Patroklos Pre-Release Discussion

Status
Not open for further replies.
Signia - No offense, but you shouldn't be teaching anyone, anything about strategy in this game. I was just talking with NFK and when the subject came up he had one major criticism of your play. You choose good options, but you are predictable about what you do... Which is exactly what I said was going to happen if you overthink things.

I have been there before, but what you will come to realize through trail and error is that any option no matter how good it looks on paper can be exploited if it can be predicted. If you really want to research and learn how to improve you need to work on pattern analysis and identification.

That is also my argument for using varied options outside of guaranteed damage situations. You want to run as many different gambits as you can to make your opponents have to use as many different ways to get out as possible. Not only does it stress them out, it makes it harder for them to keep track of your mixups and outguess you. If you have one "better" mixup on paper that you keep using you will have a low chance of getting "randomed out". But because it's so straightforward you have any kind of habits, they will be easier to identify simply because there are less variables to look at.

In short keep it simple and don't overthink things or figure out how to use as many moves as you can just for the sake of creating as many different situations as you can within the game. But if you keep it simple you better focus on being safe or you will get chewed up hard for predictability.
 
Signia - No offense, but you shouldn't be teaching anyone, anything about strategy in this game. I was just talking with NFK and when the subject came up he had one major criticism of your play. You choose good options, but you are predictable about what you do... Which is exactly what I said was going to happen if you overthink things.

I have been there before, but what you will come to realize through trail and error is that any option no matter how good it looks on paper can be exploited if it can be predicted. If you really want to research and learn how to improve you need to work on pattern analysis and identification.

You can't really teach unpredictability, so I'm not really worried about that flaw being passed down to other players. Everyone knows how to play rock-paper-scissors, I'd just be teaching them how to throw rock or when scissors is stronger or when paper isn't a threat -- aspects of the game, so they get up to speed quicker so their ability to mind read matters more.

I'm curious as to what you mean by overthinking and how it causes one to be unpredictable. Thinking more than most certainly doesn't cause unpredictability, if you think so why don't you ask NFK about babalook? You probably mean I'm thinking about the wrong things. And I actually don't think about nash equilibriums while I play, that's too much to think about while still thinking about everything else. I'd just think about common situations beforehand, consider the implications of the math, and apply it to my play. I don't even do the math actually, it's often easy to tell the result intuitively, and since you can't really apply it unless you're a computer, all it really tells you is where the metagame will eventually head, giving you a good idea of what standard play will look like. It still helps if you can understand it, it just doesn't help like you think we think it does.

On the other hand, my way of thinking about spacing (yes, I do consider all those other ranges you said I missed, I didn't list all of them because I don't know enough about Pat yet) is not too mentally taxing during play, and it's something every player should consider when thinking about matchups and how to space properly in general. I'm pretty sure your only objection to it was my lack of knowledge of Pat, though.

That is also my argument for using varied options outside of guaranteed damage situations. You want to run as many different gambits as you can to make your opponents have to use as many different ways to get out as possible. Not only does it stress them out, it makes it harder for them to keep track of your mixups and outguess you. If you have one "better" mixup on paper that you keep using you will have a low chance of getting "randomed out". But because it's so straightforward you have any kind of habits, they will be easier to identify simply because there are less variables to look at.

In short keep it simple and don't overthink things or figure out how to use as many moves as you can just for the sake of creating as many different situations as you can within the game. But if you keep it simple you better focus on being safe or you will get chewed up hard for predictability.
This is good stuff, sounds like your own theory fighting
 
Thinking more then most causes predictability. You feel certain moves are "better" than others in a given context, but the more you trim your options to optimize things the easier you are to read as a player. One thing you have to accept is that outside of a few select instances anything that can be predicted can be punished. So thinking any situation has a "best option" is the dumbest thing you can. The smart player will be able to use a counter and make the "best option" just another move you're doing predictably because you see the stars align... So once they realize the trigger they will kill you for thinking that some options are better than others.

Does this mean you can't take into consideration their counter attempts and create a gambit designed to exploit them? Hell no! But you don't know anywhere near enough to begin that process, so why bother?
This is good stuff, sounds like your own theory fighting
It is, I never said theory fighting is a bad thing. I just said doing it without getting your hands on the game is bad. It's going to draw many more bad conclusions then good ones. Why would you want to try and learn a bunch of stuff you will need to unlearn/relearn later? Do you really think that you might think of something that will otherwise go undiscovered in the first few months of play by doing all that nonsense? I would say it's doing way more harm than good, so just try and be patient.

Also try and take a moment to absorb this little nugget of information and answer honestly. I called your biggest player weakness by seeing your theory fighter and not your gameplay... What makes you think you're doing it right?
 
I think we're actually in agreement in most of the stuff we're discussing but you haven't realized it. So all this was a way of saying I don't know enough about the new game to draw any real conclusions. Well, I agree, I can't do anything like declare matchup ratios (X vs. Y is 70:30, etc).

However, I did want to take a close look at the character to see if any twitter feedback is required. He seemed really weak when I first saw him, but right now, I'm happy with what I see. If he didn't have good lows or that tracking mid, I'd be worried about his tournament viability. The strength of a character is greatly determined by how much space they cover, and how dangerous that space is. Immediately I noticed that Pat does not control very much space, he has short range and is not super fast to compensate for this. Sounds like an awful character until you look at how good his tools are.

Sure, I can't know everything about the character, but some discussion is better than nothing.

What makes you think you're doing it right?
The level of skill I've achieved despite having next to no offline experience, and the way my theories have been demonstrated to be successful as I learned the game, over and over again. I've hit a wall for lacking that experience, but I've gotten really far without it. It won't be the same in SCV, though.

Think you're so clever calling out my weakness? Predictability? There's not even any certainty that that's indeed my biggest weakness, I'm almost tempted to call BS on NFK's assessment. That's what you tell someone when they lose and you can't tell what they did wrong other than the fact that they guessed wrong.

Allow me to call out your weakness, based on what you've said, just a guess: you don't make choices for any particular reason so you end up wasting opportunities.
 
Allow me to call out your weakness, based on what you've said, just a guess: you don't make choices for any particular reason so you end up wasting opportunities.
Can't exactly speak for him, but I think I can feel where he's going with this. "The best option" isn't what works best against a character. "The best option" is what works best against a player. His choices are probably based in reason like your own, but are skewed towards what will work based on what has worked well before and what the opponent is likely to do.
 
Can't exactly speak for him, but I think I can feel where he's going with this. "The best option" isn't what works best against a character. "The best option" is what works best against a player. His choices are probably based in reason like your own, but are skewed towards what will work based on what has worked well before and what the opponent is likely to do.

Edit: nvm I misread, that isn't obvious

I agree with this though.
 
@Signia: I didn't say anything to Bibulus that I didn't tell you in person at DEV.

aka your choice of options and game knowledge are higher than everybody i've met except babalook, however they cause you to guess wrong because you know how good they are and you can't stop yourself from doing them, due to lack of offline experience.

I think this game is extremely guess heavy, no matter how you spin it. Personally I have the urge to create constants in my head as to what the best option is for certain situations, it's how mathematical people work. Unfortunately it doesn't help at all in SC beyond exploring what the most rewarding way to guess correctly is, or ensuring you take minimal damage on an incorrect guess. You probably know all this by now.

PS: I really don't know why you lose other than the fact that you guess wrong. I don't think anybody does. Maybe it's a sign that your basics are extremely solid and you need to work on that rock/paper/scissors.
 
Allow me to call out your weakness, based on what you've said, just a guess: you don't make choices for any particular reason so you end up wasting opportunities.
LOL making choices for no particular reason is impossible. Everything that happens within the game is a choice, whether you choose to look at it that way or not is your own choice. If you really want to talk about applicable theory try this one on for size.

There is no random factor beyond unknown variables.

Nobody can do something totally at random, there are patterns and decisions taking place you may not see and/or understand. If you choose to ignore them you are choosing to ignore mindgames and someone who acknowledges them to make proper concessions will wield a significant advantage.

If you really think you can truly be random, you are deluding yourself. When looking at statistical trends of really random things you need to have an adequate sample size. Basically the more variables and results to the equation, the bigger the size of results must be before you really see the trend in motion. Poker for example will not normalize where you can expect proper gameplay to mirror results until you have seen millions of hands.

Now while this doesn't apply to 50/50 mixups like flipping a coin, this will help you see things a little better. It's a random number generator set to give 100 results either 1 or 2 at random. When I tried this the first time there were 8 more 1's than 2's. If that isn't an accurate sample size to reflect a semblance of 50/50 dispersion, you can conclude the average person can't keep track of the sample size involved to see random come to fruition. So what do you expect to do in order to be really random?

In reality the reasoning for your choices is what defines your play. It's the driver that takes the theory fighting engine you designed out on the race track to wage war against other drivers. The car is only so good as the driver and if someone in a pinto makes his decisions for all the right reasons he can beat the piss out of a race car.

The level of skill I've achieved despite having next to no offline experience, and the way my theories have been demonstrated to be successful as I learned the game, over and over again. I've hit a wall for lacking that experience, but I've gotten really far without it. It won't be the same in SCV, though.
That is like saying that you don't want to erase a bad habit you got while you used training wheels on a bike. But leaning to the left was never hurting me before. I will probably learn something new after I get more experience, because I know how to move the pedals properly already and training wheels proved it.

How do you quantify your level of skill? Where and/or what comprises the control group of your experiment? Also if you lack experience, how do you prove your theories outside of using them in gameplay?

I would say your only true skill would be to determining the possible outcomes of a given situation. I think I could also say you have no way to gauge the difference in your growth as a player relative to the field. Probably the only thing you know is there are lots of people "better" than you with no way to gauge the difference or way to know how you can bridge that gap.
 
LOL making choices for no particular reason is impossible. Everything that happens within the game is a choice, whether you choose to look at it that way or not is your own choice.

I meant a reason relating to what your opponent will do next. But w/e, it was just a wild guess, I don't care much to discuss it anymore.

If you really want to talk about applicable theory try this one on for size.

There is no random factor beyond unknown variables.

Nobody can do something totally at random, there are patterns and decisions taking place you may not see and/or understand. If you choose to ignore them you are choosing to ignore mindgames and someone who acknowledges them to make proper concessions will wield a significant advantage.

If you really think you can truly be random, you are deluding yourself. When looking at statistical trends of really random things you need to have an adequate sample size. Basically the more variables and results to the equation, the bigger the size of results must be before you really see the trend in motion. Poker for example will not normalize where you can expect proper gameplay to mirror results until you have seen millions of hands.

Now while this doesn't apply to 50/50 mixups like flipping a coin, this will help you see things a little better. It's a random number generator set to give 100 results either 1 or 2 at random. When I tried this the first time there were 8 more 1's than 2's. If that isn't an accurate sample size to reflect a semblance of 50/50 dispersion, you can conclude the average person can't keep track of the sample size involved to see random come to fruition. So what do you expect to do in order to be really random?

Please read my posts a little bit more closely so you can save yourself some time. I constantly reiterate that it's impossible to implement a mixed strategy since it's impossible to be random. A mixed strategy is when you randomly vary your options but in frequencies that follow specific percentages. I already know and agree with this, and this isn't at odds with anything I've said.

In reality the reasoning for your choices is what defines your play. It's the driver that takes the theory fighting engine you designed out on the race track to wage war against other drivers. The car is only so good as the driver and if someone in a pinto makes his decisions for all the right reasons he can beat the piss out of a race car.

This doesn't contradict anything I've said. Naturally, I agree.

That is like saying that you don't want to erase a bad habit you got while you used training wheels on a bike. But leaning to the left was never hurting me before. I will probably learn something new after I get more experience, because I know how to move the pedals properly already and training wheels proved it.

Not really, since I want to erase all bad habits.

How do you quantify your level of skill? Where and/or what comprises the control group of your experiment? Also if you lack experience, how do you prove your theories outside of using them in gameplay?
I did use them in gameplay, in casuals against NFK and sometimes some other players. Normally you'd expect I'd get my ass kicked by someone at NFK's level but that wasn't the case; using my main I maintained about a 40:60 win ratio. There's also my tournament placing and ability to take games off of the top 8 at DEV (at the very least). I have only attended a total of 13-14 offline sessions over 3 years. In between that time I picked up VF5, theoried it up, went to one offline session and beat down some players that had been playing much longer than I had. I picked up Melty Blood rather quickly, too. My "theory fighting" is very efficient when consider how often I play, this is why I feel I definitely should be the one to teach new players -- they'll improve very fast under my teachings.
 
VF is such a good game to apply theory fighter. There are such defined risks for everything it all adds up nice and neat. I have never seen a VF player I couldn't take a game off back when I played Evo... Well I never took a game off Homestay, but I also never left changed stages off the arena. Akira is soooo brutal in that small cage it just wasn't happening. Throw clashing just made me not like the game system.

Some things you just need to learn with trial and error. I think the closest I can really conceptualize your lack of vision. You seem to just be looking to streamline your game and get the most from every action you take. What you should be doing is trying to find ways to get the opponent to do what will ultimately hurt him the most.

The whole streamlining your options will look to minimize risk in a lot of areas, but that will also give your opponent few weaknesses to try and exploit. Against tough opponents there is more value in a few really unsafe moves then a hundred safe ones in many contexts. Take Sophie B+K,B for example, you take some early risks and throw out that followup as much as you can get away with. Now when it comes down to the wire, you can B+K and then resume offense. You have shown you are willing to take that risk so defending feels like the most logical response in the book, but you took risks when you could afford to and capitalized on the image it left in your opponents mind later on.

You have to give opponents the hope that they can hit you with certain things, otherwise they will just stop trying them and your streamlining just made the matchup more lame. Sometimes that means what seems like the worst option can sometimes be the best. You just have to create a gambit designed to make that killer instinct to pounce on unsafe options backfire.

I used to do that with Sophie A+B a lot too. Just throw it out wantonly at disadvantage until the dude started blocking to spite his frame advantage or using something to punish it in a predictable fashion. Then I would go out of my way to use unsafe tools to stuff that shit too, like 236A instead of 2A or low guard into whiff punish. There are plenty of safe ones, but when you consider the context it's just taking full advantage of potential damage until you force the opponent to abandon their frame advantage. Who is going to default to guard after they get handed frame advantage? Not many people and those who will are easy to identify early... but it will probably cost a round at the rate I was handing out free damage. Once you get him to give up his aggression at advantage you break out the poke tools and go to work with side grabs.
 
I stole the concept straight from poker theory, but the concept of imagery fits the context perfectly. Every action you make helps to shape the opponents perception. It was one of the big breakthrough discoveries I made towards the end of SC2. Early on I tried to just maximize my efficiency, while that was actually winning me some tournaments it was actually because the people I was playing against weren't content just being lame.

This is something I didn't really pick up on at the beginning because I just assumed that when players were "better" than you they would beat you up. I was winning vs the other players, but the top dogs would always gun me down hard. There was actually a point where I was playing astaroth based entirely on low/mid/throw because I felt like highs gave no mixup value and I wanted to play a basic style with as much mixup value as possible.

The breakthrough happened when I started to play a more RO oriented nightmare and I had to find a way to get people to do what seemed like the dumbest thing possible. Like there were people who would just block until they had frame advantage and then look to sidestep away from RO pressure at frame advantage gained from blocking.

There are 2 viable ways to go about cracking that defense in SC2. The first is to choose options to abuse them for being overly defensive. Which to nightmare is essentially low poking you to death or grabbing, also keep in mind grabbing resets space and gives time to move before you can touch them, so it effectively resets the RO pressure and you have to get back into that scary I can kill you with 1 hit ring position.

The other option is to be risky about highs... and I mean REALLY risky. I started doing agA x2, 6A into sidestep to contain the escape as a baseline vs Ivy who can RO you for ducking any of them. The thing is if you get hit by any of that you are done. Once I got them to start ducking then it was time to start using more safe stuff overall at that spacing, but making sure to throw those 33_99A's where they were ducking. That would RO you for ducking and trying to fish for a free round because it just might work or it worked before.

So basically you throw away a free round to create the image that you can get away with that. So when you get that situation again in a lead they will look for that easy round and end up in the river. Really good if you can lose a round like that with a health disadvantage to plant the seed.
 

Although old, and not a very good video in terms of skill, there is some interesting things we can pull from this video...

Pat's 3B is faster than it appears and it has TC properties. And there's another TC move the uppercut with sword and shield. His AA also seems to have some good tracking properties, despite its ridiculously short range.

Another point, is the weapon trail or particle effect on Pat's sword is red, while Pyrrha's is blue? Hints of something more there? Maybe...
 
That's interesting, when I tried my hand at poker (I read some books and was serious about it), I tried to think of analogous strategies from FGs that would help my poker game. In poker, I reasoned, instead of controlling space, you control possibility space in order to limit the options of you opponent. An example of this in poker would be if you raised 3x the blind before the flop in good position, you limit possible hands that are calling against you to only upper groups unless they want to take a big risk. Then I realized you can also do this in FGs.

Still, throwing rounds and being really unsafe to establish that control is little too scary for me. I'd rather have limited control in exchange for safety. Though, I've noticed players tend not to punish unsafe moves the first time they see them in a match unless they're very commonly used or you have lots of time to react, so maybe I can just throw out unsafe options once and have my cake and eat it too.
 
That is one of the common traps in poker logic depending on how you apply it. That is like a yomi layer 2 move, so if you apply it to everyone idiots will run you over. One of the things about poker strategy is you can't let people making huge mistakes get huge payoffs. If a 10 bad call leads to a +200 pot because he went all in and you don't expect him to be holding dogshit because he had a gut feeling dogshit was going to hit the flop.

The concept of yomi layers is seldom discussed on this forum so I'll go over the basic gist of it. It's a way to try and quantify player level based on the metagame. The general idea is to establish an intelligent progression of metagame options.

Layer 1 is where you establish base line options.
Layer 2 is where you adapt to the opponents counter of baseline options
Layer 3 is where mixups happen after the opponent adapts to counter layer 2

The key to that you establish a strong base line and don't give people credit until they do something to earn it. If they show what it takes to get past level 1 you turn it up a notch. Then if they bring it for level 2 it's probably time to turn it up to full volume and throw everything at them.

When you look at the strategy it's a very basic gambit designed to give the highest chance to blow up players who won't adapt. While at the same time giving you an easy route to establish what players high level mixups will work against quickly and reliably... and it works really well because you don't need an elaborate strategy to have a winning strategy.

Also in SC2 you had to be that serious, if you weren't willing to die by the sword you had to turtle. Throwing away rounds or showing the willingness to do so was essential because defense was too strong. You can step guard and with relative ease, eliminate any possibility of RO. What kind of reward but winning a round is going to get someone like RTD to duck when it could kill them?
 
I'm aware of the yomi layers, but I see what you mean about creating a baseline. It's almost like a local metagame you're establishing, if that make sense, where you deviate from it when you need to.

Not only can you profile what layer your opponent is on, you can also profile how they adapt. Since we're engaging in the discussion to end all discussions, I thought I'd share:

Type 1 (stubborn) - they stubbornly don't change their mixup option, and instead hope you change.
Type 2 (counteracting) - they see what you did last and counter that, doing the same thing until it stops working.
Type 3 (pre-adapting) - they preemptively adapt their mixups, changing before they stop working, guessing when you'll change it up.

It's easy to see these behaviors intuitively, but I've also noticed these behave in a RPS fashion, where 3 > 2 > 1 > 3. Pre-adapters will change at the moment when counteracters will adapt, they'd be a step ahead. Counteracters will simply counter the stubborn player's straightforward strategy, punishing their slow adaptation. Stubborn players will throw off pre-adapters by not changing even though while they're getting beat down for their unpredictability.

Of course, players do switch up even these behaviors, making things a little more complicated. This theory hasn't really worked too well, tbh, I've found strong players to be impossible to profile.
 
You're missing the point. You can't categorize a player with good adaptation into a little box, but if you realize they have good adaptation you can start high level mixups. Where if you used those mixups on people who weren't paying enough attention they could 1 trick pony you to death because you expected them to adapt.

The idea is to really define 3 player skill levels.
1 - I don't really adapt and I will play the same regardless of what you do. Learn what works and do it!

2 - I learn slowly. Many mid level players will not be able to make transitions mid round, so you can exploit this on layer 2. You just need a series of mixups where upon the opponent adapting to one, you can just start using a different mixup with a different answer and watch them learn how to deal with it all over again.

3 - These are the guys who are going to adapt on the fly to multiple levels of mixups. Any systematic approach becomes worthless after it's been shown and you need to use options with real mixups. You can now run a complex gambit, because the opponent will make adaptations to counter you at a fast enough rate.

All 3 levels of player require a different set of tactics to exploit their tendencies, the only real purpose to the strategy is to define them.
 
You keep making the same basic mistake in concept so I want to define something for you. There is a distinct difference between strategies and tactics. A strategy is an idea and a tactic is a tool. So if you had the idea of using lows until your opponent started to block them and then doing mixups, that is a strategy. The moves that you use to manifest them represent your tactics.
 
Status
Not open for further replies.
Back
Top Bottom