For competitive online gamers, the conventional wisdom has long been unambiguous: run a VPN and watch your ping climb, your connection stutter, and your performance suffer. It is an assumption rooted in reasonable logic - routing traffic through an additional server adds distance, and distance adds latency. Except, as real-world testing across six major VPN providers reveals, that assumption does not always hold. In some configurations, a VPN can actually lower ping compared to a native, unprotected connection.
Why a VPN Can Reduce Latency, Not Just Add It
Understanding why requires a brief look at how internet routing actually works. When data leaves your device, it does not travel in a straight line to its destination. Your internet service provider (ISP) routes that traffic through its own infrastructure - a path that reflects the ISP's network architecture and commercial agreements, not necessarily the most direct route to the server you are trying to reach. A VPN introduces an alternative routing path. If the VPN's server sits geographically closer to the game server, or if its network path is less congested or more direct than the ISP's default route, the result can be a measurable reduction in round-trip time.
This is not theoretical. In benchmark testing conducted across multiple runs of Counter-Strike 2 Deathmatch in the UK, connected to the same game server throughout, the baseline ping without any VPN was 12.58 ms. Windscribe returned an average of 12.34 ms. ExpressVPN came in even lower at 12.22 ms. These are not dramatic differences, but they are consistent across repeated runs - which is the meaningful part. A single data point could be noise. A pattern across multiple sessions is signal.
What the Numbers Actually Show
The benchmark data covers four distinct performance dimensions, each relevant to a different aspect of the online gaming experience.
- Local ping (CS2, UK server): ExpressVPN led at 12.22 ms, followed by Windscribe at 12.34 ms and Mullvad at 12.55 ms. NordVPN sat marginally above the no-VPN baseline at 12.73 ms. Proton performed worst at 13.8 ms.
- Jitter (Cloudflare speed test): The no-VPN baseline was actually the best here at 10.66 ms. ExpressVPN was the closest VPN option at 12.07 ms. Proton VPN was a significant outlier at 20.7 ms - more than double the baseline - which matters considerably for real-time responsiveness.
- Download speed (15.3 GB via Steam): Differences were marginal. The no-VPN baseline completed in 271 seconds; most VPNs landed between 275 and 276 seconds. ExpressVPN was the slowest at 287 seconds.
- Distance ping (CS2, US server from UK): All VPNs performed closely, ranging from NordVPN at 85.13 ms to Mullvad at 87.69 ms - a narrow spread suggesting comparable international routing capability across providers.
Jitter deserves particular attention. Ping measures average round-trip time; jitter measures the variance in that time from packet to packet. High jitter produces the erratic, inconsistent feel that competitive players find far more disruptive than a uniformly slightly elevated ping. Proton's jitter result of 20.7 ms - against a baseline of 10.66 ms - is the most practically significant negative finding in the entire dataset.
Compatibility Remains the Real Obstacle
Performance is only one part of the equation. Several titles, particularly those with aggressive anti-cheat or account security systems, actively resist VPN connections. During extended testing, both Windscribe and Surfshark failed to allow Valorant to run at all - blocking login to the Riot client and preventing the game from launching. This is not a minor inconvenience for players who rely on that title.
The practical workaround is split tunnelling, a feature present in all tested VPNs that allows specified applications to bypass the VPN while everything else remains protected. In practice, this means identifying the relevant executables - including background processes that may not be immediately obvious - and manually excluding them. It functions, but it represents a meaningful configuration burden and effectively means those applications are operating outside the privacy protections the VPN is supposed to provide.
NordVPN was the only provider to pass compatibility testing across all titles tested without requiring split tunnelling workarounds, which gives it a practical advantage for players who want comprehensive coverage without managing exclusion lists. The trade-off is a marginally higher local ping than Windscribe or ExpressVPN - though the difference of roughly 0.15 ms is functionally imperceptible in play.
The Broader Case for Reconsidering VPNs in Gaming
Beyond raw performance, there are practical reasons a competitive player might want VPN coverage. ISP throttling of gaming traffic is a documented behaviour in various markets - some providers have been found to selectively deprioritise traffic to gaming servers during peak hours, particularly for high-bandwidth titles. A VPN obscures the nature of the traffic, which can circumvent that throttling. It also provides protection against distributed denial-of-service attacks, a real concern for streamers or players in ranked environments where opponents have been known to target connections directly.
The conclusion drawn from the data is not that VPNs are without cost for gaming - jitter increases and compatibility friction are real. But the assumption that any VPN inevitably degrades performance badly enough to rule it out does not survive contact with actual testing. For players already running a VPN for legitimate privacy reasons, the question is no longer whether gaming is compatible with that choice. With the right provider and, where necessary, considered use of split tunnelling, it largely is.
