The assumption that VPNs and online gaming are fundamentally incompatible turns out to be wrong. Hands-on benchmarking across multiple leading VPN services found that some can reduce latency below what a direct ISP connection delivers - a result that challenges one of the most persistent pieces of received wisdom in PC gaming circles. The implications reach beyond competitive play into the broader question of why VPN use has carried an unfair performance stigma for so long.
Why VPNs Were Dismissed - And Why That Logic Was Incomplete
The conventional objection is straightforward: a VPN routes your traffic through an additional server, adding physical and computational distance between you and the game server. More hops, more latency. It sounds airtight. But that framing misses a critical variable - your ISP's own routing decisions are not automatically optimal.
Internet traffic does not travel in a straight line. When you connect to a game server, your ISP determines the path your packets take, and that path depends on peering agreements, network load, and infrastructure that has nothing to do with your proximity to the destination. A VPN server, by contrast, may sit on a different part of the network backbone, one that connects to a game server's data centre more directly. If that alternative route is shorter or less congested, your packets arrive faster - even though they technically travelled through an extra node.
This is not a theoretical possibility. It happened measurably in real testing. With Windscribe, average ping in Counter-Strike 2 dropped from 12.58 ms on a bare connection to 12.34 ms. With ExpressVPN, it fell further to 12.22 ms. These are not rounding errors - they held up across multiple runs. Even NordVPN, which produced a marginally higher average of 12.73 ms, remained within a margin so narrow as to be functionally irrelevant in most play scenarios.
What the Benchmarks Actually Show
The data collected spans four distinct measurements, each testing a different aspect of network performance that matters to online play.
- Local ping (CS2, same country): ExpressVPN led at 12.22 ms, followed by Windscribe at 12.34 ms and Mullvad at 12.55 ms - all below the baseline 12.58 ms. Surfshark and Proton sat slightly higher.
- Jitter (Cloudflare speed test): The baseline was 10.66 ms - lower than any VPN achieved. Proton was a clear outlier at 20.7 ms. ExpressVPN came closest at 12.07 ms. Jitter measures the consistency of your connection rather than its speed; high jitter causes erratic hit registration and visual stuttering that raw ping numbers do not capture.
- Download speed (15.3 GB Steam download): All VPNs slowed the download slightly, with results ranging from 275 to 287 seconds against a bare 271. ExpressVPN was the slowest here. The overhead is real but modest - under 6% in the worst case.
- Distance ping (CS2, US server from UK): When connecting to a server on a different continent, VPN routing differences compress. All six VPNs tested clustered between 85.13 ms and 87.69 ms, with NordVPN and ExpressVPN performing marginally better.
Jitter deserves more attention than it typically receives. A connection with 12 ms average ping but volatile jitter will feel worse in fast-paced play than one with 14 ms ping and consistent delivery. On this measure, the bare connection won, which means VPN use is not a universal improvement - it depends heavily on what aspect of performance matters most to you.
Compatibility Breaks the Equation for Some Services
Raw latency numbers are only useful if the VPN actually lets you connect to the games you play. That is where meaningful differences emerged. Both Windscribe and Surfshark blocked access to Riot Games titles entirely during testing - Valorant could not be launched, and the Riot launcher itself failed to authenticate. This is not a quirk of a single game; Riot operates an aggressive anti-cheat system called Vanguard that runs at the kernel level and is sensitive to network anomalies associated with VPN use.
A workaround exists in the form of split tunnelling - a feature available on all VPNs tested that allows specific applications to bypass the encrypted tunnel while everything else remains protected. By adding the Riot launcher and game executable to an exclusion list, it is possible to run Valorant outside the VPN while keeping other traffic routed through it. This functions adequately as a practical solution, but it is worth being clear about what it represents: you are not running the game through the VPN. You are running it the same way you would without one.
NordVPN handled Valorant without requiring that workaround, which makes it the more complete option for players who want genuine whole-system coverage rather than selective routing. For players who only use Steam titles and avoid Riot's ecosystem, Windscribe presents a compelling case - lower average ping than NordVPN, credible privacy credentials, and a competitive price point.
The Broader Case for Reconsidering VPNs in Gaming
There are legitimate, non-performance reasons why a gamer might want VPN coverage. Some ISPs have been documented throttling traffic to specific services, including game download servers and multiplayer infrastructure, particularly during peak hours. A VPN obscures the nature of your traffic, making selective throttling harder to apply. If your connection degrades predictably in the evenings, that is worth investigating before dismissing VPN use as irrelevant to your situation.
There is also the question of DDoS exposure. High-profile streamers and competitive players have historically been targeted by denial-of-service attacks that exploit their visible IP addresses. Routing through a VPN masks your real IP from other participants in a lobby or from anyone watching a stream - a meaningful protection for a specific subset of users.
The conclusion is not that every gamer needs a VPN. It is that the categorical refusal - "VPNs kill ping, full stop" - is no longer defensible as a general rule. Whether a VPN helps, hurts, or makes no difference depends on your ISP's routing, your geographic relationship to game servers, and which games you play. Testing remains the only honest method. But the data here is clear: in the right conditions, the encrypted detour can be shorter than the direct road.
