What Is Polling Rate?
Polling rate is how often a controller reports its current state to the host device, measured in hertz (Hz). A 1000Hz controller sends data every 1 millisecond; a 125Hz controller every 8 milliseconds. Higher polling rates can reduce input latency, but the relationship is not linear — jitter, connection type, and platform-level caps often matter more than the headline number.
What Polling Rate means
How Polling Rate Works
Controllers don't continuously stream data — they wait to be asked, then respond with their current button and stick state. The polling rate defines how often the host asks. This frequency is governed by a USB descriptor field called bInterval (for wired connections) or the equivalent Bluetooth connection interval — neither of which can exceed what the operating system and platform allow.
- 01
The host requests data on a fixed schedule
On a 1000Hz connection, the host operating system requests the controller's state every 1 millisecond. The controller responds with a packet containing button states, stick positions, and any sensor data. This request-response cycle is the polling, and its frequency is the polling rate.
- 02
USB bInterval declares the maximum frequency
Every USB device exposes a bInterval value in its descriptor — a single byte telling the host how often (in milliseconds) it can be polled. A bInterval of 1 means 1000Hz, 4 means 250Hz, 8 means 125Hz. Sony's stock DualSense ships with bInterval of 4; Microsoft's Xbox firmware hard-codes 8. This is why third-party controllers advertising 1000Hz only achieve it on PC where the OS can override device caps.
- 03
Higher Hz can reduce theoretical minimum latency
At 125Hz, input latency from polling alone is up to 8ms. At 250Hz, up to 4ms. At 1000Hz, up to 1ms. These are best-case theoretical minimums — actual latency adds USB stack scheduling, OS interrupt handling, game engine input processing, and display response time. Polling-rate reductions in latency are real but rarely the dominant component of perceived input lag.
- 04
Jitter — the stability of polling intervals — often matters more
A controller polling at a steady 250Hz with 4ms intervals feels more responsive than one polling at 1000Hz with intervals jumping between 0.5ms and 2.5ms. The standard deviation of the polling interval is called jitter. A 'long tail' of late packets causes the sensation gamers describe as 'heavy aim' or 'input mush' — even on controllers with high peak Hz ratings.
Polling Rate what your Hz means
Polling rate bands have very different interpretations depending on your platform and the connection type. The bands below reflect what's actually achievable on each platform — not marketing claims from controller boxes.
| Polling rate | Verdict | Meaning |
|---|---|---|
| 125 Hz (8 ms) | Standard for Xbox / default Bluetooth | Microsoft hard-codes Xbox first-party controllers (including Elite Series 2) to 125Hz at the driver stack level. Default Bluetooth HID also uses 125Hz. Adequate for RPGs, story games, single-player. Noticeable in competitive FPS or fighting games. |
| 250 Hz (4 ms) | Standard for PS5 / stock DualSense wired | Sony's PS5 caps DualSense polling at 250Hz via bInterval=4 in the USB descriptor. Stock DualSense wired to PC also reports 250Hz unless overclocked via HIDUSBF. Most 2.4GHz dongle controllers also land in this range. Solid responsiveness for nearly all gameplay. |
| 500 Hz (2 ms) | Steam Deck and mid-range premium | Steam Deck's internal controls poll at a stable 500Hz. Some premium third-party controllers (Backbone One on capable phones, certain wired Razer) also reach this range. Diminishing returns vs 250Hz for most players, but real improvement for competitive FPS. |
| 1000 Hz (1 ms) | Premium PC competitive standard | True 1000Hz polling requires PC with a controller that supports it natively (8BitDo Ultimate 2, GameSir G7 SE on Xbox, SCUF Omega on PC, etc.) or an overclocked stock controller via HIDUSBF. Genuine measurable improvement in competitive aim assist and rhythm games. |
| 8000 Hz (0.125 ms) | Marketing claim — verify before paying for it | Some 2025-2026 controllers (Razer Wolverine V3 Pro) advertise 8000Hz polling. The controller's USB descriptor can request this rate, but the internal sensor sampling often cannot update that fast — the controller sends duplicate packets to fill the extra slots. The benefit beyond 1000Hz is small even when achieved cleanly. |
Sony's PS5 enforces a 250Hz cap on DualSense polling via bInterval=4 in the USB descriptor. Microsoft hard-codes Xbox first-party controllers (including Elite Series 2) to 125Hz at the XInput driver stack level. Third-party '1000Hz' controller claims apply only when connected to a PC where the operating system can override the device cap.
Test for Polling Rate
Fix Polling Rate issues
Devices most affected by Polling Rate
Related glossary terms
Polling Rate questions
No. Sony's PS5 caps DualSense and licensed third-party controller polling at 250Hz via the bInterval=4 setting in the USB descriptor. Third-party PS5 controllers advertising '1000Hz' only achieve that rate when connected to a PC where the OS can override the device cap. On PS5 itself, every controller — whether stock DualSense, SCUF Omega, Razer Raiju V3 Pro, or Victrix Pro BFG Reloaded — reports state at 250Hz regardless of marketing claims.
Mostly no. Microsoft hard-codes Xbox first-party controllers (including the Elite Series 2) to 125Hz polling at the XInput driver stack level. Even premium Xbox-licensed third-party controllers typically stay at 125Hz on Xbox consoles. The notable exception is the GameSir G7 SE, which Microsoft specifically licensed to report at 1000Hz — a uniquely high polling rate among Xbox-licensed controllers. On PC, more Xbox controllers can reach higher polling rates.
Not always. Polling rate stability (low jitter) often matters more than peak Hz. A controller at 1000Hz with high jitter — where individual polling intervals vary widely — feels less responsive than a stable 250Hz controller. A 'long tail' of late packets causes what gamers describe as 'heavy aim' or 'input mush.' Verify both your average Hz AND jitter before assuming high-Hz marketing translates to gameplay benefit.
Almost never in practice. Some 2025-2026 controllers like the Razer Wolverine V3 Pro advertise 8000Hz polling, but the controller's internal sensor sampling rate usually cannot update that fast. The result: the controller sends duplicate or interpolated packets to fill the extra polling slots. You get 1ms USB report intervals but the actual hardware state changes more slowly. Real-world benefit beyond 1000Hz is small for most games.
Sony's USB descriptor sets stock wired DualSense to bInterval=4 (250Hz / 4ms). Modern Bluetooth 5.0 implementations can reach 1000Hz polling on capable PCs and phones. The intuition that 'wired is always faster' is wrong without overclocking the wired connection. To make wired faster than Bluetooth on PC, use a tool like HIDUSBF to override bInterval to 1 — overclocking the USB connection to 1000Hz.
Jitter is the standard deviation of the polling interval — how much the time between polls varies from packet to packet. A controller polling at a steady 250Hz with intervals consistently at 4ms feels smooth. A controller polling at 1000Hz with intervals jumping between 0.5ms and 2.5ms feels uneven, even though its peak Hz is higher. Industry rule: if jitter exceeds 30% of average Hz, your connection has a stability problem regardless of the headline number.
Yes for many controllers, with caveats. On Windows, a free tool called HIDUSBF can override the bInterval setting and force USB polling up to 1000Hz for compatible controllers (DualSense, DualShock 4, some Xbox via XInput workarounds). The tool works at the driver level — it doesn't modify the controller's firmware. Risks include connection instability, slightly higher CPU usage, and potentially shortened controller motherboard lifespan. Overclocking via software does not work on Xbox consoles, only on PC.
Further reading
- Bluetooth vs Wired Controller Latency: Input Lag Tested for Competitive Gaming · Gamepad Test App · Retrieved