If you are reading this post in 2022 or later, please be aware this post was written *6 years ago*. I don’t care about Matthewmatosis. I haven’t commented on anything he’s done in the half-decade since I wrote this. I haven’t even watched it. He might have gotten better according to my weird-ass standard during that time!
I have been developing a critical theory of game design on this blog, all by myself, not involving criticism of other creators. Please read something else! I have written HUNDREDS of posts that are more interesting than this.
Join my discord if you want to tell me how much you dislike this post:
I honestly feel more than a bit of hesitation criticizing Matthew Matosis. He’s a rather popular guy. He puts in the work. He’s extremely thorough He’s not a shameful embarrassment like the Game Theorists, Extra Credits, Campster, or any number of other writers who make fools of themselves trying to critique games, but when I hear him get held up like he’s revolutionary or his reviews are brilliant I start to get irritated. His work isn’t any more exemplary than the traditional game reviews of the past few decades, it’s just more thorough in making kind of vague dry observations about the more obvious parts of games, and it accords with the opinions of many “core” gamers, taking up a lot of more traditionally held values for games and picking the right side on most issues rather than picking the wrong side occasionally, but doing an actually good job arguing his case.
Okay. Lets put it like this. Matthew Matosis is not a game reviewer or critic. He’s more of an experience reviewer.
He’s not reviewing from the perspective that the thing he’s reviewing is a game. This is why his videos are exclusively focused on single player games (‘cept for DotA 2). He reviews them like they’re the new media equivalent of films/books/etc. Thankfully he’s not as dumb as Errant Signal and others like him and doesn’t review them literally like they’re literature or film.
The topic of focus in all his reviews is centered on what the developer is trying to express and whether this comes across in the experience delivered to the player. He cares about the emotional overtones of the game, the belief in the game world through immersion, obvious user experience issues (like whether some part of the controls or camera work weirdly), places where the developer put in effort (whether it’s to success or failure), things that express the design mentality of the director, and an endless stream of little details.
To that end, I’d characterize Matthew Matosis’s critique as Superficial or Bland. He’s endlessly thorough about commenting on every element that exists in a work, but he glosses over every detail he covers. He’s popular because everyone else glosses over details too, but they’re less thorough, get the details wrong, or end up with the wrong opinion. He’s better than most simply because he’s less worse than most and puts in more effort, but in the process he’s produced a better bland review rather than one that actually captures the spirit of the game.
So how are his reviews bland exactly? Lets go with his Mario 64 review, one of his most popular. In his analysis of Bob-omb battlefield, he mentions that there are no bottomless pits, and that’s about as far as his commentary on the first level goes. He makes almost no mention of the orientation of the different elements composing the level, the trip up to the boss for the first star, or the choice or design of the enemies present in the first level. At most he mentions that it’s weird to include the chain chomp, because it’s punishing, for dealing 3 units of damage, and that the boss cannot deal damage to Mario directly. This ignores the way those elements work or the way the player interacts with them, or how, as parts of the first level, they teach the player something about how to play the game that they need to know for later. It ignores how the first level creates multiple possibilities and challenges for the player that push the player to understand and improve at the game, or replay it in different ways.
His remark that the star doors, “blow the door wide open on player progression,” doesn’t mean anything. There’s no description of how this system is beneficial to the game or interesting in its own right. He cares that the unlock system, “never feels too restrictive”. He doesn’t mention any details of how the castle is even connected or what areas and where get unlocked by getting more stars, and the relationship between the number of stars available in the currently open courses to the stars required to progress. By failing to do this, he only has an ungrounded conclusion that he can just say because he can say it. It doesn’t feel like he’s getting into the meat of the game, it feels like he’s just saying random observations that don’t seem to coalesce into any conception of how this game is successful or how to modify or replicate it. Rather than praise the level design of the castle itself, he cares more that the castle acts as a hub between levels because having a level that a character moves around in is less . He cares that the feature is there more than any particular aspect of it, which is why he says that the castle is worthy of praise, but fails to mention any aspect of the castle. He has a very results oriented focus rather than a process oriented focus in his reviews. I don’t think this is a good way to review media products, especially games. To really evaluate whether the product is good, I believe you need to analyze the composition of each element of the product, not merely the superficial presence or arrangement of the elements. He later goes on to mention some ways that the hub allows levels to be manipulated, more content that there are these elements or that they seem particularly clever than what they offer the game.
After this he goes on to discuss some of the mechanics, first stating that they form a sort of “language” with which the player “communicates” with the game. Matosis loves statements like this that have only a vague sort of meaning and don’t communicate any type of useful information. He remarks that the most thoughtfully crafted mechanics are those used for movement, but he barely puts any thought into his analysis of those mechanics.
Next he goes over the way each of the courses has 7 stars that need to be gathered separately. Rather than discuss the impacts this had on the design of each of the stages, what with extraneous elements in courses that don’t fit the current star you’re after or allow you to go down the wrong path, he discusses the reasons the developers might have done this. He has a tendency to do this in many of his videos, attempting to guess at technical or developmental reasons that developers might have chosen a specific approach, instead of simply reviewing the game as it is. He guesses that limited cartridge space meant that reusing assets meant more bang for buck. He guesses that the tools level designers had back then to model 3d objects were more constrained, so it must have been harder to develop 3d levels. He’s an amateur on these topics and frankly doesn’t know the actual limitations or developmental constraints in question. We don’t know how many or how detailed levels could be fit on an N64 cartridge (Mario 64 was 8MB, Star Fox 64 was 12MB, Zelda OoT and MM were 32MB). It’s theoretically possible they could have given every star its own level or had as many individual levels as say mario 3 or super mario world. The BSP editing tools of the time for editing 3d levels weren’t actually bad. They made Quake and Half Life after all. There’s an argument to be made that the BSP editing tools of times past allowed editors to do a better job of designing 3d levels than modern tools which have deprecated or limited many of their BSP editing functions in favor of static meshes, which is why modern level design looks really pretty but has become less engaging, because it’s tougher to quickly and accurately lay out the shapes that form the architecture of the level, allowing for faster prototyping and iteration, which helps designers build levels that play in interesting ways.
Matthew Matosis’s regular habit of attempting to guess at the developmental process (something he’s regularly shown he’s not extremely knowledgeable about) does the games he reviews a disservice. By commenting on the way the game might have been developed rather than the things the game consists of, he’s giving the game an excuse for being worse than it could have been or should have been. He’s brushing over the detail of whether the levels in question are actually good or not by merely remarking on whether they’re “good for the time”, because the constraints of the time shouldn’t be factored into whether they made a good game in the end. Good games were made before Mario 64 and good games were made after it. It should only really matter whether Mario 64 is good, but commenting on development methods, saying that the developers “put thought into things”, are ways of complimenting the developers more than the game, as if trying hard counts for more than genuine success.
His focus on the superficial is why he cares so much about “nice touches” like the decoy penguin near the penguin mother, the execution of wet dry world (purely because it’s unique in concept, he admitted he disliked the execution), the linearity of bowser’s levels versus others, that all the songs are new instead of repeated, the use of sprites instead of 3d models for some objects, that some levels are made with coins inside enemies presumably to get around the sprite drawing limitations, the entire discussion on the snow effect in the snow levels (which is probably wrong from a technical perspective, given the sprites render on top of enemies. It would be harder to make them actual objects in the world and then incorrectly assign z-depth to them. It’s really easy to make snowflakes on a 2d plane with projected depth through simple parallax math so that it matches the rotation speed of the camera, appearing to have a place in the environment, but not actually having one, requiring you to draw it on top. The idea that they’d generate and keep track of tons of particles in 3d space, then deliberately strip them of their z-depth is absurd. This whole section feels like him trying to impress us with technical knowledge, he does this frequently and is usually wrong. EDIT: Seems truth is stranger than my guess or MM’s, the particles are actually generated in 3d space in a limited area very close to the camera, explaining why they always overlap objects. This is why I usually don’t guess at technical implementation), the story of miyamoto originally making the game about chasing a rabbit until the controls felt good enough to build a larger game on (completely skipping a discussion of the actual feeling of mario’s movements), the development of the sideflip and backflip.
Even after acknowledging the huge role the movement mechanics play in the game his dissection of them is disappointing. He says the walljump “feels fantastic” making mario “look like an acrobatic master”, which skips over the details of how it actually achieves that, like the windows on when you can walljump, under what circumstances and how it affect’s mario’s momentum differently than standard jumps or other options. He continues to fail to offer such analysis for his other moves, and fails to analyze the relationships and tradeoffs between all the moves, despite himself saying they were so important to the game. Again he makes a guess into the development process of the game with the sideflip and backflip, though admittedly prefacing it with a functional purpose to the division of these moves. The sideflip very well could have been developed first for all Matosis knows, with the backflip added as an afterthought, but as critics and analysts we don’t know that, and in the end it doesn’t matter, only the product as it sits in front of us does.
His entire commentary on “inertial frames of reference” is flat-out wrong on a technical level. Unlike reality, in a computer game, there is no such thing as an inertial frame of reference. In reality there are no absolute positions, and no absolute frame of reference for inertia, in a game there is an absolute reference point for positions, and absolute velocity relative to the absolute system of coordinates. Elevators and moving platforms in the real world work by forces of friction and repulsion. Masses push on each other. The same in video games work (usually) by literally gluing the character’s feet to the floor for as long as they’re touching that surface, moving the character in lock-step with the movement of the platform each frame. Games that allow you to jump on a platform and retain momentum don’t do so by having you be joined with that platform’s internal inertial system, they do so by having you inherit the momentum of the platform as a velocity impulse applied to the character’s total X, Y, and Z velocities at the point where you cease to make contact with the platform. You can see this in games like Unreal Tournament, where you can jump off elevators to keep momentum from them, but if you don’t jump then you won’t be launched when the elevator finishes moving. This type of inertia inheriting from moving platforms wasn’t present in earlier iterations of the series despite his claim otherwise. The bigger issue is that he innately assumes that it’s a bad thing that Mario 64 lacks this, rather than evaluating the challenges in the levels for whether they’re better or worse off for lacking it. A simple point to make is that many of the cases where bars block a moving platform’s path would be simplified or trivialized if you could simply jump when it approaches you with no input on the analog stick, where lacking the inertia inheritance makes it so you need to not only jump correctly past it, but also plan your jump to land on the platform at the end too. A case where a lack such inertial inheritance is more clearly a flaw would be Crysis Warhead, in the train level, where it’s much more difficult to jump forwards up the train in the direction it’s moving than back, made especially irritating when you want to get on top of the front cabin. Realism (or the fictional equivalent, consistency/coherency) is prized by Matthew Matosis because he does not prioritize good gameplay interactions and this is a very clear case of that.
I predict that his reviewing methods would not hold up if he attempted to review a multiplayer competitive game, such as a fighting game, competitive shooter, or RTS. Reviewing competitive multiplayer interactions requires actually talking about the way the mechanics function and how they interact with each other to create a wide range of viable strategies, and can be precisely manipulated to produce different outcomes, something he’s been particularly weak at doing in his videos. That and it seems unlikely he’d be able to make fine criticisms of the small details that shape entire games, like how crouch techs, slow walking speed, 1 frame links, focus attacks, and invincible backdashes shaped the way SF4 played, changes in the pathfinding engine changed the nature of StarCraft 2 from Brood War, or the massive number of changes between Melee and the later Smash games that completely changed the way they play (changes that casual players typically gloss over). He hasn’t paid this much attention to detail to mechanics in his prior videos and it seems unlikely that he’d be able to describe it, much less model how it shapes the entire structure and strategies people play with in a multiplayer competitive game. For that matter, given his bias towards developer intent, he might even label something like later Smash Bros games, or Gunz 2 as being the correct move by the developer, since they more clearly accomplished the vision of the first games, despite being overall worse as games. He simply doesn’t have the correct perspective or methods to review games as games, and I imagine this would shine most true in his hypothetical reviews of competitive multiplayer games. He cares more about things like immersion than fun gameplay, which is regrettable, because it leaves gameplay up as something ineffable, something assumed to be good because all the accompanying elements are good, rather than evaluating the goodness of the gameplay in of itself.
And he seems to take a weirdly congratulatory stance on tons of things that don’t really matter, where he congratulates the developer for doing things one way when it doesn’t honestly matter how they did it, but seems to congratulate them less for making the right decision and more for having a thing in their game that is noticeable and affects the player. He cares a lot about superficial details that are plainly visible and show thought of some kind, more than say game mechanics or whether the details in question actually uphold the central design. Through this he seems to paint a picture that making a good game isn’t about fun interactions, but rather an endless assortment of tiny details that give character to the world and experience, even if they’re completely pointless, or if the alternatives were just as valid.
He’s not a *game* critic. He’s about the “experience.” You’re not going to leave his videos with a better understanding of what makes a game good or bad or why that particular game was good or bad. You’re going to leave it feeling self-justified and happy about whatever game it is, because hey, it has so many places where the designer “clearly put thought into it” that it MUST be good, because tiny details are what makes a game good rather than whether it sets up a challenging experience that through its challenges force you to come to understand a system of possibilities across which you can exercise many different choices between solutions to problems, playstyles, and selections of content. He’s not a game critic, and when he tries to critique, like with Dark Souls 2, he falls flat because he gets details horribly horribly wrong. Because trying to say what went wrong with something and how to fix it demands a level of accuracy and awareness of how the parts in the system fit together that he does not normally deliver.
I think his Dust Force review is a good example of his failures to because he described that game as being “built for speedrunning” when I’ve always seen it as built to discourage speedrunning, to be one of the least interesting speedgames out there. Half Life is built for speedrunning. Super Metroid is built for speedrunning. Dark Souls is built for speedrunning. Of course none of these games were literally intended to be speedrun, they just happen to have a structure that is naturally good for it, because they were built to be good games. Dustforce is built to follow the path as optimally as possible, not to find or create it. And Mirror’s Edge’s was intended to be built in the same linear and limited sort of fashion, yet outside of its intention it’s a great game for speedrunning because of all these unintended factors that rip apart the developer constraints and create freedom and challenge for the players, but he can never acknowledge a game like that because he exclusively reviews within the author’s intention. Speedrunning isn’t something explicitly designed for, it’s a function of the game that is good or bad based on a higher principle outside developer intention. Of course he’d pick a game that was deliberately built for speedrunning with a lame speedrun. It fits his modus operandi perfectly. He doesn’t care about gameplay, he only cares about developer intent and vision for the player’s experience, and this should make it clear.
He has a video that spends 10 minutes telling you, “did you know Pac man has scores? Did you know that there’s an end to the game and therefore a maximum high score? Did you know that the ghosts follow the same pattern every single level so if you move the same way, they will too?”
He never once talks about the scoring system of pac-man in that entire video. He doesn’t mention how high scores grants extra lives, he doesn’t mention how fruit appears, or the bonuses for eating multiple ghosts in a row. He doesn’t talk about eating the pellets granting score either. He’s not talking about how these elements cohesively create a challenge of conflicting motivations, making score optimization difficult, because gaining in one area means sacrificing in another. He’s clearly never watched an actual pac-man high score run, since he claims that because the ghosts are deterministic that players repeat the same thing every level, which they clearly don’t and he’d know that with even cursory research.
It’s a long rambly rant on the fact that maximum scores for games literally exist at all and the realization that competing over time supplants score when reaching the maximum score is attainable/common.
It’s not a catch 22 situation between randomness and determinism, there are perfectly deterministic score based games that people still have not reached the maximum possible score in, even after a factor of decades, because people are inconsistent. We’ve known how to get the optimal time in Super Mario Bros and Donkey Kong for years and years thanks to TASers, but no human player can get anywhere close to TAS times in most games because they’re too inconsistent.
Maybe I’ve thought these things through a bit further than most people have, but a lot of these remarks on scoring are really obvious, basic conclusions that are kind of ignorant of how people play games. A lot of people go for scores in games where they could never possibly reach the top score, because it’s fun to get a good score. Games ending isn’t necessarily a problem because perfectly optimizing for score in them is impossible because it requires too finnicky movements. Imagine a maximum score run in Touhou, a game where you get scored for brushing up against as many bullets as possible, hitting the enemy with your shots, especially out of focus fire mode, and also collecting item drops, you can collect all the items on the screen by going above a certain point on the screen, but that puts you further away from the majority of bullets and also puts you closer to the source of the bullets, which is dangerous.
That and he doesn’t address how a challenge may have randomized elements but still be fair because it’s within reactionary boundaries, like tetris, or like common reflex tests, or like practicing Fox up throw chaingrabs, or tech chasing in the 20XX hackpack.
That and some games may be so difficult that even if they’re deterministic, most people don’t last too long in endless. Especially true for real time games with a decay factor of some kind that will destroy you without sustained input (see flappy bird, even though flappy bird was somewhat random) or ones that scale up in difficulty over time.
Most of his whole rant isn’t applicable to most of all score based games, it’s really only worrying about upper boundary concerns that aren’t even relevant for the vast majority of games that have scoreboards at all, because most people have never and will never reach the maximum score. It has bloody nothing to do with scoring systems, how they’re designed to be fun, or even how pac man’s scoring system is fun. There’s almost nothing that can be learned from that video, it’s practically masturbatory thought for the sake of thought on something incredibly obscure and irrelevant.
That and he didn’t mention space invaders, where the high score was achieved over a course of days continuously playing the machine, and they took rest periods by leaving the machine on and letting all the extra lives they had racked up tank for them so they could eventually come back and keep going.
Another one I’d cite is the Game Design Companion on Warioland 4 by Daniel Johnson, which unfortunately is not a game review, not judgmental in any way, and is also too long. I expect that type of information to be embedded in game reviews in a more condensed format and used to assess games more frequently, and that’s a very different information type than I see in his reviews or any existing review. Articles like these are rare as all hell, and they’re obviously not game reviews, but they’re the direction game reviews should be moving in to more appropriately assess games from a systems perspective. That and you use weasel words like “they must have put a lot of thought into this” or “This feels like the right balance between whatever and whatever” which are filler statements that don’t really explain anything or explained how they arrived at these conclusions, or by what criteria.
A more recent example would be Joseph Anderson, who even when I disagree with him, clearly describes the gameplay and level design elements of games, using evidence from the games to back up his conclusions about games instead of mere assertions.
Matthew’s not thinking about games on a technical level, on a level of how the game accomplishes the goal of being a game, a system that you try to overcome and how you fail sometimes, but you get better so you can ultimately move forward or do better, that produces a variety of different possibilities from player decisions and uses difficulty to encourage balance between all these possibilities, so the player has to think and execute using a variety of skills. He’s not thinking about the process by which players overcome satisfying challenges, he’s thinking more about how the game is constructed to make the player feel like they’ve gone on an adventure in line with the developer’s vision for the player experience. This is why I’d like to see him comment on a multiplayer game or attempt more critique of games, establishing what went right/wrong and how to replicate/fix it, because in that context he’d have to try harder to do a better job.
But hey, lets look on the positive side: Matthew Matosis is less bad than the common game reviewer. He’s not doing anything outright offensive, like giving the wrong opinions of the wrong games like common game outlets are typically indicted for. He’s certainly thorough if nothing else. However his approach puts gameplay at a lower priority than superficial bland commentary on assorted details, and I worry that it perpetuates and inspires a new generation of critics who repeat his mistake.