How does one get good?

I'm going to point out that if it seems that though I usually browse /sci/, I am posting here because I honest to god do not know where else to ask this. It might actually require me doing some serious research which is extremely scary, but I feel it might have ramifications far beyond being good at vidya. Actually, I myself only play these things casually. Something I have come to make clean distinction is between 'intelligence' and 'smartness'. Colloquially, I do not think such cutting of hairs is necessary, but this sort of splitting of hairs is far more important for this particular discourse. By intelligence I mean domain general cognitive capacities. A intuitive idea of this can be 'brain processing power'. There are people that simply have better Working Memory, Executive Functioning, etc, and these allow information to be processed more efficiently. By smartness, I mean more so metacognitive/elaborative strategies and schema that lead one to approach general questions more efficiently. I guess you could say it is 'wisdom', though more specific.

This is a question I have been asking myself about video games for a while now. I will put it in two parts.

(1) Generally, how do metas form? When we see video games like starcraft, you observe a complex manifold of rules and interactions, that it is difficult to really explain why certain strategies evolve over others. This is a bit more nuanced than it seems in the surface. For instance, an immediate theory for why this is happening is due to natural selection, namely that people randomly try out some stupid shit and eventually they discover stuff that works. This however doesn't seem acceptable over further inspection. For instance, it is clear that you have instances of certain players trying a vast number of different strategies, but of them still seem to improve faster than others.

Attached: tenor (1).png (540x470, 533K)

Other urls found in this thread:

journals.plos.org/plosone/article?id=10.1371/journal.pone.0186621
journals.plos.org/plosone/article?id=10.1371/journal.pone.0123259
steamcommunity.com/id/tomoko/
twitter.com/NSFWRedditImage

This show didnt deserve to be as fun as it was. It was pure otaku pandering but had this weird charm to it, like it wasnt afraid to admit it. Reminds me of Monogatari

post more Jibril

Meta literally just means using info from outside the game to change how you play.
As to how you get good? It's literally just tons of grind (practice), the ability to learn and a competitive mindset.

Also being able to live without having a job helps, I guess.

>473667723

Attached: 50883686_p0.jpg (4960x3897, 1.21M)

... If trial and error and time played was the best predictor of this, then we wouldn't be able to explain players in say league of legends who have millions of mastery points, but still remain in bronze. There also seem to be individual differences that determine how effectively different players might learn and master the fundamentals of their respective games.

(2) This then brings us to the second part of the problem. Intelligence or smartness? I haven't done a full doxology of the issue, but I think an example of someone who is much more a proponent of smartness would be a coach like LS. In his video monologues he stresses a methodology to improving at any given game. More specifically, strategic invariants that aid you when you move from game to game (for instance learning the fundamentals first so the rest slows down). Of course, while these strategies are indeed helpful, it doesn't answer the question of where exactly did these fundamentals come out in the first place? Is there a similar wisdom/smartness involved?

On the other side, there is support for the intelligence supposition as well. I recall an article like this exploring the correlations between League of Legends and fluid intelligence:

journals.plos.org/plosone/article?id=10.1371/journal.pone.0186621

Which seems to point to the fact that Fluid Intelligence is indeed very much involved in this process. At the same time I remember reading this article that suggests there is some wisdom that could be utilized as well:

journals.plos.org/plosone/article?id=10.1371/journal.pone.0123259

Of course the above article only concerns with training gains as opposed to the discovery of fundamentals within a game. Still, it does give me hope that there may be ways to be smart about approaching any given video game. Ofc, there still comes the added question of, if we do find such strategies, why do they work?

Attached: dd7eaca8a708dd61e22ed0b82918144c.jpg (800x571, 187K)

People plateau at some point and most of them are just not interested in putting in the time and work to improved their personal skill and knowledge.

Dota 2 for example extensive information on the internet about how to play the game and gives you every tool you could possibly need (you can watch replays from every players perspective at anywhere between half to 8 times speed, complete logs of every single damage instance, spell duration etc) but barely anyone will get go back and try to analyze why the fuck they kept dying in the same spot to the same player over and over.

Instead they just shrug it off, blame team and move on the next match.
This is how people with a billion matches and 2000 MMR happens.

The last note I'd like to make is if we do find such general strategies, wouldn't it be fun to make a compilation into a book of some sorts?

>Meta literally just means using info from outside the game to change how you play.
That is not really what I mean by meta. To me a meta is a collection of game-specific strategies repeated by a group of players within some community. I am interested about how such information structures form. Actually now that I think of it, my distinction between intelligence and smartness is a rather misleading dichotomy. The reason being we can also look at the question from the view of social structures.

> It's literally just tons of grind (practice)
I wasn't able to further elaborate why this doesn't seem the case completely, but I went further here . I ask these questions because I have observed cases in which people might grind a shit ton, but not improve a lot for some reason.

>the ability to learn
Which then brings me to the idea of intelligence and smartness. I have a hunch that intelligence only takes you so far if you don't have a proper methodology to augment it. It's also interesting to point out how intelligence helps as well. For example, what is Fluid Intelligence that it would help in playing video games. There is the vague answer that it is the measure of how to solve complex problems, but it is more of a tautological restatement of the definition than actually investigating the cognitive processes and the prerequisites of learning.

>and a competitive mindset.
Oh thank you, I actually forgot of that. I think being competitive also probably makes you approach things different than if you just want to play something casually (I guess for one it biases you into being more of a maximizer)

She is best girl

ok

Attached: 248121.jpg (225x350, 46K)

Yeah, I agree. That's probably one of the biggest reasons why. Would you say that simply making hypotheses about why a certain thing happened in game is enough? Also I think this goes back to what I said about why it might have ramifications for other stuff, because there is a certain 'science' (that makes me sound like a dork lol) to learning how to get good at certain games

You're using too many uncommon words for a thread on Yea Forums, dude, should have lurked more. Here's how metas form.
>game comes out
>players start playing
>small portion of players who are obsessed with developing strategies discover a strategy and start dominating competition
>competitive players get butthurt and start copying the strategy for their own use (this is dopaminergically reinforced by the game itself).
>everyone else starts losing regularly to the new strategy, so everyone starts doing it
>very few strategies are perfect, so eventually the first group of players develops a counter strategy
>the cycle repeats multiple times and the meta evolves as a result

Here's why people play a lot but never try to get good:
>they don't want to get good, they just want to win
It's as simple as that.

>Intelligence or smartness?
There's no good definition of intelligence, so you're starting your investigation poorly. The answer to your question is "smartness" as you've defined it, that's why people generally are better than others. Not every intelligent person is interested in winning, while almost everyone who uses "smartness" strategies is interested in winning.

>That is not really what I mean by meta. To me a meta is a collection of game-specific strategies repeated by a group of players within some community. I am interested about how such information structures form.
This is what I meant, thought. Meta is just something that forms from people trying to exploit eachother's knowledge/information outside of the game.
If you played RPS and you knew that meta is 65% of people always pick rock for some reason then the way to exploit and beat the meta would be picking paper until enough players catch up to the fact that rock is losing games to paper. (this is obviously simplified but I hope you understand what I mean)


As for intelligence/smartness:
I've played games both with people better than me and people much worse than me.
The ones that had higher skill than me always seemed way more frustrated by bad performance and used that to motivate themselves to improve at the game.
The worse ones, I've asked them why they aren't trying to get better despite me trying to teach and help them for months. When asked, they usually told me that they have other priorities in life than excelling at a video game and just play it to relax or kill some time.
One of them literally said he cares at being good at his web dev job and doesn't want to waste time on games.

It's a multitude of factors, one part psychology, one part mathematics, one part strategy and tactics and one part skill and ability to understand these things and execute them. If you don't even have a surface level knowledge of any of these things in video games you can never get good, you'd actually be surprised at how many people don't actually know or understand the concept and prime use of strategy at all in daily life despite actually employing it from time to time. You have to ask yourself key questions as you play games which is the basis of any form of learning. Who/What/Where/When/Why/How/Should/If/Could/Will/Are etc.

You bring up Starcraft but that's just easy to explain if we break down the actions with questions that would be constantly asked in real time to a player either consciously or subconsciously. For example:
Who is my opponent and do I know how they typically play?
What are they building?
Where is my opponent in relation to my start location?
Why did they go for vespene gas so early?
Should I attack them early or focus on securing a foothold in my area?
How should I approach this matchup?
Could I end the game before they establish an economy?

All of these questions can translate over very easily to so many other games like Racing, Fighting, Puzzle, Card battling, Turn Based Strategy, Flight Sim etc.

Oh yeah, something I'd like to add because I think my question seems a bit stupider than it is already. I mentioned league and starcraft, and what you need to imagine is that you are going into the games, and there is no internet. The game teaches you all the controls, how to build different buildings, etc. How do you improve from there? Hell, how do you even know something you are doing is good when you have a whole 30+ minute game in which several variables are being involved? You really get a sense of this feel when you see some lol coaches doing vods where they might point at the tiniest details pertaining to minion waves, rotations, and general late game macro. Where did such knowledge come from?

Even better, imagine the same thing with chess. Usually when you are learning chess, there are obviously meta early game moves, but due to the complexity of the game, quickly sense of meta vanishes. Instead what we have is rather general principles of how to play chess well. Where on earth did these ideas come from? Imagine a board with wooden pieces and you have rules on how to move them. How does one go to just knowing bare basic game knowledge to such sophisticated heuristic information? I think imagining this helps deliver the idea of how monumental the tasks seem, yet "we" somehow manage.

Oh yeah also I thought of another reason why grinding doesn't seem like enough. Cause if you see computers that excel at various video games, they only get even close to human levels after what is technically decades or centuries of practice. There is something we are doing with the information that is retardedly superior, but we haven't really tapped into it yet.

Attached: tumblr_o8m1lk7gyL1ud2l90o1_500.gif (500x281, 717K)

>To me a meta is a collection of game-specific strategies repeated by a group of players within some community. I am interested about how such information structures form.
I imagine it's often just standout players leading the way and the rest copying it. Players who are actively "labbing" for new strategies and are successfully pulling them off in high-level play. Others will just ape them often without even knowing how or why this is successful until the cycle repeats.
A combination of somebody putting in some thought and a herd mentality to follow it.
I couldn't tell you how many times it happened in Smash that a character that was thought to be underpowered/low tier suddenly jumps up in tier lists (not claiming it happened often, if the wording suggests this; I literally can't say) even though the game received no patch, because some standout player showed what you can actually do with them. Amsa with Yoshi comes to mind. People just never bothered because the common lore said "this character is not good". How that lore formed in the first place is the question. I guess early impressions quickly cement preconceptions like "that character is bad" and people don't even try after that. So laziness/inertia plays a role? Had circumstances been different and somebody had found out an effective playstyle sooner, things would look much different.
Depending on the type of game, this can also happen by mathematical demonstration. E.g. in single-player games like Diablo, with somebody slapping something in a build creator, saying "look, this has more DPS than the current build we use" and convincing people that way.
Doesn't even need to be a top player in that case. E.g. all it might take is somebody taking the time to really buckle down and test something only to notice "look, if you saturate your bases this way, you get more minerals per minute". Make a nice graph and show it to people and boom, people now saturate their bases with workers differently.

Are you writing your thesis on video games and self improvement? Don't come to Yea Forums for this, you're just gonna get shitposts and wrong information.

>Where did such knowledge come from?
Good players look for ways optimize their gameplay as much as possible.
If someone finds some way to play that gives a sizable advantage it is eventually adapted by the rest of the player base.
Repeat

>>small portion of players who are obsessed with developing strategies discover a strategy and start dominating competition
So I guess you are thinking like the other user in saying it first comes from the motivation to seek these strategies. I don't disagree, though are there interesting strategy discovery methods if this is the case?

>There's no good definition of intelligence, so you're starting your investigation poorly
Obviously there isn't, which is why I gave a specific idea of what I mean. I didn't want to indulge in philosophical questions about what we mean by 'intelligence'. Also you are right, you can have intelligent people that are not interested in winning so obviously it can't be an all-deciding factor. However, where certain measures of 'intelligence' correlate with performance on video games and there is a causal factor, how does it benefit performance? For example Working Memory Capacity being related to moba performance or something. The ability to hold unrelated items in your head might be of a lot of help.

>The ones that had higher skill than me always seemed way more frustrated by bad performance and used that to motivate themselves to improve at the game.
>The worse ones, I've asked them why they aren't trying to get better despite me trying to teach and help them for months. When asked, they usually told me that they have other priorities in life than excelling at a video game and just play it to relax or kill some time.
>One of them literally said he cares at being good at his web dev job and doesn't want to waste time on games.

I see... Thanks user. I should take more note about motivational reasons too.

Thanks too. This could actually also provide us reason to believe metas evolve faster as human progress continues as we can always use previous game fundamentals as key points to ask questions about. This does mean my question might be asked way too late because what I want is primordial too

Attached: 3750fe7ec566443cce32982169a4a3b3--tales-of-berseria-magic.jpg (225x350, 15K)

No this is good so far cause there are quick points made that I had some ideas swimming in my unconscious but didn't consciously put out in my OP. Seeing them and also discussing other stuff is useful for better articulating what exactly I want, and also looking at other things to consider

No problem.
I don't think I have anything else to add but it was nice see such an atypical thread on Yea Forums for a change.

Good luck with whatever you are doing this study for.

I think another interesting thing to note is that the better a meta evolves, the harder it is to improve. You can see this with scientists. It's no longer Edison cranking out a thousand inventions single handed, it's fifty engineers working to make a 2080 8% better than a 1080.

I don't think you're factoring in the impact of the feedback loops involving discussion. Discussion (mostly consisting of but not limited to arguments) both shares knowledge among the community and stimulates generation of new approaches to various tasks.

>because there is a certain 'science' (that makes me sound like a dork lol) to learning how to get good
Just because you want to be a little girl you don't have to act like one. You outright said you're from /sci/ before.

you unget ungood

its really just repetition, critical thinking, and experimentation
meta in video games is just the best way to guarantee a win unless its a game that is completely evened out and takes more skill than numbers

Attached: perfection_is_not_an_art_its_pure_habit.png (640x360, 902K)

Dude I'm not reading all that shit

Attached: c1f703842235c394aa8c1b4e4a32b4e0.jpg (673x904, 438K)

steamcommunity.com/id/tomoko/
>tfw no getting good bf

Attached: 1437548802956.png (345x311, 57K)

Yeah, that's another good point to make. Scientists now collaborate more in groups. The scientists that worked on the fundamentals of QM pales in comparison to those working in contemporary projects. Even mathematicians are going to want to collaborate more (especially with new tools to do so)

>How that lore formed in the first place is the question. I guess early impressions quickly cement preconceptions like "that character is bad" and people don't even try after that. So laziness/inertia plays a role?
Good question. Actually it also reminds me of some coach talking about how if you gave beginners 2 champs, they might say one is worse than the other even though it is the reverse. The reason being those beginners simply arent skilled/knowledgeable enough to play the other superior champ. Like one champ might have a fundamentally less obvious playstyle.

Though overall this is a good summary of the general social mechanisms. I'm beginning to understand it is more interesting to focus the question on cognitive/subjective/mechanical reasons people discover the game different or certain aspects of the meta evolves

This is also true. Conflicts of opinion can help shatter long standing dogmas in the game. I can't think of any specific example of this rn though other than in philosophy and science.

>Just because you want to be a little girl you don't have to act like one.
That's rather presumptuous user. Why do you think that? Also yeah, though thinking about it rn I'm not sure if 'science' is the right term. There are certainly empirical practices made, but they do lack some features of science (for instance a sense of rigour in regards to parsing helpful/wasteful information/methods).

Attached: 111 - DK4nzly.jpg (539x480, 69K)

>shit tier
dont play vidya
>meh tier
dont use the control you have
>good tier
play vidya, use your own controls
>great tier
be autistic or underage puddybrain
>>meta
great tier makes it, other tiers use them to emulate autism or youth

Attached: 837680D3-4F00-4EE8-8C51-0568DCD31725.jpg (905x468, 99K)

>if you gave beginners 2 champs, they might say one is worse than the other even though it is the reverse. The reason being those beginners simply arent skilled/knowledgeable enough to play the other superior champ.
I can relate to that, thinking that Fox was bad in ssbm before any sort of exchange on the game over the internet.
And everyone being unexperienced in the beginning, I guess that's how characters can get a stigma of being "bad" initially.
I want to add that another reason to not re-examine those preconceptions is the risk involved, rather than laziness as I put it. Playing some low tier character might get you nowhere and only reaffirm the common lore, making you waste time that could have been spent on developing play of your character/a character that's agreed to be good.
But then somehow, in a game like Starcraft, people decided after over a decade that mech is viable/good in TvZ. While the game had not received a balance patch in forever, mapmaking was in the community's hand.
I didn't follow the game at the time, but I'd be surprise if the release of maps was the only factor here. The game was in turmoil with the death of the Kespa scene and went from team houses to Afreeca streamers. That must have been a factor. Perhaps this lead to more experimentation? Without a coach hammering the common wisdom into you.
Though of course even within the pro team/team house environment there was innovation, but I imagine such turmoil can shake things up.
Unfortunately I can't retrace the steps that led to this development.