14 KiB
title, url, channel, duration, transcribed_at, transcription_tool, category, tags, summary
| title | url | channel | duration | transcribed_at | transcription_tool | category | tags | summary | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Once you see this, You'll see Competitive Games Differently | https://www.youtube.com/watch?v=NgHvdCcmQ4o | unknown (yt-dlp metadata unavailable at time of transcription) | 14:01 | 2026-08-11 | whisper-cli (whisper.cpp) via MacWhisper ggml-medium model | gaming |
|
Introduces a three-axis framework for understanding competitive games: Micro (execution/aim/muscle-memory), Mezzo (mind-games/probability/reading people), and Macro (systems/routing/optimal strategy). The clearest way to identify which axis dominates a game is the "cheat test" — what single cheat would most break it? Most big competitive games sit in all three, but in different ratios. Applies the framework to CS2, Overwatch, Apex, and Rocket League with input from domain experts. Closes with the insight that players have natural leans toward one axis, and the games that frustrate you are probably demanding the axis you haven't developed. |
Once you see this, You'll see Competitive Games Differently
YouTube: https://www.youtube.com/watch?v=NgHvdCcmQ4o
Duration: 14:01
Transcribed: 2026-08-11 via whisper-cli (MacWhisper ggml-medium model)
Transcript
You know what's kind of weird? When someone asks what kind of games do you like, most people struggle to answer. You say "tactical shooters", and they respond "oh, so you must like Counter-Strike, Valorant, and Rainbow Six Siege", and you say "eh, actually only Counter-Strike". You say "strategy games", and someone recommends you Civilization when you wanted something more like Hearthstone. Game genres are kind of broken. But I think there's a better way to think about this, and it's a much simpler way to look at games. And once you see it, you'll see games much more differently. But let's start somewhere everyone can agree. Every game has something that separates the good players from the bad ones. Sometimes it's pure skill, your reflexes, your aim, your timing. Sometimes it's information, knowing things the other person doesn't. Sometimes it's something harder to name. The mental side, reading people, managing randomness, and adapting on the fly. These aren't just different skills, they are fundamentally different kinds of gameplay, and most games mix them together in different amounts. So let me give you three words. Micro, Mezzo, Macro. Now here's where it gets interesting, and a little uncomfortable. What if I told you the clearest way to understand what a game is actually testing, is to think about how you'd cheat in it. If you wanted to completely break a game, to make it trivially easy, what cheat would you use? The answer to that question tells you more about a game than any genre label ever could. So let's put this to the test. What is micro? Micro is the execution layer, it's your aim, your timing, your movement, your muscle memory. An example of a micro cheat looks like an aimbot, perfect inputs every single time. If a game can be broken by perfect execution, that's a micro game. osu!, Geometry Dash, Aim Trainers, and Tetris are a few examples. You can cheat in osu! with perfect cursor tracking, you can cheat in Geometry Dash with perfect jumps. These are execution cheats, and you can feel the microness when you play them. Mezzo is the probability layer, the place where mind games exist. You're reacting to the unpredictable nature of people, not mechanics. A mezzo cheat looks like stream sniping, knowing someone's habits before they show them, reading cooldowns, seeing things you shouldn't see yet. The best example for the purest mezzo game is Rock Paper Scissors. There's zero execution skill, it's entirely mental, and probability. Among Us also runs on mezzo, so does Liar's Bar. Even the stock market is built around this layer. If you wanted to cheat in Among Us, you'd stream snipe to know who the imposter is. If you wanted to cheat in stocks, you'd insider trade. Notice how none of these cheats involve executing anything better, they involve knowing something you shouldn't. Intro is the systems layer, it's your routing, your resource management, and your win condition. You can think of it like the mathematically correct way of winning the game. A macro cheat looks like a chess engine, economy automation, or you can imagine like a perfect coach telling you exactly what to do. Pure macro games are almost mathematical puzzles. Think Factorio, PolyBridge, Connect4, Tic Tac Toe. Once you find the optimal solution, the game is solved. But here's where it gets really interesting. Most games aren't just one of these, they're combinations of them, and the combination tells you everything about what kind of game it is. So let's look at Micro plus Macro. These tend to be single player precision games. You need to know the optimal solution and execute it fast. Think Rubik's Cube Speedrunning, Mario 64, Jump King, and Getting Over It. If you were to have a perfect micro cheat in Alderaan Ring, you'd have perfect dodges. But if you were to have a perfect macro cheat, you'd use the mathematically optimal build. But there's one weird multiplayer outlier here, 8 Ball Pool. There's a mathematically optimal shot, and then there's the skill of actually hitting it. But even though it is player versus player, the room for mind games doesn't really exist. And I think it's a rare exception. Next let's look at Micro plus Mezzo. This is PvP skill expression games. These tend to be your fighting games, Street Fighter, Tekken, Smash Bros, Brawlhalla. Your skill matters, and reading your opponent matters. But also you have some other examples like Mario Kart, like the micro of driving, and the mezzo of random items and unpredictable opponents. We even see some unique games like Tetris 99. The original Tetris is pure micro. But Tetris 99 also involves mezzo, because now you're managing who to send garbage to. You also have physics based PvP, like Fall Guys and Gang Beasts. Which also fits into Micro plus Mezzo. The next one is Mezzo plus Macro. These are the mind and math games. Using Karthstone, TFT, Pokemon VGC, and Battleship. Now chess is theoretically a pure macro game, but at the human level, there's a massive mezzo component. You're optimizing for your specific opponent's habits. That's why studying opponent matters in high level chess. And then there are some weird outliers. Bellatro is a single player game, but it's also mezzo-macro because the RNG of the card combines with the math of maximizing your hand. You're not executing anything difficult, but you have to pick and choose your cards wisely, and play to their biggest strength. Phasmophobia is also a weird game that fits into mezzo-macro. It's a co-op game, but it fits here because you're deducing a ghost using evidence, which is macro, while the ghost's behavior has RNG, mezzo. Although I suspect non-PvP mezzo-macro games slowly converge towards pure macro over time. Once players fully map the RNG, it stops feeling uncertain and becomes a speedrun problem. And then there's the middle, all three overlapping. This is where the biggest competitive games live. League of Legends, Dota 2, CS2, Valorant, Overwatch, Marvel Rivals, Rocket League, Apex, Rainbow Six, World of Warcraft. But here's the thing, they're not all the same balance. And the cheat test tells you how exactly they differ. So let's use Counter-Strike as an example. Can aimbots win a game of CS completely alone? Almost certainly yes. If you never miss, most duels are over before they start. That makes micro extremely high in CS2. But what about x-ray, seeing through walls? Pros have done experiments with one player with x-ray versus five without x-ray. The five still usually win, but the x-ray player takes rounds they have no business taking. This makes mezzo very high in CS2 too. Which means macro, the strategic layer, is relatively the lowest. CS isn't about long-term economy optimization the way MOBAs are. The rounds are short and the execution is everything. But let's try this test with Overwatch. And right away we see a unique case, because the answer actually depends on your role. But since it varies so heavily by character and roles, I decided to ask someone who knows about Overwatch way better than I do. So if you compare all three like maxed out versions of the perfect version of it, like a cheat version of it, which one would have like the highest win rate and how would the distribution be? I would just say support is more macro heavy, DPS is more micro heavy, and then tank is more mezzo heavy. Also the way that it works is you have offense, defense and control. Control refers to like map control, offense refers to offensive options, defense refers to defensive options. And a character's mixture or like kind of design links to this concept of offense, defense and control. Tanks tend to depend on more control, supports more defense, DPS more offense. And if you actually like dissect each individual hero, they will have different like varying levels or varying focuses on offense, defense or control. So I'm thinking about it like this, right? If you were to give each role one cheat, you would give aimbot to DPS and for tank you would give the ability to see enemy cooldowns as a perfect cheat, right? And for supports you would have a coach or an engine, perfect coach. That's perfect, that's exactly perfect. And also, especially in Overwatch, because it's 5v5, because you only have one tank. Like if you can win every duel as a tank, that's kind of like a free win for your team. But what about Apex? Apex sits remarkably balanced across all three. If you think about which player is most likely to win, a guy with perfect aimbot, a guy with perfect wallhacks or a guy with perfect ring knowledge, all three actually has a pretty good chance of winning the game. If you have aimbot, you can easily kill anyone rotating, but any close range fight will tend to go either way. But with perfect wallhacks, you can use your needs and abilities to take a team down, but you'll still lose to someone with good ring positioning. But even with perfect ring information, you could still easily die to someone who wants to contest that position. Let's test this in a really unique game. Rocket League I asked pro Rocket League coach Luke on his answer to this thought experiment, and this was his response. Basically, perfect micro or perfect mechanics in Rocket League would be defined as taking the ball from point A to point B as quickly and as efficiently as possible. Like how in a first person shooter, aimbot moves the crosshair to the target as quickly and as efficiently as possible. For Rocket League, smarts cannot make up for the fact that a more mechanical person can just take the ball to the open net far more than someone with less mechanics. This makes Rocket League heavily micro dominant. But after micro being the most dominant, he put macro second and mezzo last. This is because even if perfect mezzo allows you to hear enemy comms or see enemy boost, doesn't change the fact that all information in Rocket League is able to be known at all times. And for how fast paced the game is at the highest level, listening to a perfect coach verbatim would be better over hearing enemy comms or seeing enemy boost. Also, the closest thing to perfect micro in Rocket League is the AI bots. I came across footage of three semi-pro players versing three of the best Rocket League AI. And the three AI bots were able to win comfortably. So here's what this framework actually gives you. It helps classify games, but it also helps understand yourself as a player. I noticed something when I looked at which games I genuinely enjoy. I gravitate toward games in the middle, the full micro-mezzo-macro cluster, and toward pure micro and pure macro, but I don't really enjoy micro-macro games. Speedrunning doesn't scratch the right itch for me. And when I looked at why, it mapped exactly. In Apex, I'd argue I aim pretty well and have respectable movement. I'd say I also have a solid understanding of rotations and ring positioning, but I tend to tunnel vision in close quarter fights, specifically when we're in the same building. So I tend to miss mezzo-reads when the pressure is high. And this exactly explains the gap in my game. Because here's the real insight. Every player has a natural lean. Some people are micro-players, they grind mechanics obsessively, and they want the skill gap to feel fair. Some people are mezzo-players, they love reading people, and they thrive in the chaos of PvP randomness. We even see a common cluster of people who play Among Us, GTA Artee, luck-based games like gambling, along with many other party games. And then we also see that some players are macro-players, they love efficiency, they draw routes on maps, they think about the game between games more than during it. And most of us are a mix, but not an equal mix. The games that frustrate you the most, they're probably demanding something your brain doesn't naturally give. The games you keep coming back to, they're probably rewarding the part of you that already runs at full power. Genre labels never told you any of this. Action RPG doesn't tell you whether you'll be grinding execution, reading opponents, or solving systems. It just tells you there's a sword and a level bar. And maybe more importantly, you can stop blaming yourself when a game doesn't click. Because maybe the games are testing things you haven't built yet. In my case it's the underdevelopment of mezzo. And that's the three dimensions of games, micro, mezzo, macro. Every game lives somewhere inside them. And now that you know where to look, you'll never see a game the same way again.