-Must have at least played through the game twice for better understanding or finding missing details
-An explanation of the game itself
-Explanation on WHY things are good, and WHY it meshes so well with the framework of the game, instead of just saying the graphics are good or the music is good, shit like that is why I read negative Steam reviews more often
-Explanation on things from a technical standpoint (optimization, available settings, loading times etc.)
-Reviewer needs to at least know more than a thing or two about the genre it is reviewing rather than come off as someone who is playing a certain genre for the first time (I don't need to hear shit about danmakus being too hard from a casual who only reviews strategy games)
-Being able to discern bullshit from actual difficulty (like bulletsponges or trial 'n error traps)
-Reviews shouldn't end on a GET THIS GAME/DON'T GET THIS GAME note, but rather 'I would/wouldn't recommend this game, BUT - (insert argument for why a certain demographic or a fanbase might like/dislike this game)'
-Morality is very subjective, and shouldn't affect the game's score (or see the suggestion above)
-Some proper fucking disclosure about any possible relations between the reviewer and the publisher/developer
-No overuse of flowery language as if you just got an English major (see
Adam Sessler's TLoU review)
-In-depth explanation of the game's mechanics
-Reviews about old games should take its age in consideration, and/or at least mention how this game would stand up in the past and how it stands up in the present
-Taking other entries in the franchise of the reviewed game in consideration, and mention how the reviewed game has improved/regressed in comparison of the preceding games
-Mentioning game length and what factors could influence it
-Understanding good level design, enemy placement, enemy variety, weapon variety, weapon balance, enemy AI, friendly AI,weapon design, weapon 'feel' (animations and sound), encounter design etc.
-Whether the game 'lives up to the hype' is irrelevant, because after three months of the publication of your review, new readers won't be able to feel any kind of real hype towards the game
-A review of all game modes/minigames as well (unless the multiplayer/co-op is dead and there's no AI bots)
-Mentioning if the game is really original, and if it executes its concept well regardless of originality
-Whether the developer is a massive shitstain is irrelevant, maybe the very chair you are sitting on is made by a massive shitstain, or the table your PC is placed on is made by a massive shitstain, or the bed you sleep in is made by a massive shitstain, or yadda yadda you get the point
-Keep the target demographic for the reviewed game in mind, don't go ragging on Sonic games for not having romanceable party members
-The paragraph about 'good/bad graphics' should also mention artstyle, animations, UI design, particle effects, lighting, backgrounds, (overuse of) post-processing effects, whether designs match the tone of the game itself, and aesthetics
-Further explanation on what makes the atmosphere so good. Is it the sound? The world design? The ambience?
-Explanation on the controls, if they're fluid, responsive, or at the very least customizable
-Explanation whether certain elements of the game actually benefit the whole, or if are they just added for the sake of it
-Automatic disapproval of the game if it's a RPG and the bartender has less than 4 dialogue options
-Explanation on the skill/progression/difficulty curve of the game
-Humor is subjective, and different people like different kinds of humor. At least mention what kind of humor there is in the game, and if it's well executed/not too repetitive
-If mods make the game good, review the vanilla game first, and the mods afterwards
-"It's fun with friends" is an invalid argument since ANYTHING can be fun with friends
-Don't insult your fucking audience you fucking prick