I have been collecting mice for benchmarking, I think of all the peripherals available mice are the least forgiving, or at least the most transparent. 2026 has also seen a significant shift, with the introduction of the Logitech Superstrike X2 and Razer Viper V4 Pro, each taking a different approach to what the future of HUI (Human User Interface) may look like. In may ways I am a traditionalist, my belief that with hardware entering diminishing returns along with gaming markets becoming insubstantial in comparison to Enterprise A.I, we are approaching (if not already arrived) at a point where there is no longer going to be gen-on-gen advancement in pure compute. Which leads me to mice, and to the market.

It is very difficult to discuss market fluctuations — as the only real way to evaluate is in retrospect — but I do believe that we can make comparative analysis. If we look at the motor industry, or at instruments, a pattern of user experience emerges once a comparative performance is achieved. The term “feel” is often used as an ineffable reasoning for direct, often analogue, experiences, but as software and interactive media is by its very nature digital, there has always been a disconnect between player input and on-screen action regardless of mechanical actuation. With industry norms for televisions now approaching 120FPS, and the majority of OLED (Organic Light Emitting Diode) panels easily achieving the magic sub 13ms response rate, input lag will be noticeable. It is in my opinion that 125Hz potentiometer driven input is sunsetting. From the earliest iterations arcade cabinets have been known to offer Hall Effect technologies and comparatively high refresh rates to the console counterparts, with OLED technology having far more in common with CRT (Cathode Ray Tube) technology than traditional LCD (Liquid Crystal Display) panels, the market will inevitably mirror past instances. Hence the obsession with input. Whist corporate think-tanks may spend millions on marketable ideas, my goal is to ensure that Project Alpha feels cutting-edge in 203X and still above-standard into 204X. Quantum Break (Remedy, Microsoft, 2016), for instance, is limited to 1440p and clearly designed for upscaling and 30fps feeling ancient despite a fairly unique cross-medium approach (pioneered by Defiance, Trion Worlds/SyFy, 2013) when compared to Rise of the Tomb Raider (Crystal Dynamics, Square Enix, 2015), despite releasing a year after. This sets a clear precedent for future-facing technical decisions superceding creative decisions for long-tail revenue generation and brand-identity. Or at the very least, that the decisions should be made harmoniously.

Currently I have in for benchmarking purposes three separate SKU’s:

• Logitech Superlight 2

• Logitech Superstrike X2

• Razer Viper V4 Pro

The current separation comes from a differing of ideologies, Razer with Frame-Sync and Logitech with the Haptic Inductive Trigger System. The Superlight 2 will be used as a scientific control, as the most used mouse by professional e-sports athletes. Currently all mice are using stock skates on the Pulsar Paraspeed pad, but once stock for the newly released models is sourced, all skates will be X-Raypad Obsidian. Both Paraspeed pad and Obsidian skates rate approximately 75% on a scale of slow-to-fast. The primary goal of the current inventory is to ascertain the differences between methodologies, how each methodology changes the gameplay experience of targeted benchmark titles. The secondary goal is to ascertain how these technologies differ through A/B testing for future Project Alpha optimisation. A decision will need to be made if Banshee Interactive will optimise for haptic input, frame-sync, or both technologies (with the expectations that the dominant technology will be available in controllers shortly). How the player interfaces with Project Alpha is the most immediate variable and will decide almost all courses of action from AI behaviour patterns, biome layouts, and pacing, which will inevitably coincide with narrative beats. Whist Project Alpha has no immediate e-sports ambitions, the current goal of benchmarking is to have e-sports-level HUI as the foundation of all Banshee Interactive software from day-zero, prior to engine adoption.

The Logitech Superlight family is considered the standard for competitive mice since launch, despite stiff competition from Razer. Whilst there are brands such as Endgame Gear, Zowie (by BenQ) and more recently Pulsar, both the market and player adoption skews towards the Superlight, with the Hero 2 dominating theoretical specifications prior to the DeathAdder V4, now a largely irrelevant (for our purposes) SKU with the release of the Viper V4. As such the G-Pro X 2 Superlight will be where baseline evaluations are made, and deviations from that experiences will be noted. The Logitech mice have been calibrated using G Onboard Memory Manager, and the Viper V4 with Razer Synapse Web (Beta). All mice have been set to 3,200 DPI (3840×2160p canvas) at 2,000 Hz polling (i9 14900K CPU/ 240Hz) with default LOD (lift off distance). The Superstrike Mouse 1 input (left click) has been set to 0.3mm, following the current meta, and Mouse 2 input (right click) has been set to 0.5mm to mitigate against ghost input. Whilst the competition from secondary manufacturers is high, the choice to refrain from testing budget alternatives has been made, despite a curiosity over legitimacy of entry-level offerings. This is made in part so as manufacturer variance (such as implementation of the PAW 3950/3350) does not skew results, and because of the sheer market penetration of the Superlight.

It is also worth mentioning that I personally daily a Logitech Superlight 2 Dex with X-Raypad Obsidian skates and that my opinion is that the current market has reached a point of diminishing returns. These peripherals are what I consider the culmination of what was essentially a four-year industry arms-race.

Response time is unique as a singular point where interactive media has actively regressed, much like digital sensors and film, where a mature ecosystem (CRT/wired/hall effect, gated & micro switch arcade cabinets) was set aside for resolution and size of LCD technologies, along with the convenience of wireless peripherals. With resolution came three tedious generations of target 30 frames-per-second gameplay on 30-plus millisecond response panels, through 125Hz peripherals. If you have ever played a “retro” platformer or 2D fighter on “modern” systems, you will know how excruciatingly difficult they can be. This is not the Mandela Effect, but due to initial keyframe optimisation for near instantaneous response times of sub 20 milliseconds end-to-end. What was once a fair but demanding title such as Super Ghouls n Ghosts now makes “Hardest Games of All Time” lists. With rapid trigger and Razer pioneering snap tap (and Wooting dominating market perception), we have returned to what made Street Fighter II the competitive meta for decades, instant frame cancellation.

To expand on this thesis in contemporary context I have played through specific titles that I consider comparative to Project Alpha in various degrees, that either instigated change or were the last commercial outings of a specific, hard-boiled, mil-adjacent adult entertainment before the shooter adapted to PG-13 arena and hero shooters (Which, as an addendum, have more failed titles than successful). Alan Wake II is what I consider the current watershed title, but having never played Quantum Break on Xbox One, this was the first title I comparatively benchmarked. Unfortunately on this title my thesis holds, not only are the “TV” segments notoriously difficult to locate, but are also presented in what is now comically low resolution, whilst the game itself is clearly intended to run at 1440p upscaled, likely from a base count of 720p (As was the norm for Xbox One). Despite recent iterations of Northlight (Remedy’s proprietary engine) being highly optimised for current standards, Quantum Break fails to make use of greater-than 30 frames-per-seconds, with combat and animations very likely created to mask what is an inherent 30-plus millisecond handicap from framerate alone — bearing in mind that 1,000Hz polling, 120 frames-per-second, instantaneous magnetic switches and OLED technologies cleanly place end-to-end input back in the sub-20 millisecond range, if not in the “magic” sub-13, where it becomes imperceptible. Even on base peripherals such as the Xbox Series controller at 125Hz polling, the game is agonising to play. Bear in mind Quantum Break launched a whole year after The Witcher 3, which is still the go-to title for its given genre, and a year prior to Resident Evil VII which is still highly played, streamed, recommended, and an entryway into Capcom’s significant catalogue. The main personal takeaway from playing Quantum Break, is that it is unfairly and frustratingly difficult in 2026 even on medium difficulty due to the unresponsive nature of the (prior version) engine and contemporary expectations.

This article will be expanded as benchmarking progresses.

Benchmarking Mice

David Roberts, Creative Director