns is what determines how lethal you are over the course of a set. Often times you will see a trend where one player win game one of a set in WF/LF, but then gets beaten 3 times in a row afterwards. Or a set where one player wins the first match, then the other player 3 stocks them the next match. This is because the player who “adapted” became used to either his opponents options or patterns. Ganondorf is bad because his moves are generally slower than human reaction, oftentimes, you don’t have to “guess” what he’s going to do. You can watch and because of the nature of his moves, react and punish. That’s adapting to a character’s options. That’s why someone playing a top tier character may lose to a lower tier only once, and after that never lose again. In some matchups, one player dominating the other is as simple as understanding their opponents options. In other cases however, adapting is a matter of getting used to an opponent’s patterns. The speed at which you realize that your opponent repeatedly does Z if put into situation one, will allow you to start capitalizing on it. If your opponent continues doing it you can punish it continually, and expect them not to adapt very quickly. Once they realize that they’re being read they may begin trying another option, and this is where the process starts over. Matches happen so quickly that it’s extremely difficult to consciously think “my opponent did this last time, I should do X and expect Y”, so most of it comes to instinct, which is why some people just seem naturally **** at this game. The speed at which you naturally react is what determines what most people would agree on is overall skill.

Concerning watching matches on youtube and how people use them wrongly for determining matchup ratios.

I’m an avid watcher of brawl matches on youtube. Don’t believe me? Look for my comments on every ****ing video. Before I was able to go to tournaments I didn’t watch videos the right way. I would see someone get ***** by someone I hadn’t heard of and say “wow, he’s getting *****, he must suck”. While browsing the boards, I see even worse things, usually influenced by player bias and holding players on a pedestal. “M2K nearly lost that match, he must’ve been sandbagging”, or “wow Player X really doesn’t know this matchup (due to them regarding player X as amazing and assuming he should win because they read on the boards that the matchup was 60:40”. Understanding a matchup from both sides is extremely important to understanding what’s going on when you’re watching a video. That’s the only way to understand the mix up game thoroughly, and what was going through each players head when they made each decision. Then when someone is losing, you can determine whether it’s because they’re getting outplayed, don’t understand the matchup and the options they have or because their character simply doesn’t have the options to deal with something.

Judging matchup off of videos/single matches as a method is a horrible idea. Less than 10% of what happens at tournaments actually ends up on youtube, and single matches or even sets aren’t representative of matchup ratios. Before you can determine matchup ratios here is what you need to look at.

The Skill of the two players – Unfortunately this is difficult to judge, especially in brawl, where experience, matchups and knowledge actually come before skill in many situations. However if you’re going to use a video in itself to try and representative of a matchup ratio, it is important that first, the players are playing at a high level. Secondly it’s important that the players are close to equal skill level. Judging this can be difficult, as brawl isn’t “balanced” in such a way that evenly skilled players will go even however. But matchup ratios are made under the assumption that players are of roughly even skill level. I like to use my matches with Seibrik as examples because I believe us to be on a very close to even level.

How well each player knows the matchup – How well are both players aware of their options. The problem with this is that it’s generally judged off one’s knowledge of the current metagame, and in many places the metagame is underdeveloped, making it difficult to judge “what one could be doing”. (Notably, you can only learn matchups from playing them. Watching a video can tell you what your options are, but experience breeds muscle memory, and you have to do something first before you’ll remember it forever. This will speed up how quickly you can recognize things in matchups.)
“example” – I lost to Lee Martins Lucario in tournament after beating him in MK dittos. I was extremely unfarmiliar with the matchup and continuously got punished by dairs. It was only after the match that I realized that MK’s Utilt completely beats out dair from the ground. I could have played a much safe game if I had known that in advance and probably would’ve won <3 Lee inb4nojohns.

*a bit from Seibrik*

“Another example you could use is Me vs Esam, DDD vs Pikachu. Before i experimented with the matchup, ddd's were convinced it was a 70/30 or worse, because of the way they were used to playing DDD walking around, going for grabs. When they met someone (pika) that countered DDD specifically with grabs, and didn't get punished BY grabs as hard as DDD is used to, DDD players nationwide called it quits. All because they're general style was shut down, as you said.

After I did my experimented, i found that if DDD stays on platforms, (plays gay) camping bairs and dairs from the very begining of the match, it is now a dead even matchup, as pika doesnt have much to do vs bair, as it beats out his projectiles and all his attacks in general save thunder. So as you also stated regarding stages, the only time this matchup is truely what people believe is when it's played on FD, which when banned, the set as a whole, becomes even.”

How well each player is actually playing the matchup. Oftentimes in friendlies and even in tournament you’ll see people winning/losing matchups. But that isn’t necessarily related to the actual matchup ratio if one player isn’t playing the matchup correctly. This may be due to nervousness/lackof knowledge/personal style.

example: Ally vs. Mikehaze, M2K vs. Diddy. Using your “regular style” in certain matchups won’t necessarily work, even if you know that other things are require to tilt the matchup back in your favor. You may see other players who are generally regarded as “less skilled” doing better in these matchups, because they’re using a more appropriate style.
The stage. This is also important. One match on one stage isn’t necessarily representative if that stage will always be banned against that character, or the stage is neutral or counterpick. Over the course of a set this changes a lot of things, such as the likely hood of one character beating another.

“example” – Metaknight can turn a 60:40 matchup into an 80:20 matchup using a CP stage. M2K vs. Mikehaze on SV in comparison to M2K vs. Mikehaze on brinstar.
http://www.youtube.com/watch?v=mdZH_mLyEs0 – See set for details.

Once you have determined which of these factors is present in a match or tournament set you can determine roughly who has an easier time with a matchup, whether a matchup is possible and at what level (general competitive level or top player level: For example, Kirby vs. MK is impossible at the highest level, as Kirby should be unable to kill MK if he plays the correct defensive or even offensive style. At the highest levels of gameplay now top MK players play a clean and safe enough game to avoid this, and even if they do make mistakes, the amount of punishment they receive in comparison to reward should never be enough to win a match. However MK vs. Kirby at the casual level, where both players leave multiple openings and make “mistakes” you will see that Kirby doesn’t have as much trouble, as Kirby can actually get more reward than MK with single moves.) *back to before parentheses* why the player who lost did lose.

So basically the idea of this thread is to help people understand when something is a relevant statement and when something isn’t as well as helping understand when an option should be dismissed or should be looked into further. I remember when people said Snake’s C4 was useless because humanly anyone should be able to see it coming and react to it. I remember when people would dismiss techniques like reverse grenading and illusion cancelling. When people would watch Japanese players and say “oh that’s just being flashy”. Or when people looked at Ally and said “oh, he’s not that good. Most of that stuff wouldn’t work on good people”. This game is about what people are expecting. When watching a video it’s easy to say “I would’ve expected that” but you didn’t experience the previous mix ups that may have conditioned the strategy to work. It was only after extensively learning MK vs. Snake from both sides that I actually understand what’s going on when I watch M2K vs. Ally. People leave comments on youtube all the time saying “someone should’ve seen something coming because it was obvious” but you would do almost anything if you know your opponent is expecting something else.” The speed at which people are able to react as well as experience determines what gets punished and what doesn’t. People make judgements about when something is worth the risk of trying to punish and when something isn’t. That’s why some matches are closer than others. One player adapts and then the other player adapts back. The degree to which each player adapts determines how much one player wins by.

Anyway I hope that you found this to be an enjoyable read. If it didn’t make sense then I’m sorry as it was a lot to follow. With this I’m hoping to bridge the gap between theory craft and competitive matches, as well as clean up the way that general smash analyzes videos. This game has a lot of untapped and ignored potential and approaching learning the right way is a good direction for everyone.

I’ll do my best to answer questions, critics and comments.

edit:
BTW Seibrik edited this so that you dont have to read my crappy typos and broken thoughts.
 

Kewkky

Uhh... Look at my status.
Premium
Joined
Apr 20, 2008
Messages
8,019
Location
San Diego, CA
Switch FC
SW-7001-5337-8820
First!

Skimmed through, gonna read it thoroughly soon... What I read though makes me want to say... Is that how DDDs actually acted against Pika? *facepalm*
 

rPSIvysaur

[ɑɹsaɪ]
Joined
Jun 7, 2009
Messages
16,415
Good read, convinces me further why MK is too good *cough* too many options *cough*
 

RATED

Smash Lord
Joined
May 14, 2008
Messages
1,627
Location
The Grand Line... PR
I already read it and I have to say this first : *high five*


I am totally in agreement with your essay. also by some points explained in this essay is why I think DDD vs DK is not as **** as people seem to think.
 

Gadiel_VaStar

Smash Champion
Joined
Nov 12, 2009
Messages
2,066
Location
Atlanta, GA
NNID
GadielVaStar
Enjoyable read. I see more clearly how one should think when playing matches. I think I'm a good example when you talked about the many viewers commenting, “Why yo do that!” on youtube. I see Ally always fast-falling a bair after he gets hit. I tell myself, “I would have picked up on that by now,” but because I'm not there actually playing X player, I can't really comment.

So what should we be watching for besides mix-ups, when we watch Brawl vids?
 

OverLade

Smash Hero
Joined
Jun 19, 2006
Messages
8,225
Location
Tampa, FL
Good read, convinces me further why MK is too good *cough* too many options *cough*
Too many options doesn't necessarily make a character good, but the number of options he has makes learning to fight MK hard to learn because the number of things you can be expecting from is diverse. You have to learn what to do in every individual scenario, and many MKs will make different choices of his varied moveset, which is why it's so difficult. This is why repetition against MK is key as well.

However, Falco and Dedede for example, have a few options that are just **** good. They aren't simple characters but what you can expect from either character is limited, meaning you're unlikely to get punished just because something is difficult to predict. You're more likely to lose because you didn't know much about the matchup or got outplayed.
Enjoyable read. I see more clearly how one should think when playing matches. I think I'm a good example when you talked about the many viewers commenting, “Why yo do that!” on youtube. I see Ally always fast-falling a bair after he gets hit. I tell myself, “I would have picked up on that by now,” but because I'm not there actually playing X player, I can't really comment.

So what should we be watching for besides mix-ups, when we watch Brawl vids?
Almost everything in brawl is mix ups. When a matchup can be won without mix ups it's because that matchup is one sided. If I think of anything else I'll add it to the 2nd post.
 

Flayl

Smash Hero
Joined
May 15, 2006
Messages
5,520
Location
Portugal
I'm glad someone wrote up on this. Can you include a segment on human error, as in frame room and imperfect spacing (and how you can directly affect the other player's spacing by reading them/weaving/etc.)
 

rPSIvysaur

[ɑɹsaɪ]
Joined
Jun 7, 2009
Messages
16,415
Too many options doesn't necessarily make a character good, but the number of options he has makes learning to fight MK hard to learn because the number of things you can be expecting from is diverse. You have to learn what to do in every individual scenario, and many MKs will make different choices of his varied moveset, which is why it's so difficult. This is why repetition against MK is key as well.
Yes, I know this. I'm just saying that since MK has nearly an option for everything adapting to his options are much harder than adapting to a character who has a fewer but decent options because he has so many mix-ups.

Edit: Thanks Seibrik for editing the OP :D
 

Player5

Smash Rookie
Joined
Jul 29, 2009
Messages
10
Very interesting read. One thing I would like to add about watching videos, at least from my point of view. I also find it important to consider the physical level (not in-game stage) on which the game is being played and how it effects each player in terms of comfort level, general strategy, etc. I believe you discussed this a little bit in your "MK vs. Kirby" example.

The first game of a friendly set provides a different environment as opposed to the 5th game of a set in the grand finals of a tournament. For example, a person playing the same game, on the same stage, against the same person may play more offensively in one environment or more defensively in another. (Obviously, it goes more in-depth than this, but it provides a simple example)

However, I do suppose this factor could be somewhat minimized with very high amounts of experience at extremely high levels of play. (Which I do not have so I cannot state this as fact) /shrug
 

Scaphist

Smash Apprentice
Joined
Apr 10, 2009
Messages
193
Location
British Columbia
This sums up like 8 different topics that are good for newer players all in one. Everyone both new and experienced should read this.
 

OverLade

Smash Hero
Joined
Jun 19, 2006
Messages
8,225
Location
Tampa, FL
I'm glad someone wrote up on this. Can you include a segment on human error, as in frame room and imperfect spacing (and how you can directly affect the other player's spacing by reading them/weaving/etc.)
There's no reason to really. That kind of frame data would apply to individual scenarios, and every scenario in the game. And most importatly, all humans have different reaction speed.

A player punishes something when the amount of time it takes for their brain to process something and command their fingers (which is sped up by experience, which like I said creates the muscle memory that makes punishment more natural) is shorter than either the amount of time the opponent has before they can react, or before their brain can tell their fingers to. Often people technically have enough time to do something but the reaction it would take is extremely difficult to pull off.
 

Jane

Smash Hero
Joined
Jul 29, 2008
Messages
5,593
Location
Ba Sing Se, EK
amazing read.

im at that point in my smash career where i'm just starting to pick up on mix ups, and the thought process that i should be having during a match. this really reinforced what i've been thinking in my head, and gave me much much more.

thank you.
 

Xiahou Dun

Smash Ace
Joined
Nov 24, 2008
Messages
524
Location
England
Excellent thread. I would suggest you show some examples of Good Theory Craft. Theory Craft gets a lot of bad rap but it can be an important block for getting better or coming up with something new or unique if you go about it in the right way. Every good tactic was Theory once.
 

napZzz

Smash Hero
Joined
Mar 4, 2008
Messages
5,294
Location
cg, MN
very good read. ALot of people dont understand this and love to troll/make dumb arguments all the time through bad theory crafting on matchups and such when they dont have any prior experience.
 

shadyf0o

Smash Journeyman
Joined
Mar 31, 2008
Messages
202
Location
Jersey
Really great read. Experience trumps study in all walks of life. You can't read how to Box and expect to compete against a real Boxer. You can't read about sex and expect to be amazing at it. You can't read about swimming and expect to be an olympic swimmer. Just the same, you can't read about Brawl and expect to win in a tournament without experience.
 

A2ZOMG

Smash Legend
Joined
Oct 13, 2007
Messages
12,542
Location
RPV, California
NNID
A2ZOMG
Switch FC
SW 8400 1713 9427
Frankly, I appreciate the effort you put here.

One of the biggest problems about current matchup discussion is that there's really only a very small percentage of players in the community who are actually good (and even many of those players still don't compete on a comparable skill level to your average professional in Melee), and I really sometimes wonder if the masses forget that as they are explaining options and trying to cite (usually awful) video examples to make themselves look more credible.

Many matchups simply don't get represented correctly in this game due to significant skill discrepancies within the community. Hence why there's a lot of really bad theorycrafting that you described.
 

Dark 3nergy

Smash Hero
Joined
Aug 15, 2008
Messages
6,389
Location
Baltimore, MD
NNID
Gambit.7
3DS FC
4313-0369-9934
Switch FC
SW-5498-4166-5599
Facts before fiction. Theory crafting is such a unhealthy habit to get into...
 

Luigi player

Smash Master
Joined
Jul 29, 2004
Messages
4,106
Location
Austria
Metaknight’s forward tilt comes out in 4 frames. Metaknights forward tilt slightly outranges the first hit of snakes forward tilt and comes out faster. MK’s forward tilt is disjointed, and can be slowed down. This means that it beats Snakes dash attack, dash grab, and sidestep. This means the only thing that should beat it is a perfect powershield of the first hit to a shieldgrab.

Snake’s forward tilt comes out in 5 frames. The second hit of Snakes forward tilt outranges all hits of MK’s forward tilt but takes 8 frames to come out.
The frame data threads say MKs ftilt comes out on frame 3 and Snakes ftilt on frame 4...
 

OverLade

Smash Hero
Joined
Jun 19, 2006
Messages
8,225
Location
Tampa, FL
Excellent thread. I would suggest you show some examples of Good Theory Craft. Theory Craft gets a lot of bad rap but it can be an important block for getting better or coming up with something new or unique if you go about it in the right way. Every good tactic was Theory once.
No actually. Most good tactics people just came up with via understanding of "how to win". The original Top players didn't learn how to use their characters by reading the boards, they learned and invented things by playing in tournaments and with their friends.
Stickie?...
Jama?

The frame data threads say MKs ftilt comes out on frame 3 and Snakes ftilt on frame 4...
This is the point of this thread. No one cares when they come out. For the sake of the matchup, the two moves are roughly evenly matched in range and speed, just snake's move has more lag.

Facts before fiction. Theory crafting is such a unhealthy habit to get into...
Like A2 said theory crafting the wrong way is the problem.

For example, theory craft that "applies" to an actual match is (though not necessarily entirely accurate). I'm going to apply these to what makes some matchups against MK "impossible" at the highest level. The information I provide might be slightly off, but it's not important because it's the "reality" of the theory that's important.
On paper, in MK vs. GaW, when at high percent, Meta knight should be able to avoid game and watches kill moves, and pull ahead simply by forcing game and watch to approach and punishing, while outspacing GaWs predictable kill moves. While it's likely a MK player will use a move that "can" be punished, such as a mispaced shuttle loop,MK can cancel his shuttle loop and fastfall. At the worst case, he will get grabbed upon landing, but the fastfalled cancel is so fast MK should be able to roll away before GaW can even get near him. Even if he is grabbed as long as he techs, and avoids being read from his tech, GaW will not be able to kill him with a throw. GaW's best option to kill MK at high percent is Fair (besides a sidestep/roll read). This will kill a MK with good DI (with realistic depreciation) at 130%ish. Once MK is at high percent and ahead, he can play safe, and avoid even these kind of situations.
http://www.youtube.com/watch?v=gzexdyBAIBQ

Here is Havok vs. UTD Zac. For the most part you can see what I just "crafted" in action. While UTD zac doesn't get ***** first stock, Havok rely's on GaW's inability to kill to pull ahead and take another stock. GaW's options are poor if MK is aware of them. Like I said is possible, Havok makes a mistake, which is the reason it isn't a 3 stock, but you can see what would happen if he didn't make the choice that got him killed.

Also, Havok and UTD zac are generally accepted as top players for their character, and UTD Zac clearly knows the matchup very well, as can be judged from him defeating Tyrant, who didn't know the matchup, but is accepted as a top Meta knight.

This is how someone could come to the conclusion that "on paper" MK vs. GaW is 70:30. I'm not saying it is or it isn't, I'm saying this is the kind of "theory" that should back it up.
 

Xiahou Dun

Smash Ace
Joined
Nov 24, 2008
Messages
524
Location
England
No actually. Most good tactics people just came up with via understanding of "how to win". The original Top players didn't learn how to use their characters by reading the boards, they learned and invented things by playing in tournaments and with their friends.
Jama?
What? Haha. "Theory" and "Read on the boards" are not the same thing. By all good tactics were Theory once I mean that before anyone's ever done anything they have to think of it. To give example, one day plankk thought "Why even fight them why not just camp and win?" that was a Theory he hadn't done it beforehand. And when he did, it worked. Ledge Camping, Bananna Locking ZSS and GaW team bucketing. All of these things were a product of someone thinking "Hey, this could work..." they were Theory. And Theory can lead to you finding something that works. But Theory is often approached wrongly.

When a Matchup is a disadvantage "Theory" can help to alleviate this. The example by Seibrik works fantastically here.

Seibrik said:
Before i experimented with the matchup, ddd's were convinced it was a 70/30 or worse, because of the way they were used to playing DDD walking around, going for grabs. When they met someone (pika) that countered DDD specifically with grabs, and didn't get punished BY grabs as hard as DDD is used to, DDD players nationwide called it quits.

After I did my experimented, i found that if DDD stays on platforms, (plays gay) camping bairs and dairs from the very begining of the match, it is now a dead even matchup, as pika doesnt have much to do vs bair, as it beats out his projectiles and all his attacks in general save thunder. So as you also stated regarding stages, the only time this matchup is truely what people believe is when it's played on FD, which when banned, the set as a whole, becomes even.”
Seibrik saw that DDD's were having trouble with Pikachu and analysed "Why?" he saw that the Grab oriented side of DDD suffers against Pikachu. But then he started to use Theory, he believed that DDD's were losing because they were playing against Pikachu like they do against everyone else and decided to see if playing differently and forgoing grabs for a different playstyle would help. What is this if not the correct application of Theory Craft?
 

-Jumpman-

Smash Champion
Joined
Nov 3, 2007
Messages
2,854
Location
Netherlands
*Looks at example*

Seibrik should've shielded the first time. Also, if Seibrik SHd the second time to bair you out of your spot dodge, you would have been punished.

Edit: trying to punish the nado if you can just shield it is just stupid.
 

OverLade

Smash Hero
Joined
Jun 19, 2006
Messages
8,225
Location
Tampa, FL
What? Haha. "Theory" and "Read on the boards" are not the same thing. By all good tactics were Theory once I mean that before anyone's ever done anything they have to think of it. To give example, one day plankk thought "Why even fight them why not just camp and win?" that was a Theory he hadn't done it beforehand. And when he did, it worked. Ledge Camping, Bananna Locking ZSS and GaW team bucketing. All of these things were a product of someone thinking "Hey, this could work..." they were Theory. And Theory can lead to you finding something that works. But Theory is often approached wrongly.
In a match, the amount of time you have to spontaneously think of these things is a very small window. Thinking of something in a split second isn't theory, though you could still call it that if you really want. Theorycraft is what this thread is about anyway, not "theory" as you're referring to it as.

When it's last stock people don't theorize that playing more campy will help them win, they don't approach because they're afraid. That's not the same thing as theory. People may "theorize" these things, and some strategies were legitimately discovered from theory, but for the most part as this game developed, it's people who were actually playing the game realized them. Most people don't decide to camp in some matchups. You just realize you're doing it because you don't see any options to approach, that's not the same thing as theory.
When a Matchup is a disadvantage "Theory" can help to alleviate this. The example by Seibrik works fantastically here.

Seibrik saw that DDD's were having trouble with Pikachu and analysed "Why?" he saw that the Grab oriented side of DDD suffers against Pikachu. But then he started to use Theory, he believed that DDD's were losing because they were playing against Pikachu like they do against everyone else and decided to see if playing differently and forgoing grabs for a different playstyle would help. What is this if not the correct application of Theory Craft?
Camping on a platform to avoid a grab wasn't discovered by theory. It was discovered by logical realization, which isn't theory. Standing on the platform makes perfect sense to anyone who plays this game and understands a grab can mean death. If nobody posted it on the boards, tons of people would still be doing it. It would pass by word of mouth in the tournament scene. But it's not theory.
 

OverLade

Smash Hero
Joined
Jun 19, 2006
Messages
8,225
Location
Tampa, FL
*Looks at example*

Seibrik should've shielded the first time. Also, if Seibrik SHd the second time to bair you out of your spot dodge, you would have been punished.

Edit: trying to punish the nado if you can just shield it is just stupid.
Lol this is the problem right here. This is exactly what I was talking about.

Why didn't Seibrik shield? Ask Seibrik. Seibrik didn't shield because he didn't expect me to sidestep so quickly. This game is rock paper scissors. He had to guess what I was going to do. He couldn't go to dash grab but see the frames of my sidestep come out and wait. Nobody's brain works that fast.

Concerning the SHbair or w/e, of course I would've been punished, but I made that choice because I knew he wasn't going to. That's exactly what this entire thread is explaining. I didn't get lucky, I predicted correctly, and was rewarded for it. Dashing and SHing followed by bairing against MK is a terrible option in general. It would be too risky, because what if I DIDN'T do what I chose to do. He can't read my mind, he has to either guess or react, and MK is fast to the point where you have to predict to punish.
 

Veril

Frame Savant
Joined
Jun 20, 2008
Messages
3,062
Location
Kent Lakes, New York
This essay is necessary reading for 99.9% of people who think they understand the game numerically... I hate the term theory-craft personally, and I was certainly guilty of sloppy theorycrafting when I first joined smashboards.

That said, my experience with the B+ project runs contrary to what you are saying. Having an encyclopedic knowledge of the game has been tremendously useful for me. Good, theory, which is really more like "law" than "theory" requires being able manipulate and interpret data and deal with all the confounding variables that will occur in a match. I've seen only a tiny handful of people who are capable of this. Academic training really is needed for analyzing the mechanics vs. theorycrafting.

I mean, I've worked as a cognitive science research assistant, and I got really really really good at crunching numbers. I also have enough math background for a minor in it (but... nursing major so it doesn't matter lol). Bowyer's another example, with his game-theory background and supremely gay MK. "shrug" I guess I'm saying that almost nobody is capable of ugh... theorycrafting... at the highest level, much like very few people are capable of "playing" at the highest level at all. How many people are at the level of Lain or M2K? A handful. Its the same with people who work with the game's mechanics.
 

OverLade

Smash Hero
Joined
Jun 19, 2006
Messages
8,225
Location
Tampa, FL
This essay is necessary reading for 99.9% of people who think they understand the game numerically... I hate the term theory-craft personally, and I was certainly guilty of sloppy theorycrafting when I first joined smashboards.
I was also extremely guilty of theory crafting until after January 31st. It was my offline 2nd tournament. I had since only been playing wifi, and my perception of the game was heavily skewed because of how bad my internet connection was. I was always playing in red connections, which meant I had to predict almost EVERYTHING or taking the safe option (camping). It was in January when I realized that I could read and punish sidesteps almost on reaction of seeing them, and that I could force people to sidestep by rolling behind them. Since then I began playing with forcing reads and being unpredictable in mixups.

I started playing a style that would show people what I'm doing when I read them. I would wait for people to airdodge toward me, but instead of grabbing them, I would wait. They would try sidestepping on reflex. I would wait for it, and then grab them. This was to teach people how far ahead I was thinking, and how I was relying on experience and logic to get the read. It doesn't make any sense to do anything but sidestep, as if I mistimed my grab you could possibly get away with it.

But now that I'm waiting for it, I get my opponent thinking (wait, what if I tried grabbing him, because I knew he was going to wait for my sidestep?). Suddenly it's a mixup, even though it's not logical.

What you should expect is based on conditioning and learning. You guess what your opponent is going to do based off of two things, what you've experienced in the matchup before, and what you've experienced from your opponent before.
 
Top Bottom