ThundeRobot G45 Pro

Inner Deadzone

The Inner Deadzone is the area around the center of the stick where small movements are not registered. This helps prevent stick drift or accidental inputs, but if the deadzone is too large, it can make aiming less precise, especially in games requiring fine control. We evaluate the Inner Deadzone based on how much you need to move the stick before it responds—the less movement required, the better.

The ThundeRobot G45 Pro has no Inner Deadzone. The stick responds immediately to even the slightest movement, which is excellent for aiming accuracy and micro-control. This makes it a great choice for precision-heavy games like first-person shooters (e.g., Valorant or Apex Legends).

For comparison, many budget gamepads often have a moderate to large Inner Deadzone, while premium controllers typically aim for a slight or no deadzone for better precision.

Want to learn more? Check out our video explanation of how the Inner Deadzone works.

Outer Deadzone

The Outer Deadzone is the area near the edge of the stick’s range where further movement isn’t registered. This can make the stick feel less responsive at full tilt, affecting actions like quick turns or maximum speed in games. We evaluate the Outer Deadzone based on how much 'lost' range there is—the smaller the deadzone, the better, as it allows full use of the stick’s range for more precise control.

For comparison, budget gamepads often have moderate to large Outer Deadzones, while premium controllers strive for minimal or no deadzone to maximize control.

Want to learn more? Check out our video explanation of how the Outer Deadzone works.

Stick Asymmetry

Stick Asymmetry measures the consistency of the joystick's response across different directions. Ideally, if you physically deflect the stick by 80% from the center, the software should report an 80% deflection regardless of the direction. A high asymmetry score indicates a problem where for the same physical movement, the reported coordinates are inconsistent—for example, 60% in one direction and 90% in another. This creates an uneven, often 'egg-shaped,' response zone, which negatively impacts aiming and control predictability.

For the ThundeRobot G45 Pro, the Stick Asymmetry is 20.9% for the left stick and 7.7% for the right stick. Higher values can lead to noticeable inconsistencies, potentially impacting aiming or movement in games.

Testing Methodology: It's crucial to note that this test is performed at partial stick deflection (~80%), using special physical limiters (clips). Testing at 100% deflection often hides asymmetries because the controller's output is clamped at the maximum value, artificially 'smoothing' the resulting shape. Our method reveals the true performance of the stick in the ranges most critical for gameplay. This precise approach was also utilized by Linus Tech Tips in their controller review.

For comparison, many budget gamepads show asymmetry levels above 30%, while high-end controllers typically stay below 10% for better uniformity.

Learn more about how different gamepads perform in the Stick Asymmetry test and how to conduct such a test in this article. You can learn how to test joystick asymmetry yourself from this video.

Circle Error

Circle Error evaluates how closely the stick’s movement follows a perfect circle. A high Circle Error means the path is more square-like, which can cause inconsistent speeds when moving diagonally—your character might move faster or slower than expected. The lower the percentage, the better, as it ensures smooth, uniform movement in all directions.

For the ThundeRobot G45 Pro, the Circle Error is 0.7% for the left stick and 0.7% for the right stick. This is an excellent result, providing smooth, natural diagonal movement similar to premium controllers.

For comparison, budget gamepads often have Circle Errors above 12%, resulting in 'square' feeling sticks, while high-quality ones aim for under 8% for better smoothness.

Want to learn more? Check out our video explanation of how Circle Error impacts performance.

Resolution (Stick Bitness)

Stick Bitness measures the precision of the joystick’s analog input, similar to bit depth in audio. Higher bitness means more distinct positions the stick can register, leading to smoother and more accurate control. Lower bitness can result in 'stepping' or less fluid movement, especially noticeable in slow, precise actions like aiming.

Unlike declared digital resolution, our True Bitness metric is derived from actual physical stick movement, reflecting the usable positions the stick can produce in practice.

For the ThundeRobot G45 Pro, the movement-based True Bitness is 8.0 bits on both sticks. This is moderate precision. You might notice slight stepping or unevenness during slow, fine aiming adjustments.

This corresponds to a measured Step Resolution of 0.00781 on the left stick and 0.00781 on the right, with about 128 SFC on the left stick and 128 SFC on the right.

For comparison, many budget gamepads have around 8 bits, while premium ones often exceed 10 bits for superior accuracy.

Want to learn more? Check out our video explanation of how Stick Bitness affects control. It is important to note that the video specifies the resolution of the stick, not the bit depth; the higher the bit depth, the higher the resolution.

Center Error (Stick Centering)

Center Error (also referred to as Stick Centering) measures how accurately the joystick returns to its neutral (center) position after you release it. A low Center Error prevents stick drift—a common issue where your character or camera moves slightly in a game, even when you're not touching the stick. The lower the percentage, the better the centering, and the less likely you are to experience drift.

For the ThundeRobot G45 Pro, the Center Error is 3.2% for the left joystick and 3.2% for the right stick. This is a poor result. High center error indicates that the sticks do not reliably return to neutral, which will likely cause stick drift unless a significant inner deadzone is applied.

This test methodology intentionally employs a more rigorous approach by implementing small-angle deflection and release, which produces the most challenging conditions for stick re-centering. This technique differs from the conventional maximum-deflection method where the stick is pulled to its full range and released, as small-angle deflection better simulates the micro-adjustments typically executed during actual gameplay scenarios, providing more representative data on potential stick drift occurrence during normal use.

Want to learn more? Check out our video explanation of how Center Error works.

Cardinal Snapping

Cardinal Snapping (sometimes referred to as Axis Magnet) is a form of stick processing where the controller's output artificially 'snaps' or clings to the cardinal (horizontal and vertical) axes when the stick passes close to them. While this can make pure horizontal or vertical movements feel perfectly straight, it distorts the natural movement path and makes diagonal aiming or fine steering less predictable.

The ThundeRobot G45 Pro shows no Cardinal Snapping. This means the stick does not artificially cling to the horizontal or vertical axes, preserving your real movement path for consistent aiming and natural analog control.

Want to learn more? Check out our video explanation of how Cardinal Snapping affects stick behavior.

Disclaimer

We tested the ThundeRobot G45 Pro gamepad using a single unit, so keep in mind that other units of this model might perform slightly better or worse. In most cases, these differences are minor and shouldn’t affect your experience significantly. The results were obtained with the Stick Tracer program, and some values might vary if you use different software or testing methods.

Testing conditions, such as the gamepad’s firmware version (FW: ) or connection type, can also influence the results. If you have this gamepad, we’d love for you to share your own test results! This will help us build a more comprehensive picture of the ThundeRobot G45 Pro’s performance across different units.

Full test results can be viewed on the test page.