Pillars 3D

Overview

Walls rush toward you, and each one has exactly one gap. Pillars 3D is a perspective dodger built around a single mechanic: a wall of pillars approaches down a corridor toward the camera, and you slide left or right to line yourself up with the one opening before the wall reaches you. There is no jumping, no shooting, no collecting — just lateral movement and a rising sense that the gaps are arriving faster than you can read them. The 3D perspective is the whole point. Reading where a gap actually is, as opposed to where perspective makes it appear to be, is a learned skill, and the speed ramp means the practice never quite catches up to the difficulty. It is pure reflex and timing, the kind of game where a single run lasts under a minute and you immediately restart to try the same wall better.

How to Play

Slide with the left and right arrow keys, move the mouse left and right, or drag your finger on a touch screen. The camera is fixed at the end of the corridor, so movement is purely lateral — you are positioning yourself along the width of the approaching wall, not moving forward or back. Each wall has a single gap; align with it before the wall hits you. Miss the gap and the run ends on contact. The speed at which walls arrive increases the longer you survive, so the input you had time to line up early becomes a snap decision later.

Tips & Strategy

Trust the center of the gap, not the edge. Perspective makes a gap look wider as it approaches, but your hitbox does not change, so lining up on the visual middle of the opening gives you margin on both sides for the input arriving a beat late. Read the gap early — by the time the wall is close, you should already be sliding into position, not deciding to. Small inputs beat big ones. A long slide across the corridor overcorrects and forces a counter-correction, while short deliberate taps let you hold a line and adjust as the gap clarifies. Watch the rhythm of consecutive walls: if two gaps in a row land on the same side, you can pre-position for the second while the first is still closing, which buys back the reaction time the speed ramp is stealing.