Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

To get good, go after the metagame

commoncog.com

Post reply on HN
Print view View original

151–160 of 307 posts

Re: To get good, go after the metagame

#151

Post by echelon »

screye wrote: post #138
pretendscholar wrote: post #129

When did meta come to mean strategy? It can't really be 'meta' if its about one specific game or implementation.

MetaGame is not strategy. The metagame is often the behind-the-scene variables that enable a certain set of strategies to flourish. For example, in Dota, changes to armor or magic resistance scaling can change the metagame to be magic or physical focused, but the strategies and trends themselves develop in ways that take the meta-game into account. During one meta game, it is common to see various different strategie…

To further expound upon this, the vast majority of a game is encoded in the engine or ruleset. Core strategy is in dealing with these these ground truth fundamentals. This is the realm of beginners and casual players.

Small tweaks to the rules or weights (simple numeric multipliers!) can dramatically change how the game is played amongst those who are incredibly skilled at the game's fundamentals.

Metagame is truly emergent phenomena.

Re: To get good, go after the metagame

#152

Post by munchbunny »

rc-1140 wrote: post #58
Proziam wrote: post #3

Getting good in almost all games is based on understanding and mastering the fundamentals to the extent that you can make consistently correct (or at least a high degree of 'correctness') decisions based on them. This is true for all esports titles, and probably all games in general. Mastering the fundamentals will make you 'good' to a level that very few people ever reach. It's not until you reach a level where ever…

Agreed. Since a bunch of people seem to be responding to you about competitive games/experiences, my friend left me with the following pointer when he gave me a lengthy rundown on 2D fighters (cleaned up a bit for HN posting): "Tech and meta aren't things to chase after because they're things that come with learning fundamentals. You practice fundamentals until you die". It was like that when I was playing Quake, whe…

In my experience, there's a sort of hierarchy of skills relevant to the game or competition. There are basic skills (like being fast and accurate with aiming, or predicting the opponent's movement), and then there are skills that are dependent on basic skills (leading targets), skills that are only possible once basic skills become automatic (map awareness), or skills that are situational.

The "fundamentals" are usually the 80% mark on this graph: it's the generally agreed upon set of skills that are relevant in 80% of situations, so if you have your fundamentals down better than your opponent, you will win in 80% of those situations. The meta might win you the other 20%, but that's capped at 20%.

Re: To get good, go after the metagame

#153

Post by 83457 »

scott_s wrote: post #144

A saying I have related to this is "Rules make sports." The skills and strategies that matter in a sport develop around the rules. Change the rules and you change the sport. The judo example the author presented is actually one of my go-to examples as well. Not only did disallowing grabbing the legs take out an entire suite of offensive options, it took out defensive options in judo. In judo, the main way to win is t…

The 2004 Olympics Judo representative for USA at 99kg, Rhadi Ferguson, retired from Judo competition after leg grabs were made illegal because that was such a big part of his game.

https://www.youtube.com/watch?v=l0Jj7r3CLwQ

Re: To get good, go after the metagame

#154

Post by WhompingWindows »

The metagame is PARAMOUNT in Starcraft Brood War, which has had no patches in nearly 20 years. The only thing that's changed is the map pool and the players' skill/knowledge (finding some bugs, mapping out defense to rushes, etc.). Thus, players have years and thousands of hour to grind "standard" or "optimal" strategies, and someone who is less creative but more mechanically gifted can advance just by copying cookie-cutter strategies but executing them 5% better.

However, the Brood War leagues know this tendency, so they often add crazy maps to the mix. This season, ASL added Inner Coven, which is a really bizarre island-ish map, and has created a totally new meta. Check out this TvT, it's one of the weirdest games I've seen in years, all due to a map prodding the meta game.

https://youtu.be/yF6GczAXpJI?t=3185

Re: To get good, go after the metagame

#155

Post by juliend2 »

gbasin wrote: post #78

I love this concept. I've been thinking about ways of "traversing skill trees" and identifying meta-games for some time. Collecting ideas here: https://garybasin.com/thinking-toys/ I think this can be done systematically...

The "skill tree" image seem like a good mental model[1] for learning.

[1] https://fs.blog/mental-models/

Re: To get good, go after the metagame

#156

Post by ska »

clairity wrote: post #137
ska wrote: post #107

Earlier quoted context omitted.

I think the idea is that if you aren't operating at "pro league" level in anything, you should focus mostly on getting the basics right and not making mistakes. But if you are operating at that level, this is table stakes and you need to do more, or your competition will overrun you. I suspect most people/organizations operating at the highest level understand or have internalized this, but many in the other category…

> "I think the idea is that if you aren't operating at "pro league" level in anything, you should focus mostly on getting the basics right and not making mistakes." still, that's just a matter of definition and perspective. even at a pro level, if you don't win, you made at least one mistake relative to your opponent. so to win, make fewer mistakes, right? i don't play video games anymore (and never played more than…

> so to win, make fewer mistakes, right?

I don't think that is correct according the to thesis (and again, haven't read it so may be misrepresenting).

They are saying that at that (pro) level, not making mistakes is not enough, you have to make things happen too. Obviously making too many mistakes will cost you, but they are suggesting that playing "perfectly safe" can still easily lead to a loss. I don't think anyone was suggesting that pros don't make mistakes (hell, there are entire stats dedicated to this), but rather than merely focusing on minimizing them will not win you a championship.

The idea is that conversely at lower levels of performance, this isn't the case, and it really is mostly about lowering mistake rates. More tellingly, it suggests that the opportunity cost of focusing on anything else is likely too high.

To tie it back to the actual post, this suggests if you are spending a bunch of time and effort "metagaming" but you haven't already mastered the game, you are probably going to have poorer returns than just getting better at the game. This at least sounds plausible to me.

Re: To get good, go after the metagame

#157

Post by _hardwaregeek »

scott_s wrote: post #144

A saying I have related to this is "Rules make sports." The skills and strategies that matter in a sport develop around the rules. Change the rules and you change the sport. The judo example the author presented is actually one of my go-to examples as well. Not only did disallowing grabbing the legs take out an entire suite of offensive options, it took out defensive options in judo. In judo, the main way to win is t…

Fencing exhibits this phenomenon nicely. Fencing does not come even close to mimicking a real sword fight. One reason for this is right of way, basically the idea that the person who attacks first gets precedence and the onus is on their opponent to defend against the attack before attacking themselves. However what is defined as an "attack" in fencing is very different from what a layperson would assume. Really, it's defined as forward movement without a clear intention to take the blade. Which means that tactics such as advancing with ones arm held back are rewarded.

A lot of people, upon hearing this, respond with something along the lines of "fencing is stupid! We should make our own sword fighting system that encourages real fighting!". HEMA is a nice example of this. Putting aside the questionable logic of a martial art based around swords, what inevitably happens is that as the system develops, people want to compete, to see who is the better fighter. Since they clearly can't judge fights by, well, murder, they need to come up with a rules and points system. Once this system develops, someone starts to realize "hmm, if I do x action, I can win fairly easily". Thus a meta develops. Once a meta develops, everybody starts using the meta to win and the fighting becomes less mimicking killing people with a sword and more competing in a sport.

Some argue that the way to prevent this is to not have competitions or rules. But...then you have a bunch of people waving around swords with no clue as to whether it's actually effective.

Side note, fencing has had a few of these major meta developments in its history too. Johan Harmenberg famously pushed epee's meta to be a lot more athletic and dynamic.

Re: To get good, go after the metagame

#158

Post by rasz »

majos wrote: post #66
asood123 wrote: post #20

One of my favorite books of all times: The Art of Learning by Josh Waitzkin. He was the chess prodigy written about in Searching for Bobby Fisher . He quit chess shortly after and became a world champion in Tai Chi. The book is about learning two very different skills and how they are the same. Thesis that learning one thing deeply helps learn other (unrelated) things makes total sense to me.

If anyone reacts as I did and wonders how it's possible to compete in Tai Chi, the competitive sport [1] is not the slow-movements-in-the-park activity I had in mind. [1] https://en.wikipedia.org/wiki/Pushing_hands

its pure bullshido https://www.youtube.com/watch?v=d-KX99T5r3A

Re: To get good, go after the metagame

#159

Post by WhompingWindows »

Proziam wrote: post #3

Getting good in almost all games is based on understanding and mastering the fundamentals to the extent that you can make consistently correct (or at least a high degree of 'correctness') decisions based on them. This is true for all esports titles, and probably all games in general. Mastering the fundamentals will make you 'good' to a level that very few people ever reach. It's not until you reach a level where ever…

I disagree slightly. The metagame is relevant whenever you're facing opponents of similar skill. If you're way better than your opponent, you can use highly sub-optimal, non-meta strategies and win through sheer experience and skill. For instance, I could kill 50% of Starcraft opponents lower than me by building only one unit and even announcing which unit I'll mass up.

However, if you're making a similar number of errors as your opponent, then the meta does come into play. Regardless of the raw error rate, where pros make few and amateurs make many, if this rate is similar to your opponent, then it still matters if your opponent has a strategic/meta counter to your strategy.

Re: To get good, go after the metagame

#160

Post by rasz »

dillonmckay wrote: post #84
majos wrote: post #66

Earlier quoted context omitted.

If anyone reacts as I did and wonders how it's possible to compete in Tai Chi, the competitive sport [1] is not the slow-movements-in-the-park activity I had in mind. [1] https://en.wikipedia.org/wiki/Pushing_hands

Yes, it is a martial art, and there are some interesting moves. Also, if you enjoy that sort of thing and action movies, I recommend: Tai Chi Master https://m.imdb.com/title/tt0108281 Man of Tai Chi https://m.imdb.com/title/tt2016940/ There are also random youtube videos of various ‘push hands’ techniques.

Its TMA, as effective as ballroom dancing. Do not try to delude yourself into thinking it has any defense qualities.
Post reply on HN