Teslagrad 2

First Impressions

Steam reads Mostly Positive: 447 of 585 reviews are positive, 76% (snapshot 2026-09-12). English alone gives 216 of 292, or 74%. Splitting by language barely moves it.

What is striking is that both camps are pointing at the same thing: a body that moves under physics.

The helpful positive reviews praise movement almost in unison — the slide keeps its momentum, and simply running around is a treat. The top helpful negative (33 votes) says the slide and the ground slam share a button, and that this drops you into pits you never aimed for.

Both are talking about the controls. The thing being praised and the thing being blamed are one and the same.

The original Teslagrad was a game about stopping to read a room. The sequel asks you to read while running. The split in the reviews follows that switch almost exactly.

Putting the Mechanics into Words

The original ran on two verbs, more or less: flip your polarity, and blink a short distance. Both are verbs you use after standing still and reading the room.

The sequel adds sliding, ground-slamming, and repelling off a thrown axe. It looks like addition, but what actually happened is closer to substitution: verbs of observation were swapped for verbs of momentum.

Line up the words the positive reviews keep reaching for and the swap is visible on its own — fluid, flow, momentum, a blast. One of them names Ori and its unbroken motion. What they are praising is not a solution. It is that the speed never stops.

The negative side points at the same place with other words. One review says that as far as it can tell, this game has no puzzles. That overstates it, but the structure it names is right: magnetism is no longer a grammar for solving, it is thrust for moving.

The lineage that treats magnetism as a tool for thought is collected at Magnetism & Gravity. This sequel deliberately steps half a pace outside it.

Feel of the Controls

This is where the split lives. A body that runs on physics lets the same behaviour be called delightful and untrustworthy at once.

The positive side writes that running around is a treat. The negative side writes about grazing a wall and being flung somewhere random, clipping into geometry while carrying a box, falling through the floor into a softlock. Both are touching the same physics.

The sharpest piece of design criticism in the whole pool is in that top negative. The slide and the ground slam share a button, and it can fire on the press or on the release. Slide, jump, let go, and you slam yourself into the pit.

That is not a story about rough physics. It is a story about two verbs colliding. Put two verbs on one button and you open a gap between what the player meant and what the board did. What needed subtracting was not the verb count but the double duty on the button.

The bosses are disliked for the same reason: one-hit deaths, attacks with no tell, cutscenes you cannot skip, no checkpoint mid-fight. One reviewer reports watching the same pre-fight scene something like fifteen to twenty-five times near the end.

Teaching

Teslagrad's signature was teaching without text. In the original that held up: there were two verbs, and each room did the teaching by itself.

The sequel kept the no-text policy and grew the verbs and their combinations. In places that turns not-teaching from an aesthetic into a debt.

The combination everyone names is throwing the axe and then repelling off it. What matters is that the complaint is not confined to the negative side — one positive reviewer admits to being stuck there and genuinely believing the game had softlocked.

The map carries the same fault. Unexplored regions are drawn as blur, and the blur does not clear once you have been everywhere there is to go. Places you cannot reach and places you have not reached yet look identical.

Subtraction is a virtue only when what remains can teach itself. Fez and INSIDE worked without words because, having deleted the explanation, they let each room do the explaining.

Place in the Lineage

The critic aggregate splits the same way. OpenCritic shows a top-critic average of 72 and a 59% recommend rate across 30 reviews. An 8/10 and a 50/100 sit on the same game.

The gap between users and critics is small, though. Both split along the same line: people looking at the art and the movement score it high, people looking at the structure and the endgame score it low.

The widest gap is with the developer's own store copy. The page sells new puzzle-solving tools and epic boss battles. In the review pool those two are precisely what gets treated as the weak points, while the praise goes to the movement the copy barely mentions.

The original arrived in 2013. A decade later the same studio built a different kind of game out of the same material. The useful question is not whether that is faithful, but that the material was pointed somewhere else.

Sources

This piece was written from the user reviews on the Steam store page as of 2026-09-12. Nothing is quoted directly; recurring claims are reconstructed.

・Steam: Teslagrad 2 (Mostly Positive, 447 of 585 across all languages = 76%; 216 of 292 in English = 74%)

・Read: the top 10 helpful positive reviews, the top 10 helpful negative, the 8 most recent in English, and 6 most recent across all languages.

・Critic aggregate from OpenCritic (top-critic average 72, 59% recommend, 30 reviews).

・Developer, publisher, release date and store copy come from Steam's appdetails API (retrieved 2026-09-12). Completion times are approximations from reviewers' own reports and are not verified against a primary source.

Closing

My score is 7.0. It lands close to Steam's 76%, but for a different reason.

What I rate is the decision itself: re-reading magnetism as thrust. Moving the poles from something you read to something that pushes you is not laziness, it is a redefinition of the material. The points come off not for that decision but for not carrying it all the way.

The bosses never once use what the movement is good at. A negative reviewer's daydream — a boss where you dodge red and blue orbs with the magnet — reads as accurate design criticism almost verbatim. The thing this game does best is absent from its biggest set pieces.

Reviewers put the main path at roughly three to four hours, and more than double that if you chase the collectibles and the true ending. Difficulty divides people: the traversal is easier than the original, the bosses feel more unfair.

Buy it expecting the original's stop-and-think and you will be disappointed. Buy it to climb a beautiful northern tower with speed and you will not be. If you want something that makes you solve quietly, like The Swapper, this is not the address.