The twelve levels of cognitive reflection in RTS

Similar to Yomi Layers in fighting games:
https://www.sirlin.net/ptw-book/7-spies-of-the-mind

I think what differs from fighting games is the time component.
I think you have to break down the thinking in more layers in RTS.

Fighting games resets to neutral state quite fast and often after a combo has been done.

The timescales and cycles of actions and reactions are quite short and limited in scope in comparison to RTS.
Most RTSes also have a fog of war where you can't see the exact state of the game at all time. Fighting also have hidden state in the sense that humans have reaction times and will have a delayed response to their opponents input/moves in the game.

If you see a Mace coming at your base early in a game of Zero-K, there several time sensitive strategic choices you have to make.

Information from the past.
When you spot the Mace, you can gauge when the enemy made the Mace and how all-in this rush is. If you are familiar with the map, you might even be able to tell if the enemy used their commander to boost out the Mace or not.
You should also have perfect information about your buildorder and how much you resources you spent on non-fighting units such as solar collectors, mexes and buildpower.

Information in the present.
When you spot the Mace, you can access all the information in the present to gauge a proper response.
Information about the enemy position is also important, as to whether the enemy commander is joining the Mace or staying behind and building economy. How early you scout the enemy is an overall strategic decision that goes into the metagame of Zero-K.

Information about the future.
You should be able to gauge when the enemy Mace will reach your base or expansion, and whether to pull back cons or the commander to protect your base, or just keep expanding. Pulling the commander to protect the base, might be the wrong decision if this is just a slow push to try and keep you from expanding and the enemy is ecoing behind this.



If I were to try and simplify this I would divide it into 4 components.
Execution
You
The opponent
The metagame

The most important part is arguably the metagame as it feeds back into all other aspects of gameplay.
However, gauging the "actual" metagame of an RTS is one of the most difficult parts of understanding the game.
Here I imagine the metagame applying to even small micro decisions such as raiding and jiggling of Glaives and stuff.
Glaive jiggling is strong specifically against Scorchers, so how good you are at Glaive jiggling, affects the Rover viability on many maps.
I find that when I play Helwor who likes to jiggle his Glaives a lot, I cannot let him just focus on one group against my Scorchers as that will make my Scorchers lose hard.
However, if I spread my Scorchers into several groups, it forces him to spread his attention and micro, as it is almost impossible to jiggle several groups of Glaives at once against my Scorchers.

So the pregame factory choice is affected by Glaive micro, which is affected by my opponents skill and multitasking, which can be countered by spreading up my Scorchers into several groups and multitasking them to avoid single group fights against Glaives.

Hm, having played SSBM competitively, I think that the prediction level is a lot more detailed in fighting games. If ZK had fighting game mechanics, a single glaive could beat a paladin if it just predicted all it's moves correctly.

quote: Fighting also have hidden state in the sense that humans have reaction times and will have a delayed response to their opponents input/moves in the game.

It is fascinating how our brains can live in the future. I onced watched a match of SSBM on stream, where both opponents seemingly reacted without any delay. Suddenly, the CRT turned off and the 2 players needed around 300ms until that info showed in their faces. Basically, that input was unpredictable, so it took full time to react.
See also Ping Pong for a classic example of the brain having to anticipate where the ball is going.

There's something off about this.

The points are relevant and some require more advanced knowledge than others, but imo there isn't a clear hierarchy.


A player with a level 2- thought process making decisive plays may often outperform a flawed mid-level player that just overthinks things and spreads itself too thin trying to react to or anticipate different threats (raiders here! skirmishers there! oh they have air too! ...)

Alright, as the dude in here that actually travels for fighting games, I feel it is my duty to set some of this conversation straight.

Levels

First of all, I have better arranged/simplified the levels in the first post, because there is no need for 12 of them.

Level 0 - Mashing buttons.
Level 1 - Reacting to what is in front of you on the screen.
Level 2 - Thinking about your actions in terms of executional ability and routing.
Level 3 - Considering stuff your opponent is likely to throw at you (usually referred to in situations as an OS), and your response.
Level 4 - Actions taken outside of the game in order to prepare, such as labbing setups or counterpicking based on game balance. This generally can be considered as other preparation rather than deliberate actions in game.

This is really all there is to it. Any other level above is simply a combination of these, or these levels combined into something else.

RTS vs Fighting Games and why we really can't compare these two

For this, I am going to present some truly strange evidence. I am going to tell you guys about a time I got high.

So, some non-ZK related background: I am a semi-pro fighting game player (formerly ranked world top 50 in a number of games), and I also have played a lot of Smash Brothers. Like 4,000 hours plus. I have also likely put in some similar amount of time in Zero-K over the years, although I will never know because hour tracking was not a thing prior to Steam. Thus, it is not a stretch to say that I have about the same amount of time playing both.

One year, at a New Year's party me and my friends (also competitive high ranking FG players) decided it would be a wonderful experience to get extremely drunk and high and try playing Smash. However, what we discovered is that for fighting games, a large part of the experience happens largely without any intentional thought, at least for experienced competitive players. We ultimately just ended up playing the game like normal, and we were all slightly surprised (and disappointed) the day after because of how little changed,

However, could I do the same thing for Zero-K? Absolutely not, and it doesn't really have anything to do with being an experienced player or not.

Zero-K is a game where much more individual thinking is required based on the situation you find yourself in. Unlike fighting games, it's quite hard to outplay your opponent in ZK on a technical level, as micro plays a much smaller (albeit present) part in the game. While it does matter, macro play in ZK is vastly more important, and it's much much harder to boil down macro play to the same type of flowchart we can simplify micro play into. That being said, I have to actually think quite hard every time I play Zero-K compared to fighting games, and the type of thinking I'm doing is largely different, so I feel as if the levels here might not directly apply too much.

Now, someone has been reading this post and has been thinking, how is he going to drag Magpie into this discussion? Believe me, we are getting there.

Zero-K is interesting because it takes intentional steps to limit some of the micro that could be present in the game in order to force the player to interact with the game in specific ways. In doing so, some of the generic logic about high intensity micro play being very good in ZK has to be interpreted in some unorthodox ways.

For example, for those of us who have been around for awhile, there used to be a projectile dodging widget. In response to someone making this, the devs actually went through the trouble to limit information avaliable to widgets (although a workaround was found by someone later on) and even though a PR was made to potentially add it to the main game, momentum for it died since adding that sort of high micro nonsense was bad for the game's balance.

Fast forward to the year 2026, and we have a certain bird in our midst. On top of being somewhat annoying to fight en masse (although not particularly strong), it also happened to gain coincidental strengths through intensive micro, making it a very appealing unit to spam and micro around all the time. Thus, the unit was nerfed, and is now much harder to micro with due to having a lower HP pool.

There is a area of micro that ZK devs seem willing to support, but we can safely assume that most micro level play in this game directly involves raiders and or single big unit micro and not micro in terms of large scale armies, although perhaps this can be said to also sometimes exist. Roping back into the levels discussion, this sort of intentional limitation of micro play and the directly of ZK even compared to other RTS games and then the huge leap across the rift to compare it to fighting games is sometimes not very compelling, so it's very hard for me to apply this same principles to ZK as a fighting game player. For example, if I were trying to apply the concept of high low strike throw (4ways) to ZK, I cannot think of a real comparison. Plenty of tricks exist in ZK to force 2 way mix, but I can't really think of anything else that forces a real 4 way scenario like FGs do.

This guy gets it.